CN1308654C

CN1308654C - Position testing method and its device

Info

Publication number: CN1308654C
Application number: CNB2003101199813A
Authority: CN
Inventors: 岡山喜彦
Original assignee: Azbil Corp
Current assignee: Azbil Corp
Priority date: 2002-11-28
Filing date: 2003-11-28
Publication date: 2007-04-04
Anticipated expiration: 2023-11-28
Also published as: JP3858994B2; CN1504721A; JP2004177335A

Abstract

The invention provides an edge detection method and device capable of performing position detection with high precision according to the light intensity distribution on the light receiving surface generated by Fresnel diffraction. The hyperbolic sine function sech(x) is used to approximate the change in light intensity at the first rising portion of the light intensity distribution on the light receiving surface of the line sensor caused by Fresnel diffraction of monochromatic parallel light, and the hyperbolic sine function is used to analyze the line The intensity of light received by each light-receiving element of the shape sensor. In particular, the light-receiving element for which the light-receiving intensity sandwiched by the reference light-receiving intensity [0.25] is obtained, and the light-receiving position at which the light-receiving intensity is obtained is calculated on the light-receiving surface of each of the light-receiving elements based on a hyperbolic sine function approximating the light intensity distribution. . Then, based on the above-mentioned light-receiving position, the position that becomes the reference light-receiving intensity is obtained by interpolation calculation, and is used as the edge position of the above-mentioned blocking object.

Description

Method for detecting position and position detecting device

Technical field

The present invention relevant can be high-speed, high Precision Detection is for example from the reel uncoiling and towards the method for detecting position and the position detecting device of the change in location of folk prescription on the edge part Width of the shoestring that transmits at a high speed.

Background technology

As the position detecting device of the marginal position that detects film and thin slice class article have the pick-up unit of optical profile type, this device comprise to the light-projecting portion (light source) of article (inspection target object) irradiation directional light, and with the light accepting parts such as CCD (line sensor) of this light-projecting portion subtend setting.The position detecting device of this optical profile type, basically be to accept the directional light do not blocked by above-mentioned article at light accepting part, just the border of the light area of directional light and non-light area (lightproof area) is detected as the marginal position of described article (inspection target object) on this light accepting part.

In addition, advocate a kind of pick-up unit recently, this device is to utilize monochromatic collimated beams such as laser, and be conceived to the Fresnel diffraction of the described monochromatic collimated beam that produces on the edge of article (inspection target object), according to the light intensity distributions on the sensitive surface of described line sensor (light accepting part), detect the marginal position of described article (inspection target object) accurately, the device that proposes in the open communique (spy opens flat 8-247726 number) of Japan's patented claim is for example arranged.

But, when the light intensity distributions that produces at the Fresnel diffraction that utilizes monochromatic collimated beam detects the marginal position of checking target object, must obtain above-mentioned distribution of light intensity characteristic in advance accurately on the sensitive surface of line sensor (light accepting part).Additional disclosure once, the light intensity distributions that above-mentioned Fresnel diffraction produces sharply rises near marginal position as shown in Figure 8, and along with restraining while vibrating away from marginal position.Be λ at the wavelength of hypothesis monochromatic collimated beam, be marginal position on z, the sensitive surface during for [x=0] from checking target edges to the distance of sensitive surface, with ∫ as expression from [x=0] to [(2/ λ z) ^1/2X] the sign of operation of integration, the characteristic of then such light intensity distributions can be represented with following formula.

Light Strong degree=(1/2) { [1/2+S (x)] ²+ [1/2+C (x)] ²}

S(x)＝∫sin(π/2)·U ²dU

C(x)＝∫cos(π/2)·U ²dU

In the formula, U is the parameter of hypothesis.And, to establish when the convergent light intensity is for [1.00] on the sensitive surface, the light intensity (relative value) of marginal position [x=0] becomes [0.25].

Also have,, then shown in special mathematical formulae, utilize the Fresnel function to be similar to respectively with following formula about above-mentioned function S (x), C (x).

S(x)’(1/2)-(1/πx)cos(πx ²/2)

C(x)’(1/2)+(1/πx)sin(πx ²/2)

Therefore, basically by with above-mentioned approximate expression S (x) ', C (x) ', can calculate described marginal position from the light intensity that is subjected to of each photo detector generation of described line sensor.

But, in fact try as if calculating, then undeniable, as shown in figure 22, it is very approximate that function S (x), C (x) rise later contraction section (after the 2nd peak) thereon with its approximate expression S (x) ', C (x) ', but at initial rising part (the 1st peak) bigger deviation is arranged.Especially the characteristic of this initial rising part is being undertaken important effect in rim detection, and the skew of its characteristic will become the main cause that the marginal position accuracy of detection reduces.

On the other hand, the line sensor that detects described monochromatic collimated beam has a plurality of photo detectors that the regulation sensitive surface will be arranged spaced component structure with regulation.Specifically, the line sensor of general cheapness for example has the spaced structure that 102 photo detectors that the sensitive surface of 85 μ m * 75 μ m is arranged are pressed 85 μ m.And utilize the signal and the corresponding output of its position of components that each photo detector will be suitable with being subjected to light light quantity (light intensity) respectively.

Therefore, undeniable strictly speaking, the light intensity distributions that produces from the output of above-mentioned line sensor gained and actual Fresnel diffraction has some differences.Promptly, for the correct light intensity distributions that detects Fresnel diffraction, just the monochromatic collimated beam that the produces Fresnel diffraction light intensity to each point place on the straight line of sensitive surface seamlessly must be detected, but the photo detector of line sensor has the sensitive surface of certain area as mentioned above, output be the signal of light quantity (light intensity) that is equivalent to whole summations (integration) of suffered light on the sensitive surface.Thereby the light intensity distributions that detects by line sensor is according to the arrangement pitch of photo detector and the distribution of stepped variation.So though can utilize the arrangement pitch of photo detector is too narrow to about 7 μ m, resolution is high line sensor to improve its measuring accuracy in the past, undeniable, it is very expensive that line sensor itself also just becomes.

The present invention system In view of the foregoing proposes, its purpose is to provide a kind of can carry out edge position detection and the device that high precision position detects, this method and device can be similar to the light intensity distributions on the sensitive surface of Fresnel diffraction generation, the characteristic of particularly initial rising part accurately.

Another purpose of the present invention is for providing a kind of edge detecting device, though this device when the line sensor that the arrangement pitch that utilizes a plurality of photo detectors is dredged, still can high precision, detect marginal position at a high speed.

Summary of the invention

For reaching above-mentioned purpose, method for detecting position of the present invention is applicable to the position detecting device with following formation, it include along folk prescription to the line sensor (light accepting part) of in accordance with regulations spaced a plurality of photo detectors, with this line sensor subtend setting and the configuration direction of shelter that exist in the light path that detects described monochromatic collimated beam towards the light-projecting portion of above-mentioned a plurality of photo detectors projection monochromatic collimated beams of this line sensor and the output of resolving described line sensor at described photo detector on the rim detection portion of marginal position.

Especially in described rim detection portion, it is characterized by the intensity variation of utilizing hyperbolic sine function sech (x) to be similar to rising part initial in the light intensity distributions that described shelter produces because of the Fresnel diffraction of monochromatic collimated beam on the sensitive surface of described line sensor, utilize this hyperbolic sine function sech (x) resolve described line sensor each photo detector be subjected to light intensity, obtain the marginal position of described shelter.

That is, method for detecting position of the present invention finds following relation and forms, i.e. the initial rising part of the light intensity distributions that on sensitive surface, produces of the Fresnel diffraction of monochromatic collimated beam, and particularly the distribution character of its 1st peak-to-peak extremely is approximated to

y＝a/cosh(bx+c)

The inverse of the hyperbolic cosine function cosh (x) of (establish a, b, c is respectively coefficient) is promptly with hyperbolic sine function sech (x).And, it is characterized by, utilize this hyperbolic sine function sech (x) to resolve the output (light intensity) of described line sensor, in the distribution of light intensity, detecting light intensity (relative value) becomes the marginal position of the position [x=0] of [0.25] as described shelter on the sensitive surface that described Fresnel diffraction produces.

Best, utilize described hyperbolic sine function sech (x) resolve described line sensor each photo detector be subjected to light intensity, for example it is characterized in that,

(the 1st stage), obtain acquisition respectively and be subjected near this benchmark of light intensity [0.25] to be subjected to the big photo detector that is subjected to light intensity of light intensity and to obtain to be subjected to the little photo detector that is subjected to light intensity of light intensity than said reference than predetermined benchmark,

(the 2nd stage) is according to the inverse function ln{[1+ (1-Y of hyperbolic sine function sech (x) ²) ^1/2]/Y} obtains respectively after becoming this light receiving position that is subjected to light intensity on the sensitive surface of these photo detectors,

(the 3rd stage), obtaining from above-mentioned light receiving position becomes the position that described benchmark is subjected to light intensity, as the marginal position of described shelter.

In addition, position detecting device of the present invention comprises, have along the line sensor of folk prescription in accordance with regulations spaced a plurality of photo detectors, be provided with this line sensor subtend and to the light-projecting portion of described a plurality of photo detectors projection monochromatic collimated beams of this line sensor, and and according to because of described shelter since the output that the light intensity distributions that the Fresnel diffraction of monochromatic collimated beam produces on the sensitive surface of described line sensor is resolved aforementioned line sensor with the rim detection portion of the marginal position of shelter in described photo detector orientation that exist in the light path that detects described monochromatic collimated beam, it is characterized in that, particularly following device is set as described rim detection portion

Photo detector is determined device, is used for determining to obtain predetermined benchmark respectively according to the output of described line sensor and is subjected to be subjected to the big photo detector that is subjected to light intensity of light intensity and to obtain to be subjected to the little photo detector that is subjected to light intensity of light intensity than described benchmark than this benchmark near the light intensity [0.25];

The light receiving position calculation element is used for according to the inverse function ln{[1+ (1-Y by hyperbolic sine function sech (x) ²) ^1/2The approximate light intensity distributions of]/Y} is obtained the light receiving position that is subjected to light intensity that becomes this photo detector respectively in the sensitive surface of determining the photo detector that device is determined with described photo detector; And

The interpolation operation device is used for detecting according to the light receiving position that this light receiving position calculation element is obtained respectively and becomes described benchmark and be subjected to the marginal position of the position of light intensity as described shelter.

