JP2001008100A - Infrared imaging device and element defect compensation method - Google Patents

Abstract

(57) Abstract: In an infrared imaging apparatus having an infrared detector having a plurality of elements and a defective element compensation method, the linearity characteristics of the elements are also determined. An optical system for performing infrared scanning, an infrared detector having a plurality of elements, and a reference heat source at room temperature.
, A high-temperature reference heat source 11, normal-temperature data from the normal-temperature reference heat source 10 during the invalid scanning period, and a high-temperature reference heat source 11.
And a signal processing circuit 6 for performing sensitivity correction of each element of the infrared detector 3, element determination by S / N determination or DC offset determination, and replacement of defective elements based on the high-temperature data obtained by Further, at the time of power-on / replacement command, the temperature of the reference heat sources 10 and 11 is changed to obtain the difference between the respective elements with respect to the average value of all the elements, and the element whose difference exceeds the threshold value is determined as a defective element. It has a processing function of performing linearity determination.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to an infrared imaging device and a method for compensating element defects in the infrared imaging device. An infrared imaging device scans a target direction optically, mechanically, or electrically using an infrared detection element, detects infrared light emitted from the target, and processes the detection signal. It is used in a wide range of fields such as surveillance cameras, night vision devices, thermography, remote sensing, and remote monitoring devices mounted on vehicles and airplanes.

Also, depending on the detection wavelength range of infrared rays, for example, 3
In general, the band is divided into a band of about 5 μm and a band of 8 to 10 μm.
A configuration using a material such as PtSi, InSn, or HgCdTe is applied, and a configuration using HgCdTe is applied in the 8 to 10 μm band. In addition, the infrared detector has one of a single element, a one-dimensional array element, and a two-dimensional array element as an element configuration. In the case of the one-dimensional array and the two-dimensional array, the sensitivity characteristics of each element are completely completed. Therefore, it is necessary to correct the variation. If there is an element that is not uniform even after sensitivity correction, it must be replaced with an adjacent normal pixel. The present invention relates to an infrared imaging device that performs such replacement of a defective element and a method for compensating for a defective element.

[0003]

2. Description of the Related Art FIG. 48 is an explanatory view of a conventional infrared imaging apparatus, wherein 401 and 402 are optical systems, 403 is an infrared detector in which infrared detecting elements are arranged in a one-dimensional array, for example, 40
4 is an amplifier, 405 is an AD converter (A / D), 406 is a signal processing circuit for performing sensitivity correction and the like, 407 is a DA converter (D / A), 408 is a monitor using a cathode ray tube, a liquid crystal display panel, or the like. Reference numeral 411 denotes a reference heat source.

[0004] The reference heat sources 410 and 411 are heat sources set at different reference temperatures, respectively. The optical system 402 is turned on during a period other than an effective scanning period for scanning and detecting infrared rays emitted from a target, that is, during an invalid scanning period. Via the reference heat source 41
The infrared rays from 0 and 411 enter the infrared detector 403, the sensitivity of each detection element is corrected in the signal processing circuit 406, and the infrared image of the target is displayed on the monitor 408.

FIG. 49 is an explanatory view of the optical system, and shows an outline of the optical systems 401 and 402 in FIG. 48, and the same reference numerals indicate the same parts. In FIG. 49, reference numerals 412 and 413 denote lenses constituting the optical system 401, reference numerals 414 and 415 denote lenses which constitute the optical system 402, reference numerals 416 and 418 denote condensing parts by lenses, and reference numerals 417 and 419 denote reflecting mirrors.

An optical system 402 using lenses 414 and 415
In the subsequent stage, an infrared detector 403 having a one-dimensional array configuration is arranged, and the optical element 402 scans in a direction orthogonal to the one-dimensional array direction of the detecting elements constituting the infrared detector 403, and two-dimensionally scans the target. Is what you do. Optical system 402
Is generally performed in correspondence with display scanning on the monitor 408. For example, when interlaced scanning is performed and two fields constitute one frame, the first field is an odd-numbered field. The second field scans the even-numbered lines in the second field, and uses the infrared detection output signal obtained by the interlaced scanning to perform scan conversion for detecting a specific target or for displaying on the monitor 408. This is performed in the circuit 406.

When the infrared detector 403 has a one-dimensional array or a two-dimensional array of multiple elements, the sensitivity characteristic is greatly deviated from the standard by utilizing the large correlation between the detection signals of adjacent elements. Means for performing control to replace a defective element with an adjacent element are known. That is, the switching control is performed so that the detection signal of the adjacent normal element is used instead of the detection signal of the defective element whose sensitivity characteristic or the like largely deviates from the reference.