Best described photo detector determines that device constitutes like this, promptly after the output normalization with described line sensor in advance, the definite acquisition is subjected to the big photo detector that is subjected to light intensity of light intensity [0.25] and obtains to be subjected to the little photo detector that is subjected to light intensity of light intensity [0.25] than described benchmark than described benchmark, is specifically to determine to obtain described benchmark to be subjected near at least two photo detector Cn, the Cn-1 mutually adjacent each other that are subjected to light intensity of light intensity [0.25].

Position detecting device according to such formation, but because the hyperbolic sine function sech (x) of the initial rising part characteristic by the light intensity distributions that produces with the approximate Fresnel diffraction of high precision, can obtain accurately respectively and obtain at least benchmark is subjected to the light receiving position that is subjected to light intensity that becomes this photo detector in the photo detector that is subjected to light intensity of light intensity [0.25] in the middle of being clipped in, so, can obtain the marginal position on the sensitive surface of line sensor according to these light receiving positions accurately, promptly be subjected to light intensity to become the position of [0.25].And, for the inverse function ln{[1+ (1-Y of above-mentioned hyperbolic sine function sech (x) ²) ^1/2]/Y}, because of carrying out computing to it according to series expansion or according to the instruction of being adorned among the CPU, so its calculation process speed (position probing speed) is very fast.

Also have, also can obtain the photo detector of obtaining peak value at first and hithermost photo detector thereof respectively, according to the inverse function ln{[1+ (1-Y of described hyperbolic sine function sech (x) in line sensor output ²) ^1/2]/Y} obtains marginal position from the light intensity that respectively is subjected to of above-mentioned each photo detector.Detect if carry out marginal position like this,, in the time of can not covering monochromatic collimated beam fully by the detected object object, still can detect the edge part position of described detected object object accurately even constitute by the translucent body group at the detected object object.

In addition, being characterized as of another kind of method for detecting position of the present invention, comprise and having at least at folk prescription to by the line sensor of certain spaced a plurality of photo detectors and to the light-projecting portion of this line sensor projection monochromatic collimated beam, when the light intensity distributions that produces on described line sensor sensitive surface because of the Fresnel diffraction of described monochromatic collimated beam on the shelter edge that exists in the light path according to above-mentioned monochromatic collimated beam detects the marginal position of described shelter, carry out following detection and handle.

Be method for detecting position of the present invention, it is characterized by

(1) after the output with described line sensor is normalized into [1], obtain acquisition respectively and be subjected to the big photo detector Cn that is subjected to light intensity of light strength ratio [0.25] and obtain the described little photo detector Cn-1 that is subjected to light intensity of light strength ratio [0.25] that is subjected to,

(2) then, with the function of the light intensity distributions on the described line sensor sensitive surface that utilizes the described Fresnel diffraction of expression the to produce light intensity of hyperbolic sine function sech (x) expression for example, each arrangement pitch to each photo detector carries out integration, according to light income A (xn), the A (xn-1) of the described photo detector Cn, the Cn-1 that obtain, obtain center Xn, the Xn-1 of this photo detector Cn, Cn-1 respectively.

, according to center Xn, the Xn-1 of these photo detector Cn, Cn-1, utilize interpolation handle, obtain the described marginal position x0 that is subjected to light intensity to become [0.25] (3) thereafter.

Another kind of method for detecting position of the present invention also can be constructed as follows,

(1) after the output with described line sensor is normalized into [1], obtains acquisition respectively and be subjected to the big photo detector Cn that is subjected to light intensity of light strength ratio [0.25] and obtain the little photo detector Cn-1 that is subjected to light intensity of described suffered light strength ratio [0.25].

(2) then, position xn, xn-1 according to the element arrangements direction that is subjected to light intensity A (xn), A (xn-1) and each photo detector Cn, Cn-1 of these photo detectors Cn, Cn-1, utilize the described Fresnel diffraction of expression at the described function that is subjected to the light intensity distributions that produces on the sensitive surface of optical sensor hyperbolic sine function sech (x) for example, obtain respectively described be subjected to the position x of the element arrangements direction that light intensity becomes [0.25] and from the described sensitive surface of optical sensor that is subjected to the distance z between the edge of described shelter.

(3) then, utilize these positions x and distance z, the sensitive surface of described each photo detector Cn, Cn-1 is being regarded as when a bit, that utilizes that described function obtains this photo detector Cn, Cn-1 respectively is subjected to light intensity Yn, Yn-1, simultaneously described function is carried out integration to each arrangement pitch of each photo detector, obtain light income yn, yn-1 after the sensitive surface upper integral of each photo detector Cn, Cn-1 respectively.

(4) then, the difference Δ yn, the Δ yn-1 that described each photo detector Cn, Cn-1 are seen as light income yn, yn-1 after the sensitive surface upper integral of light income Yn, Yn-1 when some and each photo detector Cn, Cn-1 are as correcting value, and be subjected to light intensity A (xn), the A (xn-1) to described photo detector Cn, Cn-1 proofreaies and correct respectively.

(5) after this, be subjected to light intensity A (xn) ', A (xn-1) ', and position xn, the xn-1 of the element arrangements direction of these photo detectors Cn, Cn-1 according to described each photo detector Cn, Cn-1 after proofreading and correct, use described approximate function once more, obtain respectively and describedly be subjected to position x that light intensity becomes [0.25] and to the distance z at the edge of described shelter.

Also have, the correction that is subjected to light intensity A (xn), A (xn-1) about described photo detector Cn, Cn-1, also can be that difference Δ y ' n, y ' n-1 that each photo detector Cn, Cn-1 see light income yn, yn-1 after the sensitive surface upper integral of light income A (xn) ', A (xn-1) ' behind the point calibration and each photo detector Cn, Cn-1 as are obtained as new correcting value, and carry out repeatedly, utilize corrected repeatedly light intensity A (xn) ', the A (xn-1) ' of being subjected to, obtain the position x and the distance y of described shelter.

According to this method for detecting position, even for example when the line sensor of the cheapness of using element arrangements to dredge at interval, because can proofread and correct the light income on the face of photo detector of line sensor, with its light income be subjected to the situation Approximate Equivalent of light in the mode of point in fact, so and utilize the light intensity distributions of Fresnel diffraction for example that hyperbolic sine function sech (x) high precision approximation method combines, also can improve its measuring accuracy fully.

In addition, position detecting device of the present invention comprises, have at least folk prescription to the line sensor of in accordance with regulations spaced a plurality of photo detectors, towards the light-projecting portion of this line sensor projection monochromatic collimated beam, and obtain the pick-up unit of the light intensity distributions that produces on the sensitive surface of Fresnel diffraction in described line sensor of edge part because of described monochromatic collimated beam of the shelter that exists on the light path of described monochromatic collimated beam according to the output of described line sensor, particularly it is characterized by; Comprise following device,

(1) photo detector is determined device, is used for be normalized into by the output of optical sensor to obtain respectively when [1] to obtain to be subjected to the big photo detector that is subjected to light intensity of light strength ratio [0.25] and to obtain above-mentioned light strength ratio [0.25] little the photo detector Cn, the Cn-1 that are subjected to light intensity of being subjected to described;

(2) position of components pick-up unit, be used for the described Fresnel diffraction of expression the described function that is subjected to the light intensity distributions that produces on the sensitive surface of optical sensor to each arrangement pitch of each photo detector carry out integration, and according to light income A (xn), center Xn, the Xn-1 that A (xn-1) obtains this photo detector Cn, Cn-1 respectively of described each photo detector Cn, the Cn-1 that obtain like this;

(3) marginal position pick-up unit is used for utilizing the interpolation processing to obtain the described marginal position x0 that is subjected to light intensity to become [0.25] according to center Xn, the Xn-1 of these photo detectors Cn, Cn-1.

Or being characterized as of position detecting device of the present invention, comprise following device

(1) photo detector is determined device, is used for be normalized into by the output of optical sensor to obtain respectively when [1] to obtain to be subjected to big photo detector that is subjected to light intensity of light strength ratio [0.25] and acquisition to be subjected to light strength ratio [0.25] little the photo detector Cn, the Cn-1 that are subjected to light intensity with described;

(2) edge detecting device, be used to utilize expression because of described Fresnel diffraction the described function that is subjected to the light intensity distributions that produces on the sensitive surface of optical sensor for example hyperbolic sine function sech (x), according to position xn, the xn-1 of the element arrangements direction that is subjected to light intensity A (xn), A (xn-1) and each photo detector Cn, Cn-1 of described each photo detector Cn, Cn-1 obtain respectively described be subjected to position x that light intensity becomes [0.25] and from the described sensitive surface of optical sensor that is subjected to the distance z between the edge of described shelter;

(3) the 1st are subjected to light intensity to calculate device, utilize the position x that obtains by this edge detecting device and distance z according to described function obtain respectively with the sensitive surface of described each photo detector Cn, Cn-1 see as this photo detector Cn, Cn-1 when some be subjected to light intensity Yn, Yn-1,

(4) the 2nd are subjected to light intensity to calculate device, are used to utilize the position x that obtained by described marginal position calculation element and distance z and described function carried out integration to each arrangement pitch of each photo detector and obtain respectively and be subjected to light intensity yn, yn-1 after the sensitive surface upper integral of described each photo detector Cn, Cn-1;

(5) light income means for correcting is used for being subjected to light intensity A (xn), A (xn-1) what the difference Δ yn, the Δ yn-1 that each photo detector Cn, Cn-1 are seen as light income Yn, Yn-1 when some and light income yn, yn-1 after the described sensitive surface upper integral found the solution and proofreaied and correct respectively described each photo detector Cn, Cn-1 as correcting value; And

(6) operation control device repeatedly is used for position xn, the xn-1 of the element arrangements direction that is subjected to light intensity A (xn) ', A (xn-1) ' and these photo detectors Cn, Cn-1 of corrected described photo detector Cn, Cn-1 are supplied with described edge detecting device and carry out the computing of described marginal position once more.

Position detecting device according to such formation, even when the linear transducer of the cheapness of for example using element arrangements to dredge at interval, because can proofread and correct the light income on the face of photo detector of line sensor, its light income is subjected to the situation roughly equiv of light in fact and with the form of putting, so also can improve its measuring accuracy fully.And, because of can be with the light intensity distributions of approximate Fresnel diffraction accurately of hyperbolic sine function sech (x) for example, thus can be simply and carry out its calculation process at high speed, the formation of whole device can be simplified more, and cost further reduces.Inverse function ln{[1+ (1-Y for hyperbolic sine function sech (x) ²) ^1/2]/Y}, additional disclosure be because can be with it by series expansion, or carry out computing, so its calculation process speed (position probing speed) is very high according to the instruction that is contained among the CPU.

Description of drawings

Fig. 1 is the basic comprising figure of the position detecting device of expression the present invention one example.

The synoptic diagram that Fig. 2 arranges for the photo detector of expression line sensor.

Fig. 3 is for the theoretical value of the light intensity distributions that will produce because of Fresnel diffraction and utilize the approximate characteristic of function to make the synoptic diagram of contrast.

Fig. 4 is the synoptic diagram of the edge detection process step 1 example of the method for detecting position of expression the present invention one example and position detecting device.

Fig. 5 be illustrated in try to achieve in two adjacent photo detectors be subjected to light intensity, and according to the graph of a relation between the marginal position that obtains trying to achieve the above-mentioned position that is subjected to light intensity.

Fig. 6 is the synoptic diagram of expression because of the different sensor output examples that produce of resolution of line sensor.

Fig. 7 is used to illustrate that the detected object object is the synoptic diagram of the effect that detects of the situation lower limb of translucent body.

The light intensity distributions performance plot that Fig. 8 produces because of Fresnel diffraction for expression.

Fig. 9 is the basic comprising figure of the position detecting device of the other example of expression the present invention.

Figure 10 is the concise and to the point pie graph of the light-projecting portion of expression position detecting device shown in Figure 9.

Figure 11 analyses and observe the synoptic diagram that the direction light-projecting portion penetrates the optical system of parallel beam for the A-A from Figure 10.

Figure 12 analyses and observe the synoptic diagram that the direction light-projecting portion penetrates the optical system of parallel beam for the B-B from Figure 10.

Figure 13 is the synoptic diagram of edge detection process step 1 example basic in expression method for detecting position of example shown in Figure 9 and the position detecting device.

Figure 14 is illustrated in being subjected to light intensity and tried to achieve between relation between marginal position by the position of light intensity according to obtaining these of trying to achieve on the two adjacent photo detectors.

Figure 15 is the concept map of the calculation process of the distance z of expression relevant calculation edge and sensitive surface.

The synoptic diagram of Figure 16 for comparing as the light intensity distributions that produces because of Fresnel diffraction that is subjected to distribution of light intensity and reality that the output of line sensor is obtained.

Figure 17 is subjected to Error Graph between distribution of light intensity and the actual light intensity distributions that produces because of Fresnel diffraction for expression as what the output of line sensor was tried to achieve.

Figure 18 proofreaies and correct the treatment step figure of usefulness for expression to the error between distribution of light intensity and the actual light intensity distributions that produces because of Fresnel diffraction that is subjected to of trying to achieve as the output of line sensor.

Figure 19 is the concept map that the light income of two photo detector Cn, Cn-1 in the middle of expression is clipped in for the point with light intensity [0.25] carries out treatment for correcting.

Figure 20 is the synoptic diagram of the accuracy of detection contrast with the accuracy of detection after the correction of marginal position x0 and before proofreading and correct.

Figure 21 is the synoptic diagram of the accuracy of detection contrast with the accuracy of detection after the correction of distance z and before proofreading and correct.

Figure 22 is used to illustrate utilize the Fresnel function to be similar to the key diagram of the problem that exists to the light intensity distributions that produces because of Fresnel diffraction.

Label declaration

1 line sensor

2 light-projecting portions

3 rim detection portions (device body)

3b diffraction pattern pick-up unit

3c normalization device

The 3d edge detecting device

3e the 1st is subjected to light intensity to calculate device (light income on the point)

3f the 2nd is subjected to light intensity to calculate device (light income on the face)

3g treatment for correcting portion

7 shelters (detected object object)

Embodiment

Following with reference to accompanying drawing, the method for detecting position and the position detecting device of the present invention's one example are described.

Fig. 1 is the concise and to the point pie graph of the position detecting device of this example.This position detecting device comprises having in the folk prescription line sensor (light accepting part) 1 of a plurality of photo detector 1a of arranging of in accordance with regulations interval P upwards basically as shown in Figure 2, be provided with the sensitive surface subtend of this line sensor 1 and to the light-projecting portion 2 of a plurality of photo detector 1a projection monochromatic collimated beams 4 of this line sensor 1, and the output of resolving described line sensor 1 is positioned at the rim detection portion 3 of the marginal position of the shelter that is for example formed by shoestring (detected object object) 7 on the configuration direction of described photo detector 1a on the light path of described monochromatic collimated beam 4 with detection.

Also have, light-projecting portion 2 is made up of the optical fiber 2b that comprises the monochromatic light (laser) that the light source 2a that guides laser diode (LD) for example to constitute sends and the two parts that will be transformed into projection lens (collimation lens) 2c that throws behind the directional light by the monochromatic light of this optical fiber 2b guiding.This light-projecting portion 2 and described line sensor 1 are packed in the housing 5 of the コ font that forms specified gap together, separate the mutual subtend setting in described gap, are assembled into one, form as a sensory package.

The effect that described rim detection portion 3 plays is, according to the output of described line sensor 1, detects the position, end (edge) of the described shelter (detected object object) 7 in the gap by above-mentioned sensory package.Specifically, described rim detection portion 3 constitutes like this, it is conceived to be blocked thing (detected object object) 7 when blocking at the described monochromatic collimated beam of part, its end (edge part) produces Fresnel diffraction, and the light intensity of the sensitive surface of the described line sensor 1 of arrival is shown in Figure 7 as the aforementioned behind the generation Fresnel diffraction, have the position of keeping to the side sharply rise, along with away from marginal position while vibrating the convergent distribution character, thereby and detect the position of the end (edge) of aforementioned shelter 7 accurately according to light intensity distributions on the sensitive surface.

Basically by comprise above-mentioned line sensor 1, light-projecting portion 2, and the position detecting device that constitutes of rim detection portion 3 be to be provided with like this, it for example detect from reel uncoiling and shoestring from high-speed mobile to folk prescription that transmit to (shelter: 7 offsets on the Width detected object object), as the variation of its edge part position.And the position feedback of the edge part (edge) by will detected above-mentioned shoestring 7 is used to control it and moves driving to the mobile drive system of this shoestring 7.

In the position detecting device that is above-mentioned formation substantially, method for detecting position of the present invention and position detecting device is characterized in that, described rim detection portion 3 is when the output according to line sensor 1 detects the marginal position of shelter 7, be to utilize the hyperbolic sine function sech (x) of the light intensity distributions of approximate Fresnel diffraction generation to calculate marginal position, promptly be characterised in that the light intensity distributions of utilizing the approximate Fresnel diffraction of hyperbolic sine function sech (x) on the sensitive surface of described line sensor 1, to produce, particularly be similar to the intensity variation of its initial rising part (the 1st peak), according to utilizing the approximate light intensity distributions of this hyperbolic sine function sech (x), resolve the intensity variation of each photo detector 1a generation of aforementioned line sensor 1, thereby obtain the marginal position of described shelter 7.

Below, the light intensity distributions of utilizing approximate this Fresnel diffraction of hyperbolic sine function sech (x) to produce is described, as previously mentioned, with the Fresnel function time, exist the very large problem of error of the initial rising part (the 1st peak) of light intensity distributions.Therefore, only be conceived to the initial rising part (the 1st peak) of light intensity distributions, according to the shape at this peak (the variation tendency of light intensity) try to utilize 2 powers rational function, hyperbolic cosine function, and exponential function come to be similar to respectively.

Specifically can consider following three functions

Rational function as 2 powers

y＝a/{(x+b) ²+c}

Inverse as hyperbolic cosine function

y＝a/cosh(bx+c)

As exponential function

y＝a·exp{-b(x+c) ²}

To coefficient a, the b shown in these functions, the c suitable value of substitution respectively, find the solution its family curve simultaneously after, obtain result of calculation shown in Figure 3.

Additional disclosure once, characteristic A represents the theoretical value of light intensity distributions among Fig. 3, the variation of light intensity y when characteristic B represents that coefficient a, b, c are respectively for [0.057], [0.38], [0.0417] in the rational function of above-mentioned 2 powers, characteristic C represents the variation of coefficient a, b, the light intensity y when c is respectively [1.37], [6.29], [2.40] in the inverse of aforementioned hyperbolic cosine function, and the variation of the light intensity y of characteristic D when representing that coefficient a, b, c are respectively [1.37], [16.30], [0.38] in the aforementioned exponential function.But these calculate all be with monochromatic wavelength X as 670nm, from the edge of shelter 7 to the sensitive surface of line sensor 1 apart from small letter as carrying out under the prerequisite of 300mm.

Shown in these result of calculations, as can be known as with the inverse of hyperbolic cosine function, be hyperbolic sine function sech (x), but the very characteristic of the light intensity distributions that produces of the nearly Fresnel diffraction in highland, the characteristic of rising part (the 1st peak) especially at first of precision then.Find that particularly the benchmark that becomes regulation is subjected to the error with respect to theoretical value of the marginal position of light intensity [0.25] to be 6.77 μ m, and is very little, can precision be similar to the light intensity distributions that Fresnel diffraction produces highly.

Again the inverse of aforementioned hyperbolic cosine function is applied in passing in the formula of the light intensity distributions that aforesaid Fresnel diffraction produces, part till the initial rising part (the 1st peak) of approximate this light intensity distributions, its hyperbolic sine function sech (x) can represent by following formula, promptly

Light intensity=1.37sech{1.983 (2/ λ z) ^1/2X-2.386}

And can confirm that this approximate expression is consistent with the theoretical formula of light intensity distributions with about 3 precision.

The present invention is based on above-mentioned understanding, so light intensity distributions with the approximate Fresnel diffraction generation of above-mentioned hyperbolic sine function sech (x), particularly approximate initial rising part, hyperbolic sine function sech (x) with approximate this light intensity distributions, according to the output of aforesaid line sensor 1, detect the marginal position of shelter 7 accurately.

At this moment, in order to simplify its computing,, done following improvement to scheme to improve the detection processing speed of marginal position.Now the algorithm to this computing describes, and can represent by following formula as previously mentioned with the light intensity that hyperbolic sine function is approximate, promptly

Light intensity=1.37sech{1.983 (2/ λ z) ^1/2X-2.386}

In the formula, the position of establishing light intensity (relative value) and be [0.25] is (x-a), then can derive following relation from following formula

0.25＝1.37sech(x-a)

Above-mentioned position (x-a) can be calculated as following formula, promptly

(x-a)＝sech ^-1(0.25/1.37)＝1.866

Therefore, as the position that light intensity (relative value) is become [0.25] as the X-Y coordinate of initial point on the displacement above-mentioned light intensity distributions formula, represent light intensity y, then become

y＝1.37sech{1.98(2/λz) ^1/2x-2.39}

And calculate its inverse function, if substitution

Y＝y/1.37，X＝1.98(2/λz) ^1/2x

Then can with

X＝2.39-ln{[1+(1-Y ²) ^1/2]/Y}

Form is represented.