[0008]

SUMMARY OF THE INVENTION Conventional infrared detector 4
It is general that the determination of each element constituting 03 is made based on one or both of S / N and DC offset. When those values deviate from the reference values, the element is determined to be a defective element, and the element replacement process using the detection output signal of the adjacent normal element without using the detection output signal of the defective element is performed. Done.

For each element of the infrared detector 403,
When the sensitivity is within the range of the reference value,
The correction processing is performed on the detection signal of each element.
However, actually, the relationship between the detection signal level and the temperature is generally a curve, and furthermore, this characteristic varies among the elements. Therefore, it was not easy to obtain an accurate infrared image only by the sensitivity correction. SUMMARY OF THE INVENTION It is an object of the present invention to obtain an accurate infrared image by correcting characteristics of each element of an infrared detector and performing processing of replacing an element that deviates from a reference value including a linearity characteristic.

[0010]

The infrared imaging apparatus according to the present invention comprises: (1) an infrared detector 3 comprising a plurality of elements;
A high-temperature reference heat source 11, a normal-temperature reference heat source 10, a reference heat-source control circuit 12 for controlling the temperatures of the high-temperature and normal-temperature reference heat sources, and an infrared detector that scans infrared rays in a target direction during an effective scanning period. 3, the optical systems 1 and 2 that scan infrared rays from the reference heat sources 11 and 10 at high temperature and normal temperature to enter the infrared detector 3 during the invalid scanning period, and the high temperature taken during the invalid scanning period. Based on the high-temperature data from the reference heat source 11 and the room temperature data from the room-temperature reference heat source 10, sensitivity correction corresponding to the element is performed, and a defect of each element constituting the infrared detector is determined. An infrared imaging apparatus comprising a signal processing circuit 6 for performing a process of substituting a signal by an element, wherein a reference heat source control circuit 12 is configured to switch between a high temperature and a normal temperature at the time of power-on / replacement command.
The temperature of 0 is switched to two levels for each predetermined number of fields, and is controlled to a predetermined constant temperature during normal operation. The signal processing circuit 6 converts the high-temperature data and the normal-temperature data captured during the invalid scanning period. Based on the sensitivity correction coefficient calculation, sensitivity correction, and S
/ N determination, DC offset determination, and / or linearity determination, and has a processing function of performing replacement address generation / update based on a result of element determination of a defective element and performing a replacement process on the defective element by each determination. Things.

(2) At the time of power-on / replacement command, the reference heat source control circuit 12 sets the temperature of the reference heat source 10 at room temperature to A [° C.] of the average temperature of the scene, The temperature of the heat source is controlled to A + α [° C.], and after a predetermined field, the temperature of the reference heat source 10 at room temperature is A + ΔT ₁ [° C.]
And the temperature of the high-temperature reference heat source 11 is A + α + ΔT ₂ [°
C], and during normal operation, the temperature of the reference heat sources 10, 11 at room temperature and high temperature is controlled to the initial state,
Further, the signal processing circuit 6 stores the normal temperature data acquired during the invalid scanning period in accordance with the temperatures of the reference heat sources 10 and 11 at normal temperature and high temperature at the time of power-on / replacement command and during normal operation. Based on the high-temperature data, a combination of the S / N determination, the DC offset determination, and the linearity determination based on the difference of each element with respect to the average value of all the elements is switched to perform the element determination, and the replacement address generation for the defective element by each determination is performed. -It has a processing function for updating.

(3) The method of compensating for a defective element comprises the steps of: an infrared detector 3 composed of a plurality of elements;
1, a reference heat source 10 at room temperature, and infrared rays in the direction of the target object during the effective scanning period and incident on the infrared detector 3, and infrared rays from the reference heat sources 11 and 10 at high and normal temperatures during the invalid scanning period. Optical systems 1 and 2 incident on the infrared detector 3
And a signal processing circuit 6 for determining each element of the infrared detector 3 based on high-temperature data from the high-temperature reference heat source 11 output from the infrared detector 3 and normal-temperature data from the normal-temperature reference heat source 10. A defective element compensation method for replacing a defective element signal with a normal element signal, wherein the temperature of the normal heat source 10 is set to A [° C.] and the high temperature The temperature of the heat source 11 is controlled to A + α [° C.], the normal temperature data and the high temperature data are acquired during the invalid scanning period, and the element is determined to be a defective element by one or both of the S / N determination and the DC offset determination. After the determination, a sensitivity correction coefficient is calculated, and then the temperature of the standard heat source
± ΔT ₁ [° C], and the temperature of the high-
+ Α ± ΔT ₂ [° C.], fetch normal temperature data and high temperature data during the invalid scanning period, compare the difference of each element with the average value of all elements for each element as a first linearity judgment, and compare it with a threshold value. The element exceeding the threshold value is determined as a defective element, and in the next normal operation, the temperature of the reference heat source 10 at normal temperature is set to A [° C.] and the temperature of the reference heat source 11 at high temperature is set to A + α.
The temperature is controlled to [° C.], the normal temperature data and the high temperature data are fetched during the invalid scanning period, and the element is determined as a defective element based on at least one of the S / N determination, the DC offset determination, and the linearity determination. And performing a process of replacing defective elements based on the logical sum of the determination results of the respective determinations.