Also have, λ is the wavelength of monochromatic collimated beam, and its unit is [nm], and Z is the distance from inspection target object edge to sensitive surface, and its unit is [mm], and x is the marginal position on the sensitive surface in addition, and its unit is [μ m].

So, in rim detection portion 3, for example, be transformed into X-Y with above-mentioned coefficient [1.37] except that respectively be subjected to light intensity y1, the y2～ym that a plurality of in the line sensor 1 (m) photo detector 1a produces earlier according to step shown in Figure 4, the light intensity Y1 on the coordinate, Y2 ,～Ym (step S1).Then, in these a plurality of photo detector 1a, for example obtain adjacent mutually respectively and obtain than the big photo detector Cn that is subjected to light intensity of aforementioned reference light intensity [0.25] and obtain that (photo detector is determined device than the little photo detector Cn-1 that is subjected to light intensity of said reference light intensity [0.25]; Step S2).Promptly a plurality of photo detector 1a (C1, C2 ,～each element between Cm) in, obtain and be subjected to light intensity to become [0.25] two adjacent photo detector Cn, Cn-1 mutually.

Then, according to aforesaid approximate expression, utilize inverse transformation the acquisition of above-mentioned each photo detector Cn, Cn-1 to be subjected to position Xn, Xn-1 on the sensitive surface of this photo detector Cn, Cn-1 of light intensity Yn, Yn-1 calculate (light receiving position calculation element respectively; Step S3), calculating formula is

Xn＝2.39-ln{[1+(1-Yn ²) ^1/2]/Yn}

Xn-1＝2.39-ln{[1+(1-Yn-1 ²) ^1/2]/Yn-1}

According to above-mentioned position Xn, Xn-1, notion as shown in Figure 5 is such, utilizes interpolation operation to calculate and is subjected to light intensity to become the position (marginal position) (interpolation operation device, step S4) of [0.25].Though can carry out with aforesaid approximate expression about this interpolation operation,, can use simple linear interpolation the intensity variation between above-mentioned two photo detector Cn, Cn-1 being used as under the linear situation.

Also have, here be to find that between adjacent photo detector 1a light intensity becomes the position of [0.25], determine two photo detector Cn, the Cn-1s of this position, but also can only determine above-mentioned position is clipped in central plural photo detector as the element border.But in this case, must carry out interpolation operation, as long as by preventing that like this its operational precision from reducing with aforesaid approximate expression.In addition, about above-mentioned inverse transformation,, can alleviate its operation processing burden greatly, and can in near-instantaneous, carry out by using the reckoner of for example storing its calculated value in advance.

Like this, method for detecting position and position detecting device according to the marginal position of aforesaid detection shade 7, owing to utilize the hyperbolic sine function sech (x) of the light intensity distributions of approximate accurately Fresnel diffraction generation, be subjected to light intensity y according to what a plurality of photo detector 1a of line sensor 1 produced, calculating its light intensity becomes the position X of [0.25], so can improve its accuracy of detection greatly.In addition, in common microprocessor with floating-point operation (FPU) function, natural logarithm function (ln function) is included among its instruction, even but do not possess the microprocessor of this FPU function, for above-mentioned hyperbolic sine function sech (x), its inverse function ln (x) particularly for example can be as following formula

ln(x)＝-1.0537+1.4285x-0.0382x ²

+ 0.0607x ³-0.0040x ⁴+ ... carry out series expansion, because its convergence is also fast, so calculate easily.Thereby play can be simply and carry out the effect that marginal position detects accurately.

In addition, the output of line sensor 1 changes according to arrangement pitch p and the number of elements of each photo detector 1a of this line sensor 1.When using arrangement pitch p by 7 μ m to have the high-resolution image sensor of 5000 elements, for example can obtain the sensor output of the very dense shown in Fig. 6 (a).This point when using arrangement pitch p by 85 μ m to have the cheap line sensor commonly used of 102 elements, can only obtain the comparatively sparse like that sensor output shown in Fig. 6 (b).But just because number of elements is few, thereby sensor output can be read at a high speed.

Even but utilize the low cheap line sensor 1 of this resolution, according to aforesaid method for detecting position of the present invention and device, owing to the hyperbolic sine function sech (x) of the light intensity distributions that produces with the approximate Fresnel diffraction of high precision, so the also variation that is subjected to light intensity between interpolation photo detector 1a accurately.Therefore, can adopt the cheap line sensor 1 of low resolution to export with very fast speed pickup, and utilize simple calculations to handle and can detect marginal position etc. accurately, this obtains sizable effect in practical application.

In addition, not the occasion of complete occulter at detected object object 7, for example the occasion of translucent body formation just can not cover monochromatic collimated beam fully.At this moment, the output of line sensor 1 just becomes the component of the light that sees through detected object object 7 of also having superposeed, and as shown in Figure 7, it is subjected to light intensity all to surpass [0.25] in the whole light area that spreads all over line sensor 1 sometimes.So,, just can not detect marginal position according to aforesaid algorithm.

So, in this occasion, for example the detected object object 7 that constitutes with translucent body earlier covers whole light areas of line sensor 1, the monochromatic collimated beam when obtaining being subjected to the light figure and not having detected object object 7 of detected monochromatic collimated beam at this moment be subjected to the poor of light figure.Then, differ biasing and the gain of adjustment according to this with respect to the output of line sensor 1.

Each photo detector 1a of the line sensor 1 when specifically, obtaining no detected object object 7 respectively be subjected to light intensity Ai (i=1,2 ,～m), and when covering whole light area of line sensor 1 with detected object object 7 this line sensor 1 be subjected to light intensity Ci (i=1,2 ,～m).Then, with above-mentioned be subjected to light intensity Ci (i=1,2 ,～m) minimum Cmin is provided as the biasing with respect to the output of line sensor 1, on this basis, adjust its output gain, make be subjected to light intensity Ai (i=1,2 ,～m) and the mean value of the difference of above-mentioned minimum Cmin become half of maximum output of this line sensor 1.Thereafter, obtain the output of aforementioned monochromatic collimated beam once again, obtain with this output as coefficient (normalized parameter) Ni of [1] (i=1,2 ,～m).Here, above-mentioned biasing and gain are adjusted owing to detected object object 7 causes for translucent body, so when the light and shade resolution of the output of line sensor 1 is little, be that purpose is carried out, under the sufficiently high situation of resolution, just there is not this necessity to remedy its deficiency.

Then, in the marginal position of reality detects, ask its be subjected to light figure Yi (i=1,2 ,～m), and according to above-mentioned coefficient Ni (i=1,2 ,～m), with the output normalization of line sensor 1.Obtain its peak value and for example output valve of an one hithermost photo detector 1a respectively from the initial rising part that is subjected to the light figure again, determine to obtain above-mentioned two photo detector 1a that respectively are subjected to light value respectively.Then according to aforesaid approximate expression (contrary Fresnel function)

Xp＝2.39-ln{[1+(1-Yp ²) ^1/2]/Yp}

Xp-1＝2.39-ln{[1+(1-Yp-1 ²) ^1/2]/Yp-1}

To be subjected to light intensity Yp, Yp-1 inverse mapping as shown in Figure 7 on X-axis.Then can be according to light receiving position Xp, the Xp-1 of inverse mapping, calculate as shown in Figure 7 and be subjected to light intensity to become the marginal position of [0.25].

Through above-mentioned processing, though on the photo detector 1a of line sensor 1 be subjected to light intensity to surpass [0.25] time, when promptly detected object object 7 is translucent body, still can detect its marginal position accurately.Just shown in this example, even uncertain benchmark is subjected to the element that be subjected to light intensity of light intensity [0.25] in the middle of being clipped in, also can be according to the photo detector 1a that for example gets its peak value and its photo detector 1a that is subjected to light intensity in front, calculate the marginal position of detected object object 7, can obtain the effect same with aforesaid example.

Also have, quantity and the arrangement pitch p thereof of the photo detector 1a that has about image sensor 1, as use with the corresponding sensor of its detection specification and just can meet the demands.In addition, about rim detection portion 3, can utilize general purpose microprocessor to realize, also aforesaid arithmetic expression can be solidificated among the ROM provides.

The method for detecting position and the position detecting device of other example of the present invention below are described.

Fig. 9 is the concise and to the point pie graph of the position detecting device of this example of expression.The light-projecting portion 2 of this example for example shown in the concise and to the point formation of Figure 10, comprise the catoptron that monochromatic light (laser) that the light source 2a that will be made up of laser diode (LD) sends reflects (for example by aluminium-vapour deposition to implement the prism of mirror process) 2d, will be constrained to by the homogeneous beam shape of this catoptron 2d guiding slit-shaped slit anti-dazzling screen (light projector) 2e, and will be transformed into projection lens (collimating mirror) 2c that throws behind the parallel beam by the light of this slit anti-dazzling screen 2e again.The detected object object is that shelter 7 is between this projection lens 2c and aforementioned light accepting part 1, by the marginal position of aforementioned light accepting part (line sensor) 1 detection at the above-mentioned shelter 7 of the length direction top offset of the slit of slit anti-dazzling screen 2e.

Specifically, slit anti-dazzling screen 2e is a slit of its opening shape being made rectangle, so aforementioned light source 2a is provided with like this, makes it towards the extended corner ejaculation monochromatic light of above-mentioned slit with regulation.Particularly using under the situation of LD as light source 2a, penetrating intensity from this LD and be the laser that ellipse garden shape distributes, throwing as shown in phantom in FIG. facing to slit anti-dazzling screen 2e.At this moment, optically this LD and slit anti-dazzling screen 2e are disposed like this, make the major axis of its above-mentioned laser be in the length direction of the slit of aforementioned

slit anti-dazzling screen

2e, 2 miniaturizations are good for light-projecting portion for this.Also have, catoptron (prism) 2d forms the light path that the laser that LD is sent is approximated to the right angle reflection, thus both kept LD and slit anti-dazzling screen 2e and and the optical range of projection lens 2c, also play the effect of the whole compact shape that makes light-projecting portion 2.Also have, this light-projecting portion 2 is for example packed in the housing 5 of the コ font that forms specified gap together with aforementioned line sensor 1, separate the mutual subtend in above-mentioned gap, and the composition one can realize as a sensory package.