(4) An infrared detector 3 composed of a plurality of elements, a high-temperature reference heat source 11, a normal-temperature reference heat source 10, and an infrared detector that scans infrared light in the direction of a target object during an effective scanning period. 3, optical systems 1 and 2 that scan infrared rays from the reference heat sources 11 and 10 at high and normal temperatures during the invalid scanning period and enter the infrared detector 3, and a signal processing circuit 6. The infrared detector 3 is based on the high temperature data from the reference heat source 11 and the room temperature data from the room temperature reference heat source 10.
Is a method of compensating for a defective element, and replacing a signal of a defective element with a signal of a normal element, wherein the temperature of the reference heat source 10 at room temperature is A [° C.]
In addition, the temperature of the high-temperature reference heat source 11 is controlled to A + α [° C.], the normal-temperature data and the high-temperature data are taken in the invalid scanning period, the first sensitivity correction coefficient is calculated, and then the normal-temperature reference The temperature of the heat source 10 is controlled to A + ΔT ₁ [° C.], and the temperature of the high-temperature reference heat source 11 is controlled to A + α + ΔT ₂ [° C.].
Of the first and second sensitivity correction coefficients are compared with a threshold value as a second linearity determination, an element exceeding the threshold value is determined as a defective element, and during normal operation, By controlling the temperature of the reference heat source 10 at normal temperature to A [° C.] and the temperature of the reference heat source 11 at high temperature to A + α [° C.], the normal temperature data and the high temperature data are acquired during the invalid scanning period, and the S / N The method includes a step of performing an element determination as to whether or not the element is a defective element by at least one of the determination, the DC offset determination, and the linearity determination, and performing a defective element replacement process based on a logical sum of the determination results of each determination. .

[0014]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is an explanatory view of an embodiment of the present invention. Reference numerals 1 and 2 denote an optical system, 3 denotes an infrared detector, 4 denotes an amplifier, 5 denotes an AD converter (A / D), 6 is a signal processing circuit,
7, a scan conversion circuit; 8, a DA converter (D / A);
11 is a reference heat source, 12 is a reference heat source control circuit, 13 is a control circuit, 21 and 22 are digital signal processors (DSP1 and DSP2), 23 to 26 are random access memories (RAM1 to RAM4), and 27 and 28 are leads. 3 shows only memories (ROM1 and ROM2). Below 2
1 and 22 are abbreviated as DSPs and 23 to 28 are abbreviated as memories.

Optical systems 1 and 2 for receiving infrared rays from a target object direction, an infrared detector 3, an amplifier 4 for amplifying a detection output signal, an A / D converter 5, and the infrared detector 3 during an invalid scanning period. Reference heat source 1 for injecting infrared light of reference temperature
The configuration of the D / A converter 7 for inputting an analog video signal to a monitor (not shown) can be the same as that of the conventional example.

In accordance with the scanning by the optical system 2, the trigger signals 1 and 2 are controlled by the control circuit at the respective scanning start timings for causing the infrared rays from the reference heat source 10 at normal temperature and the reference heat source 11 at high temperature to be incident on the infrared detector 3. 13 and a trigger signal 3 to the control circuit 13 at the scanning start timing of the effective scanning period.
3 to instruct the signal processing circuit 21 to start various arithmetic processing and to switch modes. This control circuit 13
Has a function of receiving the above-mentioned trigger signals 1, 2, 3 and a replacement command for performing replacement of a defective element during operation, operating as a sequence controller, and inputting a timing signal to the signal processing circuit 6. Have.

If the scanning efficiency of the optical system 2 is, for example, 80% and one field is 1/60 s, the effective scanning period is (1/60) × 0.8 = 13.33 [ms], and the invalid scanning period is , (1/60) × 0.2 = 3.33 [ms]
Becomes In this invalid scanning period, infrared rays from the reference heat sources 10 and 11 at normal temperature and high temperature are made incident on the infrared detector 3, and in the effective scanning period, infrared rays from the direction of the target via the optical system 1 are emitted. The light enters the infrared detector 3 according to the scanning by the optical system 2.