Utilize the light-projecting portion 2 of this formation, its optical system is respectively as signal Figure 11 and shown in Figure 12, and parallel beam (monochromatic collimated beam) 4 that is transformed into the cross sectional shape with slit-shaped of directional light by above-mentioned slit anti-dazzling screen 2e and projection lens 2c throws to line sensor (light accepting part) 1.The size of the cross sectional shape of this parallel beam is for example for long limit 9mm * minor face 3mm, and is corresponding, and the sensitive surface size of accepting the line sensor 1 of above-mentioned parallel beam for example is long limit 8.7mm * minor face 0.08mm.Promptly long limit size separately about equally.

Additional disclosure once, why the minor face size (3mm) of parallel beam cross sectional shape is set for very large more than the minor face size (0.08mm) of the sensitive surface of line sensor 1, be because for the ease of adjusting the depth of parallelism of light projector and light-receiving device, when even light projector or light-receiving device tilt simultaneously, as shown in figure 12, also can avoid the influence of the Fresnel diffraction that the long side edge 2h of the slit of slit anti-dazzling screen 2e produces.But, parallel beam (monochromatic collimated beam) 4 for this slit-shaped, do not deny when utilizing aforesaid slit shadow shield 2e that beam shape is carried out shaping, as shown in figure 11, comprise owing to slit anti-dazzling screen 2e influences the uneven light component that produces at the Fresnel diffraction of the short brink 2f of slit.But influence for this not parallel light component, as long as when for example in light path, not having shelter 7, obtain the light quantity variation that produces because of above-mentioned Fresnel diffraction for each photo detector 1a respectively according to the output of line sensor 1, again according to its light quantity changing unit, the output normalization of line sensor 1 (photo detector 1a) is proofreaied and correct get final product respectively.

Here, constitute the input buffer 3a that the device body of aforementioned rim detection portion 3 has the light intensity distributions on output (light income of each photo detector 1a) that is taken into aforementioned line sensor 1 and the sensitive surface of the obtaining this line sensor 1.Particularly device body (rim detection portion) 3 also has diffraction pattern pick-up unit 3b, this diffraction pattern pick-up unit 3b accepts from whole monochromatic collimated beams of the regulation width of light beam of aforementioned light-projecting portion 2 projections with aforementioned line sensor 1 in advance, handle as its initial setting, according to light intensity distributions at this moment, obtain the diffraction pattern of the monochromatic collimated beam of aforementioned light-projecting portion 2 projections, simultaneously as described later, obtain normalized parameter according to the inverse of this diffraction pattern for the light income of aforementioned each photo detector 1a.This diffraction pattern is that the short brink edge 2f of the slit that forms owing to above-mentioned slit anti-dazzling screen 2e goes up the not parallel light component that the influence of Fresnel diffraction produces and causes.

Have again, device body 3 also have according to by the detected normalized parameter of above-mentioned diffraction pattern pick-up unit 3b with the normalized normalization device of the output of aforementioned line sensor 1 3c, and edge detecting device 3d, this edge detecting device 3d is according to the output (normalization output) of carrying out the aforementioned line sensor 1 that normalized crosses with this normalization device 3c, detecting the end position of aforementioned shelter (detected object object) 7, is the distance z that detects between the sensitive surface of the position x of orientation of photo detector 1a of line sensor 1 and above-mentioned edge and line sensor 1 specifically.

This edge detecting device 3d constitutes like this, when it is conceived to the aforementioned monochromatic collimated beam of shelter (detected object object) 7 shield portions basically, go up the generation Fresnel diffraction at its end (edge), and the light intensity of the sensitive surface of the aforementioned line sensor 1 of arrival is shown in Figure 8 as the aforementioned behind the generation Fresnel diffraction, have near marginal position and sharply rise, along with convergent distribution character then while vibrating away from marginal position, thereby, detect the position, end (edge) (position x and distance z) of aforementioned shelter 7 accurately according to the light intensity distributions on the sensitive surface of line sensor 1.Particularly as previously mentioned, the light intensity distributions of utilizing the approximate Fresnel diffraction of hyperbolic sine function sech (x) to produce, thereby the marginal position of the above-mentioned shelter 7 of high Precision Detection (position x0 and distance z).

Additional disclosure once, when supposing with shelter 7 shield portions monochromatic collimated beams, light intensity distributions on the sensitive surface of aforementioned line sensor 1 begins to rise from light intensity [0], be converged in the position of [1.0], then as previously mentioned, obtain the marginal position x0 of element arrangements direction in the position that its initial rising part (the 1st peak) becomes [0.25] as light intensity.

In addition, about the distance z between the sensitive surface of edge on the optical path direction and line sensor 1, because be subjected to the light intensity distributions on the sensitive surface of the line sensor 1 that Fresnel diffraction influences, especially the light intensity distributions of its rising part depends on the wavelength X and the above-mentioned distance z of aforementioned monochromatic collimated beam, so can be according to a plurality of locational light intensities that are subjected to of sensitive surface of the aforementioned line sensor 1 of this rising part, for example become the light intensity that is subjected on two photo detector 1a of front and back of [0.25] according to light intensity, and position of components, calculate according to the characteristic of aforementioned lights intensity distributions.

Except such basic function, aforementioned body apparatus 3 comprises that also the 1st is subjected to light intensity to calculate device 3e, be used for when the output with aforementioned line sensor 1 is normalized into [1], be subjected to the big photo detector Cn that is subjected to light intensity of light strength ratio [0.25] and obtain the above-mentioned little photo detector Cn-1 that is subjected to light intensity of light strength ratio [0.25] that is subjected to for obtaining, above-mentioned each photo detector Cn, Cn-1 regarded by light to utilize when some as aforementioned hyperbolic sine function sech (x) obtains this photo detector Cn, Cn-1 respectively is subjected to light intensity Yn, Yn-1; And the 2nd be subjected to light intensity to calculate device 3f, it is used for utilizing carries out the function of integration with aforementioned hyperbolic sine function sech (x) to each arrangement pitch of each photo detector, obtains respectively to be subjected to light intensity yn, yn-1 on the whole sensitive surface of aforementioned each photo detector Cn, Cn-1.

And, also comprise the 3g of treatment for correcting portion, this one by the light income calibration function and repeatedly the s operation control function form, the former be used to obtain light income yn, yn-1 on light income Yn, Yn-1 on the point of above-mentioned each photo detector Cn, Cn-1 and each photo detector Cn, Cn-1 whole difference Δ yn (=Yn-yn), Δ yn-1 (=Yn-1-yn-1) as correcting value, that proofreaies and correct aforementioned each photo detector Cn, Cn-1 then respectively is subjected to light intensity A (xn), A (xn-1); The latter is used for position xn, the xn-1 of the element arrangements direction that is subjected to light intensity A (xn) ', A (xn-1) ' and above-mentioned photo detector Cn, Cn-1 of aforementioned each photo detector Cn, Cn-1 after proofreading and correct are offered the aforementioned rim detection 3d of portion, carries out aforesaid marginal position once again and calculates.To narrate afterwards about the 1st and the 2nd above-mentioned detailed description that is subjected to light intensity to calculate device 3e, 3f and the 3g of treatment for correcting portion.

In the position detecting device that constitutes like this, the feature part of method for detecting position of the present invention and device is following this point, promptly according to utilizing the approximate light intensity distributions of aforesaid two-sided secant sech (x), resolve the light intensity that is subjected to that each photo detector 1a of aforementioned line sensor 1 produces, obtain the marginal position x0 and the distance z of aforementioned shelter 7.

Specifically, in edge detecting device 3d, for example according to step shown in Figure 13, at first according to a plurality of (m) the photo detector 1a on the line sensor 1 try to achieve normalized respectively be subjected to light intensity y1, y2 ,～ym, obtain adjacent mutually acquisition respectively than the big photo detector Cn that is subjected to light intensity of aforesaid reference light intensity [0.25] with obtain the photo detector Cn-1 (step S11) that be subjected to light intensity littler than said reference light intensity [0.25].Promptly a plurality of photo detector 1a (C1, C2 ,～obtain in each element between Cm) and be subjected to light intensity to become [0.25], adjacent each other two photo detector Cn, Cn-1.That uses then that above-mentioned coefficient [1.37] removes above-mentioned each photo detector Cn, Cn-1 is subjected to light intensity yn, yn-1, is transformed into light intensity Yn, Yn-1 (step S12) on the X-Y coordinate again.

Thereafter, according to aforesaid approximate expression, with above-mentioned each photo detector Cn, Cn-1 obtain being subjected to position Xn, Xn-1 on the sensitive surface of this photo detector Cn, Cn-1 of light intensity Yn, Yn-1, with following formula

Xn＝2.39-ln{[1+(1-Yn ²) ^1/2]/Yn}

Xn-1＝2.39-ln{[1+(1-Yn-1 ²) ^1/2]/Yn-1}

Utilize the relative position (step S13) on the inverse transformation calculating X-axis respectively, again according to above-mentioned position Xn, Xn-1, its notion is utilized interpolation operation to calculate the position of photo detector Cn and is subjected to light intensity to become the difference Δ X of the marginal position of [0.25] as shown in figure 14, promptly

ΔX＝W·[Xn/(Xn-Xn-1)]

(step S14).Above-mentioned in addition poor Δ X is because be from being subjected to light intensity to become the distance of the marginal position x0 of [0.25] to the position of photo detector Cn, so at the absolute position x when the 1st photo detector C1 measures on the whole sensitive surface of line sensor 1, establishing n is the arrangement pitch of the element number that obtains the photo detector 1a of light quantity Y2, photo detector 1a when being W, becomes

x＝n·w-Δx

In addition, the relative position Xn, the Xn-1 that try to achieve in above-mentioned inverse transformation are shown below

X＝1.98(2/λz) ^1/2x

Be [1.98 (2/ λ z) ^1/2] doubly value, but by getting their ratio with above-mentioned interpolation operation, thereby in fact this can be deleted.

Also have, here have and find between the adjacent photo detector 1a that light intensity becomes the position of [0.25], determine two photo detector Cn, the Cn-1s of this position, but also can determine it singly is that above-mentioned position is clipped in central plural photo detector as the element border.But in this case, must carry out interpolation operation, can prevent that its operational precision from reducing with aforesaid approximate expression.In addition, for above-mentioned inverse transformation, for example, can greatly alleviate operation processing burden, and in instantaneous, finish by using the reckoner of storing its calculated value in advance.

On the other hand, as shown in figure 13, according to relative position Xn, Xn-1 on the sensitive surface of aforementioned photo detector Cn, Cn-1, be subjected to light intensity become position (marginal position) x0 of [0.25] and photo detector Cn the position difference Δ x or be subjected to light intensity, and the wavelength X of aforementioned monochromatic collimated beam at photo detector Cn place, the distance of the edge by calculating shelter 7 according to aforementioned hyperbolic sine function sech (x) and the sensitive surface of line sensor 1, thus can obtain distance z (step S15).Concrete distance calculation is the aforementioned formula according to the Fresnel diffraction at approximate aforesaid the 1st peak basically

Light intensity A (x)=1.37sech{1.98 (2/ λ z) ^1/2X-2.39}

Find the solution distance z,

z＝(2/λ){1.98·x/[arcsech(A(x)/1.37)+2.39]} ²

Distance z between the edge by calculating shelter 7 and the sensitive surface of line sensor 1 is carried out.

At this moment, when the marginal position of the orientation of asking aforesaid photo detector, utilize the position of the photo detector Cn that obtains the big intensity of light strength ratio [0.25], the difference Δ x according to this position and marginal position calculates with following formula,

z＝(2/λ){1.98·Δx/[arcsech(yn/1.37)+2.39]} ²

Then can obtain simply between the sensitive surface of the edge of shelter 7 and line sensor 1 apart from Z.Particularly the denominator of following formula is one,

Xn＝2.39-ln{[1+(1-Yn ²) ^1/2]/Yn}

Since suitable with following formula, promptly

z＝(2/λ){1.98·Δx/Xn} ²

Then calculating can be simpler.

Concrete as Figure 15 (a) shown in, (y2＞y1), n is that the element number, the W that obtain the photo detector 1a of light quantity y2 are the interval between photo detector, and to establish optical wavelength be λ, at this moment in order the light intensity after the normalization [0.25] is clipped in 2 central light quantity to establish y1, y2

(1)Y1＝y1/1.37

(2)Y2＝y2/1.37

(3)x1＝2.39-ln{[1+(1-Y1 ²) ^1/2]/Y1}

(4)x2＝2.39-ln{[1+(1-Y2 ²) ^1/2]/Y2}

(5)Δx2＝W[x2/(x2-x1)]

(6)x0＝W·n-Δx2

(7)z＝(2/λ)(1.98·Δx2/x2) ²

Can obtain the marginal position of x direction (orientation of photo detector 1a) and z direction (optical path direction) with above-mentioned formula simultaneously.

In addition, when becoming 2 the computed range z in front and back of position of [0.25] according to aforementioned lights intensity, increase when having problems in the lower and error of its resolution, also can be shown in Figure 15 (b), obtain the position xa of the crest pre-set A of light intensity arbitrarily before, that for example be assumed to be [0.8] or [1.0] that arrives the 1st peak, obtaining this position xa and aforementioned lights intensity becomes the difference Δ x of the position x0 of [0.25], according to this difference DELTA x, calculates afore-mentioned distance z.

For example, asking light intensity to become under the situation of position xa of [1.0], because become following formula

1.0＝1.37sech(X’-α)

X’-α＝arcsech(1.0/1.37)＝0.83

So as the position x that light intensity y is become [1.0] then will as initial point

y＝1.37sech(x’-0.83)

As approximate expression, can obtain light intensity y.So, in the formula of inverse transformation,

Y＝y/1.37，

Then become

X＝0.83-ln{[1+(1-Y ²) ^1/2]/Y}

So can be with the following various aforesaid calculating of carrying out

(1)Y1＝y1/1.37

(2)Y2＝y2/1.37

(3)x1＝0.83-ln{[1+(1-Y1 ²) ^1/2]/Y1}

(4)x2＝0.83-ln{[1+(1-Y2 ²) ^1/2]/Y2}

(5)Δx2＝W[x2/(x2-x1)]

(6)xa＝W·n-Δx2

(7)z＝(2/λ){1.98·(xa-x0)/[arcsech(Y2)+2.39]} ²

But being light intensity, above-mentioned x0 becomes the marginal position of [0.25].

Also have, as obtain the crest location xp at the 1st peak that becomes the Fresnel diffraction of [1.37] as default any light intensity, then because the disappearance of the arcsech (Y2) in the following formula, so distance z can be calculated simply by following formula.

z＝(2/λ)[1.98·(xp-x0)/2.39] ²

Yet as previously mentioned, the photo detector 1a of line sensor 1 has the sensitive surface of prescribed level respectively, accepts the suitable signal of total amount of light on output and this sensitive surface.Therefore, become by light intensity A will principal value of integral to be carried out in each position of each photo detector 1a in its element width scope with the approximate variation that is subjected to light intensity of aforesaid hyperbolic sine function sech (x) from what each photo detector 1a tried to achieve, this is subjected to distribution of light intensity for example just to present the variation of bar chart as shown in figure 16.Because, the error of obtaining as the output of line sensor 1 that is subjected to the light intensity distributions that light intensity distributions and actual Fresnel diffraction produce, then as Figure 17 (a) shown in, it is that the crest location of [1.37] becomes maximum at minus side that light intensity is gone up in its normalization output.In addition, on the position that is about to two photo detectors of light intensity [0.25] in the middle of being clipped on the marginal position, produce positive error, light intensity for [0.75] near the generation error become the point of [0].The elongated situation of tendency that this error produces and the distance z between edge and the line sensor 1 too, shown in Figure 17 (b), its error profile just enlarges on the element arrangements direction.

No matter these what states, aforesaid processing just will have the light intensity that is subjected on certain photo detector 1a that is subjected to optical width (sensitive surface) to handle as the representational light income on this photo detector 1a, and the light income that is subjected to each point on the optical width direction of sensitive surface of not considering this photo detector 1a is along with the variation of Fresnel diffraction.And, though this does not have the processing of attention insignificant, become the main cause that makes its measuring accuracy variation.

So, in method for detecting position of the present invention and device, tackling error photo detector 1a, that cause owing to the sensitive surface that prescribed level is arranged of above-mentioned line sensor 1 proofreaies and correct, to improve the accuracy of detection of aforesaid marginal position x0 and distance z, as described below, the light intensity A (x) that the photo detector 1a according to aforementioned line sensor 1 is tried to achieve proofreaies and correct.

That is, because the edge produces the rising part of the light intensity distributions of Fresnel diffraction, as representing, then with the approximate expression of above-mentioned hyperbolic sine function sech (x)

A(xn)＝1.37·sech[1.98(2/λz) ^1/2-2.39]

With last different be, the rising part of the light intensity distributions of obtaining from each photo detector 1a of line sensor 1, be to carry out principal value of integral as the light that the output of each photo detector 1a is produced Fresnel diffraction in the width range of its sensitive surface to provide, this integrated value is because the indefinite integral of hyperbolic sine function sech (x) is [2arctan (ex)], therefore can represent with following formula, promptly

A(xn)＝{1.37×2/[1.98(2/λz) ^1/2]}

×{arctan(α)-arctan(β)}

α＝exp{1.98(2/λz) ^1/2xne-2.39}

β＝exp{1.98(2/λz) ^1/2xns-2.39}

In the formula, xn represents the position at the sensitive surface center of n photo detector 1a, and xns represents that front position, the xne of the sensitive surface of above-mentioned photo detector 1a represent the back-end location of the sensitive surface of above-mentioned photo detector 1a.And, the light income A (xn) at the xn place, position of n photo detector 1a of line sensor 1, as know that the distance z between the sensitive surface of the edge of its position x and shelter 7 and line sensor 1 just can calculate simply.

So, when the error that the sensitive surface of proofreading and correct because of photo detector 1a causes, treatment step as shown in figure 18 for example, be normalized into the light intensity distributions of [1] earlier according to output with aforementioned line sensor 1, promptly utilize with a plurality of photo detector 1a (C1, C2,～Cm) each self-monitoring light income A (xn) that goes up measurement (is subjected to light intensity y1, y2,～ym), obtain the acquisition adjacent each other photo detector Cn that is subjected to light intensity bigger as previously mentioned respectively than aforesaid reference light intensity [0.25], the photo detector Cn-1 (step S21) that be subjected to light intensity littler than said reference light intensity [0.25] with acquisition.Then, obtain as previously mentioned above-mentioned each photo detector Cn, Cn-1 center xn, xn-1 and and marginal position x0 between distance, delta xn, Δ xn-1 (step S22), again as described above, obtain distance z (step S23) between the sensitive surface of edge and line sensor 1.

Then, the sensitive surface of aforementioned photo detector Cn, Cn-1 being regarded as when a bit, establish

p＝(2/λz) ^1/2

According to aforementioned hyperbolic sine function sech (x), with following formula

Yn＝1.37·sech[1.98pΔxn-2.39]

Yn-1＝1.37·sech[1.98pΔxn-1-2.39]

Obtain light income Yn, the Yn-1 (step S24) of the center that the distance of only leaving marginal position x0 respectively is photo detector Cn, the Cn-1 of Δ xn, xn-1 (point) respectively.

In addition, use following formula

yn＝{2.74/1.98pW}·{arctan(α1)-arctan(β1)}

α1＝exp{1.98p(Δxn+W/2)-2.39}

β1＝exp{1.98p(Δxn-W/2)-2.39}

yn-1＝{2.74/1.98pW}·{arctan(α2)-arctan(β2)}

α2＝exp{1.98p(-Δxn-1+W/2)-2.39}

β2＝exp{1.98p(-Δxn-1-W/2)-2.39}

Obtain light quantity yn, yn-1 (step S25) when measuring on the face of aforementioned each photo detector Cn, Cn-1 respectively.Also have, the division arithmetic of above-mentioned utilization [1.98pW] is to carry out following processing, promptly uses the light income of trying to achieve as area according to Integral Processing and the optical width that is subjected to of photo detector 1a to carry out division, obtains the average light income of each point of photo detector 1a thus.

Then, as shown in figure 19, use

Δyn＝Yn-yn、Δyn-1＝Yn-1-yn-1

Difference Δ yn, the Δ yn-1 (step S26) of light quantity yn, yn-1 when obtaining above-mentioned upper integral respectively and light income Yn, the Yn-1 that should on the center of photo detector Cn, Cn-1, try to achieve during with point measurement, with above-mentioned difference Δ yn, Δ yn-1 as correcting value, respectively with following formula

A(xn)’＝A(xn)+Δyn

A(xn-1)’＝A(xn-1)+Δyn-1

Proofread and correct light income A (xn), the A (xn-1) (step S27) of aforementioned each photo detector Cn, Cn-1.

Then, utilize light income A (xn) ', the A (xn-1) ' of each photo detector Cn, Cn-1 after the above-mentioned correction, carry out the processing that aforesaid step S11 begins repeatedly, respectively edge calculation position x0 and apart from Z.Its result, can be according to the light income of the face upper integral that is subjected to center xn, xn-1 that light intensity A (xn), A (xn-1) obtain respectively of photo detector Cn, Cn-1 shown in Figure 19, approach to regard as on above-mentioned position xn, the xn-1 in the light intensity distributions that actual Fresnel diffraction produces the light intensity when some on the sensitive surface of line sensor 1.Thereby, can improve according to these and regard light income Yn when some, marginal position x0 that Yn-1 obtains respectively and the accuracy of detection of distance z as.Particularly there is interdependent relation, so, then can improve its accuracy of detection more as carrying out above-mentioned treatment for correcting repeatedly till marginal position x0 or distance z convergence owing to marginal position x0 and apart from Z and aforementioned corrected amount Δ yn, yn-1.

Specifically, by carrying out once above-mentioned treatment for correcting, shown in Figure 20 (a), can reduce the detection error of marginal position x0 widely.In addition, as twice above-mentioned treatment for correcting repeatedly, be subjected to light income Yn, Yn-1 on center xn, the xn-1 that light intensity A (xn), A (xn-1) obtain respectively, that regard as when some as previously mentioned according to photo detector Cn, Cn-1 are restrained more, its error is reduced shown in Figure 20 (b) more.And then as with treatment for correcting three times repeatedly, then its error in fact also might become [0].In addition,, shown in Figure 21 (a), also can reduce its measuring error,, shown in Figure 21 (b), also can make its error be approximately zero [0] by twice treatment for correcting repeatedly simultaneously for distance z.

Also have,, then also can when its initial adjustment, only carry out the calculating of a distance z in advance, use this distance z later on, carry out the detection of position x0 as described later and handle if the sensitive surface distance z of the edge of shelter 7 and line sensor 1 is certain.At this moment, can carry out common interpolation operation as shown in figure 13, for the processing of obtaining relative position Xn, Xn-1, as long as carrying out inverse operation, the integration type that described function is carried out integration just can, but do not knowing under the situation of distance z, because the distance z of obtaining as previously mentioned also contains error, so the z that preferably simultaneously adjusts the distance as previously mentioned finds the solution repeatedly, one side is carried out its treatment for correcting.

In addition, under the situation of known distance z, in order to replace utilizing aforesaid following formula

X＝2.39-ln{[1+(1-Y ²] ^1/2/Y)

Ask center Xn, Xn-1 and the light intensity of light-emitting component Cn, Cn-1 to become difference Δ x, the Δ xn-1 of the marginal position x0 of [0.25] respectively, if utilize the aforesaid quantometer formula that carries out integration for face

A(xn)＝{1.37×2/[1.98(2/λz) ^1/2W]}

×{arctan(α)-arctan(β)}

α＝exp{1.98(2/λz) ^1/2xne-2.39}

β＝exp{1.98(2/λz) ^1/2xns-2.39}

Inverse function, its light intensity is mapped on the X-axis, carry out interpolation again and handle, just can not need above-mentioned repeated calculation, and eliminate its error.

At this moment, when calculating above-mentioned inverse function, utilize numerical computation method such as Newton method be comparison its find the solution better method with resolving.Be about to the above-mentioned quantometer formula that carries out integration for face and represent with following formula,

\begin{matrix} a = 1.37 \times 2 / 1.98 {(2 / λz)}^{1 / 2} w \\ b = 1.98 {(2 / λz)}^{1 / 2} \\ c = 1.98 {(2 / λz)}^{1 / 2} w / 2 - 2.39 \\ d = - 1.98 {(2 / λz)}^{1 / 2} w / 2 - 2.39 \end{matrix}\}

In the aforementioned calculation formula, coefficient a, b, c, d respectively are

Y＝a{arctan[exp(bx+c)]-arctan[exp(bx+d)]}

Its differential expression is

Y’＝ab{[exp(bx+c)]/[1+[exp2(bx+c)]}

-ab{[exp(bx+d)]/[1+[exp2(bx+d)]}

Thereby, if utilize the numerical evaluation of Newton method repeatedly, in the error range that is allowing with the error shown in [Y/Y '],, and obtain position X then as long as error is restrained fast.In addition, for inverse transformation,, can alleviate its operation processing burden greatly, the instantaneous inverse transformation result that obtains by the prestore reckoner of its calculated value of utilization.

More particularly, be to obtain respectively when output with line sensor 1 is normalized into [1] to obtain to be subjected to big element Cn that is subjected to light intensity of light strength ratio [0.25] and acquisition to be subjected to the little photo detector Cn-1 that is subjected to light intensity of light strength ratio [0.25].Then, according to will be according to the auxiliary function of the light intensity shown in the aforesaid hyperbolic sine function sech (x), promptly to each arrangement pitch integration of each photo detector

A(xn)＝{1.37×2/[1.98(2/λz) ^1/2W]}

×{arctan(α)-arctan(β)}

α＝exp{1.98(2/λz) ^1/2xne-2.39}

β＝exp{1.98(2/λz) ^1/2xns-2.39}

And be subjected to light intensity Yn, Yn-1 on photo detector Cn, the Cn-1 utilize following formula to represent, promptly

Yn＝a{arctan[exp(b·xn+c)]-arctan[exp(b·xn+d)]}

Yn-1＝a{arctan[exp(b·xn-1+c)]-arctan[exp(b·xn-1+d)]}

Then obtain center xn, the xn-1 (position of components pick-up unit) of aforementioned photo detector Cn, Cn-1 respectively.Then, as long as according to the interval W following formula between the photo detector of above-mentioned position xn, xn-1 and aforementioned line sensor

Δxn＝W[xn/(xn-xn-1)]

x0＝W·n-Δxn

Obtaining the aforementioned marginal position x0 that is subjected to light intensity to become [0.25] (marginal position pick-up unit) gets final product.

Like this, according to the present invention, owing to be conceived to each photo detector 1a of line sensor 1 because the light of the light intensity distributions that Fresnel diffraction produces is subjected to optical width direction (Fresnel diffraction direction) to carry out detecting under the state of integration in its whole sensitive surface upper edge, after the light intensity A (x) that obtains from each photo detector 1a is proofreaied and correct, according to the light intensity distributions that produces Fresnel diffraction, obtain marginal position x0 and edge distance z, so can improve its accuracy of detection greatly to sensitive surface.Especially combine with the light intensity distributions of utilizing aforesaid hyperbolic sine function sech (x) to be similar to Fresnel diffraction accurately, even the arrangement pitch at the photo detector 1a of line sensor 1 more slightly is under the situation of 85 μ m, also can detect marginal position x0 and distance z etc. respectively with the precision below the 0.05 μ m, advantage of the present invention is a lot of in practicality.

Also have, in this example, for example quantity, its arrangement pitch W of the photo detector 1a that has about line sensor 1 just can as long as use with its detection specification corresponding sensor.In addition, among the present invention, be to be conceived to be subjected to light intensity [0.25] to be clipped in central adjacent two photo detector Cn, Cn-1 and to carry out treatment for correcting, carry out treatment for correcting but can be conceived to that also [0.25] is clipped in central any two photo detector Cn+m, Cn-k etc.

Also having, in above-mentioned example, is to utilize the function of hyperbolic sine function sech (x) as the intensity distributions of approximate Fresnel diffraction, but also available other function.At this moment, as long as for example calculate this function in advance with mainframe computer etc., the list data after again the data of this function being solidified as ROM etc. provides and gets final product.At this moment, also making form in advance for its inverse function etc., also will be useful to its inverse operation.In addition,, can utilize general microprocessor to come computing, also aforesaid arithmetic expression can be carried out providing after ROM solidifies for edge detection process etc.

In addition, in the above description for ease of understanding, be that photo detector (line sensor) with one dimension is that example describes as photo detector.But certainly, also can utilize the sensor (face sensor) of two dimension to realize the present invention as photo detector.Subsidiary being noted that, as two dimension to be subjected to optical sensor to have photo detector to be arranged in chessboard cancellate, also be arranged in honey comb like, but no matter be any, as long as a plurality of axles for photo detector alinement shape adopt the embodiment of the relevant line sensor of above-mentioned explanation to get final product respectively.In addition, as long as in the scope that does not deviate from aim of the present invention, can do various distortion and enforcement.

As mentioned above, according to the present invention, owing to, utilize the output edge calculation position of this approximate expression, so the position that energy is simple and easy, high precision also detects the edge at high speed according to line sensor with the light intensity distributions that is subjected to of the approximate Fresnel diffraction generation of hyperbolic sine function sech (x).Especially with the low cheap line sensor of resolution the time, can fully improve its accuracy of detection, sizable practical function is arranged.

In addition, be subjected to the output of each photo detector of optical sensor, will be owing to can be used as because the light of the light intensity distributions formation that Fresnel diffraction produces is subjected to optical width direction (Fresnel diffraction direction) to carry out principal value of integral in its whole sensitive surface upper edge obtains, therefore the deviation that exists between the light intensity distributions of actual Fresnel diffraction generation can be proofreaied and correct effectively and the error that causes, the accuracy of detection of marginal position x0 and distance z can be improved widely.Especially owing in each photo detector, ask the light income of face upper integral with the hyperbolic sine function sech (x) of the light intensity distributions that is similar to Fresnel diffraction, and according to regarding this light income and aforementioned photo detector as any light income of obtaining poor, proofread and correct the light income of obtaining from its photo detector, just can fully improve its accuracy of detection effectively so do simple processing.Its result is even in the occasion of using the sparse cheap line sensor of for example photo detector arrangement, still can realize method for detecting position and device that accuracy of detection is high.

Claims

1. A position detection method comprising a line sensor having a plurality of light receiving elements arranged at predetermined intervals along a single direction, and the plurality of light receiving elements disposed opposite to the line sensor and facing the line sensor The light projecting part that projects unidirectional parallel light analyzes the output of the linear sensor in the edge detection part to detect the edge position of the blocker existing in the optical path of the monochromatic parallel light in the direction in which the light receiving element is arranged. characterized in that,

The edge detection unit detects the rising part of the light intensity distribution of the blocking object due to the Fresnel diffraction of the monochromatic parallel light from the light intensity received by each light receiving element of the line sensor, and also uses the dual A sinusoidal function sech(x) is used to approximate the light intensity variation among the light-receiving elements in the rising part, and then the hyperbolic sine function sech(x) is used to respectively analyze the light-receiving intensity distribution of the light-receiving elements, according to the analysis The position at which the received light intensity becomes specific in the direction in which the light receiving element is installed is obtained from the distribution of received light intensity at each of the light receiving elements, and this position is detected as the edge position of the shield.

2. The position detection method according to claim 1, wherein:

Utilize the hyperbolic sine function sech(x) to analyze the received light intensity distribution at each light receiving element of the linear sensor, is

When the outputs of the light-receiving elements of the line sensor are normalized to [1], the light-receiving elements whose light-receiving intensity is greater than [0.25] and the light-receiving elements whose light-receiving intensity is less than [0.25] are obtained respectively. After receiving the light intensity of the light-receiving element, the

Utilize the inverse function ln{[1+(1-Y ² ) ^1/2 ]/Y} of the hyperbolic sine function to transform the amount of light received at each of the above-mentioned light-receiving elements, and obtain the light-receiving at each of the above-mentioned light-receiving elements respectively intensity distribution,

The position where the light receiving intensity is [0.25] in the installation direction of the light receiving element is obtained from the light receiving intensity distribution at each light receiving element, and this position is detected as the edge position of the blocking object.

3. A position detection device comprising a linear sensor having a plurality of light-receiving elements arranged at predetermined intervals along a single direction, and projecting monochromatic light toward the plurality of light-receiving elements of the linear sensor disposed opposite to the linear sensor. The light projecting part of the parallel light, and the edge detection part analyzing the output of the linear sensor to detect the edge position of the blocking object existing in the optical path of the monochromatic parallel light in the installation direction of the light receiving element are characterized in that,

The edge detection unit analyzes the light intensity distribution generated by the Fresnel diffraction of monochromatic parallel light by the blocking object from the light intensity received by each light receiving element of the line sensor to obtain the obtained light intensity distribution. The detection part of the position of the edge of the occluder, including

The light-receiving element determination device is used to respectively determine the light-receiving element that obtains the light-receiving intensity greater than [0.25] according to the normalized output obtained by normalizing the output of each light-receiving element of the linear sensor and obtain the obtained light-receiving element. A light-receiving element whose light-receiving intensity is smaller than [0.25];

light-receiving position calculating means for each determined by said light-receiving element determining means by using an inverse function ln{[1+(1-Y ² ) ^1/2 ]/Y} of the hyperbolic sine function sech(x) converting the received light amount at the light receiving element to obtain the light receiving intensity distribution at each of the light receiving elements, and respectively obtaining the light receiving position on each of the above light receiving elements where the light receiving intensity is equal to the average light receiving intensity; and

The interpolation operation means is used for obtaining the light reception intensity in the installation direction of the light reception element by interpolation calculation according to the average light reception intensity and the light reception position on each of the above-mentioned light reception elements respectively obtained by the light reception position calculation means: [0.25], and detect this position as the edge position of the occluder.

4. The position detection device according to claim 3, wherein:

After normalizing the output of each light-receiving element of the line sensor to [1] in advance, the light-receiving element determining means respectively determines the light-receiving element that obtains a light-receiving intensity greater than a predetermined reference light-receiving intensity and obtains a light-receiving intensity higher than the above-mentioned reference light-receiving element. A light-receiving element with low light-receiving intensity.

5. The position detecting device according to claim 3 or 4, characterized in that,

The light-receiving element determining device is used to determine at least two adjacent light-emitting elements whose light-receiving intensity is around [0.25].

6. The position detection device according to claim 5, wherein:

Using the hyperbolic sine function sech(x) to analyze the received light intensity produced by each light-receiving element of the linear sensor, is

After the output of each light-receiving element of the linear sensor is normalized to [1], the light-receiving element with the peak value of the initial light-receiving intensity and the closest light-receiving element thereof are calculated respectively,

Utilize the inverse function ln{1+(1-Y ² ) ^1/2 ]/Y} of the hyperbolic sine function sech(x), transform the received light quantities at the above-mentioned light-receiving elements, and obtain the above-mentioned The light intensity distribution at the light receiving element,

The position where the light receiving intensity is [0.25] in the installation direction of the light receiving element is obtained from the light receiving intensity distribution at each of the above light receiving elements, and this position is detected as the edge position of the blocking object.

7. The position detection method according to claim 1, wherein:

When the output of each light-receiving element of the line sensor is normalized to [1], the light-receiving element Cn that obtains a light-receiving intensity greater than [0.25] and the light-receiving element Cn that obtains a light-receiving intensity smaller than [0.25] are respectively obtained. Element Cn-1,

The hyperbolic sine function sech(x) approximating the light intensity variation between the light receiving elements in the rising portion of the light intensity distribution on the light receiving surface of the linear sensor is integrated for each arrangement interval of the light receiving elements to obtain the After the received light amounts A(xn) and A(xn-1) of the respective light receiving elements Cn and Cn-1, the light receiving elements Cn, Cn- The center position Xn, Xn-1 of 1,

According to the center position Xn, Xn-1 of each light receiving element Cn, Cn-1, the light receiving amount A(xn), A(xn-1) of each light receiving element Cn, Cn-1 is interpolated, The position at which the received light intensity is [0.25] is obtained, and this position is detected as the edge position x0 of the obstruction.

8. The position detection method according to claim 1, wherein:

When the output of each light-receiving element of the linear sensor is normalized to [1], obtain the light-receiving element Cn whose light-receiving intensity is greater than [0.25], and obtain the light-receiving element Cn whose light-receiving intensity is smaller than [0.25] Light-receiving element Cn-1 of light intensity,

Based on the received light intensity A(xn), A(xn-1) of each light receiving element Cn, Cn-1 above, and the element arrangement direction position xn, xn-1 of each light receiving element Cn, Cn-1, using the approximate line The hyperbolic sine function sech(x) of the light intensity distribution between the light receiving elements in the rising part of the light intensity distribution on the light receiving surface of the shape sensor, respectively obtain the position where the light receiving intensity is [0.25] in the direction of the element installation After x and the distance z from the light-receiving surface of the linear sensor to the edge of the barrier,

Using the above-mentioned position x and distance z, when the light-receiving surfaces of the light-receiving elements Cn and Cn-1 are regarded as a point, the hyperbolic sine function sech(x) is used to find the values of the light-receiving elements Cn and Cn-1 respectively. Received light quantities Yn, Yn-1, at the same time, use the hyperbolic sine function sech(x) to obtain respectively the received light quantities yn, yn-1 after integration on the light receiving surfaces of the respective light receiving elements Cn, Cn-1,

The difference Δyn, yn between the received light amount Yn, Yn-1 when each light receiving element Cn, Cn-1 is regarded as a point, and the light receiving amount yn, yn-1 integrated on the light receiving surface of each light receiving element Cn, Cn-1 -1 is used as a correction amount to correct the received light intensities A(xn) and A(xn-1) of the light receiving elements Cn and Cn-1 respectively,

According to the corrected light receiving intensity A(xn)', A(xn-1)' of each light receiving element Cn, Cn-1 and the position xn, xn of the element arrangement direction of the light receiving element Cn, Cn-1 -1, use the hyperbolic sine function sech(x) again to find the position where the received light intensity is [0.25], and detect this position as the edge position x0 of the occluder.

9. The position detection method as claimed in claim 8, characterized in that,

The received light quantities Y(xn)', Y(xn-1)' corrected by considering each light receiving element Cn, Cn-1 as a point, and the received light quantity integrated on the light receiving surface of each light receiving element Cn, Cn-1 The difference Δy'n, ΔY'n-1 between yn and yn-1 is used as a new correction amount, and the correction of the received light intensities A(xn) and A(xn-1) of the light receiving elements Cn and Cn-1 is repeatedly performed .

10. The position detecting device according to claim 3, characterized in that, comprising

The light-receiving element determination device is used to separately calculate the light-receiving element whose light-receiving intensity is greater than [0.25] when the output of each light-receiving element of the linear sensor is normalized to [1] and the light-receiving intensity is less than [0.25] The light-receiving elements Cn, Cn-1 of the light-receiving intensity;

element position detecting means for applying the hyperbolic sine function sech(x) approximating the change in light intensity between the light receiving elements in the rising portion of the light intensity distribution on the light receiving surface of the linear sensor to each of the light receiving elements Integrate each arrangement interval, and obtain the center positions Xn, Xn, Xn-1;

The edge position detection device is used to measure the light receiving amount A of each light receiving element Cn, Cn-1 according to the center position Xn, Xn-1 of each light receiving element Cn, Cn-1 obtained by the element position detecting device. (xn) and A(xn-1) perform interpolation processing to obtain the position where the received light intensity is 0.25, and detect this position as the edge position x0 of the obstruction.

11. The position detecting device according to claim 3, characterized in that, comprising

The light-receiving element determination device is used for normalizing the output of each light-receiving element of the linear sensor to [1] to obtain the light-receiving element whose light-receiving intensity is greater than [0.25] and obtain the light-receiving intensity less than [0.25] ] light-receiving elements Cn, Cn-1 of light-receiving intensity;

an edge detection means for using said hyperbolic sine function sech(x) approximating a change in light receiving intensity between light receiving elements in a rising portion of light intensity distribution on a light receiving surface of said line sensor, based on said light receiving elements The received light intensity A(xn), A(xn-1) of Cn, Cn-1, and the positions Xn, Xn-1 of the element arrangement direction of each light receiving element Cn, Cn-1 respectively obtain the described light received intensity to be [0.25 ] position x and the distance z from the light-receiving surface of the linear sensor to the edge of the barrier,

The first received light intensity calculation device is used to calculate the received light of each of the light receiving elements Cn and Cn-1 by using the hyperbolic sine function sech(x) according to the position x and the distance z obtained by the edge detection device. The received light intensity Yn, Yn-1 of the light-receiving elements Cn, Cn-1 when the surface is regarded as a point;

The second received light intensity calculation means is used to calculate the hyperbolic sine function sech(x) by integrating the hyperbolic sine function sech(x) for each arrangement interval of each light receiving element according to the position x and the distance z obtained by the edge position calculation means. Output the received light intensity yn, yn-1 integrated on the light receiving surface of each light receiving element Cn, Cn-1;

A light-receiving amount correction device for a difference Δyn-1 between the light-receiving amounts Yn, Yn-1 when each light-receiving element Cn, Cn-1 is regarded as a point, and the light-receiving amounts yn, yn-1 integrated on the light-receiving surface , Δyn-1 is solved as the correction amount, and the received light intensity A(xn), A(xn-1) of each light receiving element Cn, Cn-1 is corrected respectively; and

Iterative calculation control device, used to arrange the corrected light receiving intensities A(xn)', A(xn-1)' of each of the light receiving elements Cn, Cn-1 and the element arrangement of the above light receiving elements Cn, Cn-1 The positions xn, xn-1 of the directions are supplied to the edge detection means and the calculation of the edge positions is performed again.