The signal processing circuit 6 includes DSPs 21 and 22.
And memories 23 to 28. The DSP 21 performs processes such as data acquisition, element determination, replacement address generation and updating, sensitivity correction coefficient calculation, and sensitivity correction.
The configuration for performing sensitivity correction will be described. Also memory (RAM
1) 23 is for storing a sensitivity correction coefficient and element determination data, and a memory (RAM2) 24 is for storing a sensitivity correction coefficient, a memory (RAM3) 25 and a memory (RAM4) 2
Reference numeral 6 is for storing a replacement address. Also memory (ROM1)
27 and the memory (ROM 2) 28
1, for storing the firmware of the DSP 22.

The scan conversion circuit 7 outputs the detection output signal of the infrared detector 3 which has been corrected by the signal processing circuit 6,
It is converted into a scanning method in accordance with, for example, the NTSC system for displaying on a monitor, and the defective element is replaced in accordance with the replacement address stored in the memories 25 and 26, that is, the defective element detection signal is converted into a normal element detection signal. This is replaced and output.

Further, the reference heat source control circuit 12
The temperature of 0, 11 is controlled to be constant based on the data from the infrared detector 3 which is converted into a digital signal by the AD converter 5. Assuming that the average value of the temperatures of the infrared image) is A ° C., for example, as temperature 1, the temperature of the reference heat source 10 is A [° C.], the temperature of the reference heat source 11 is A + α [° C.], and the temperature 2 is The temperature of the reference heat source 10 is A ± ΔT
_{At 1} ° C., the temperature of the reference heat source 11 is controlled to A + α ± ΔT ₂ . For example, α = 10, ΔT ₁ = −10, ΔT ₂ =
+20. The timing of such switching is instructed by the control circuit 13.

A power-on / replacement mode when the infrared imaging device is powered on to start operation and when a command to replace a defective element is input to the control circuit 13; The mode is switched to the normal operation mode for imaging processing of an infrared image. In the mode at power-on / replacement command, the infrared image processing is stopped, the quality of the element is determined, and the like. In addition, the infrared image processing is performed together with the determination of the quality of the element, the replacement processing of the defective element, the sensitivity correction processing, and the like.

FIG. 2 and FIG. 3 are explanatory diagrams of functions at the time of power-on / replacement command, and show the functions of the signal processing circuit 6 when the infrared detector 3 has a one-dimensional array of 180 elements. The first embodiment in which the linearity (1) is used (first linearity determination) at the time of power-on / replacement command and during normal operation will be described as the element determination.

That is, in the power-on mode / replacement command mode, the temperature of the reference heat source 10 at room temperature is set to the average value A of the scene, and the temperature of the reference heat source 11 at high temperature is A + α = B [°
C], based on the room temperature data from the reference heat source 10 and the high temperature data from the reference heat source 11 taken in the invalid scanning period, perform element determination by S / N and DC offset determination in the same manner as in the conventional example, and determine defective elements. Generates and updates the replacement address, takes in the normal temperature data and the high temperature data again, calculates the sensitivity correction coefficient, and then sets the temperature of the reference heat source 10 to A ± ΔT ₁ = C [° C.] The temperature of No. 11 was changed to B ± ΔT ₂ = D [° C.], sensitivity was corrected, data was acquired, and the average value of all elements was calculated.
A first linearity determination for obtaining a difference from the average value of all the elements is performed for each element. When the difference exceeds a threshold value, the element is determined to be a defective element, and a replacement address is generated and updated. Then, the temperature of the reference heat source 10 is changed to A [° C.], the temperature of the reference heat source 11 is changed to B [° C.], the mode is shifted to the normal operation mode, infrared image processing is performed, and S / N,
The replacement address is generated and updated by one or more of the DC offset and linearity determination processing, and the defective element is replaced in real time at the field interval.

In FIG. 2, reference numeral 30 denotes an intake (high-temperature) portion;
31 and 35 are adders, and 32, 35 and 37 are memories (RA
M1), 32 and 36 are dividers, 40 is a load (normal temperature)
The sensitivity correction coefficient calculation units 41 and 45 include adders, 42, 44,
47 is a memory (RAM1), 43 and 46 are dividers, each of which has a function corresponding to the first element of the infrared detector 3, and has the same function as the second element to the 180th element.

In FIG. 3, reference numeral 50 denotes a high-temperature linearity determination unit, 51 and 54 are adders, 52 is a divider, and 53 is B
A register, 55 is an absolute value calculation circuit, 56 and 58 are memories (RAM1), 57 is a comparator, 60 is a normal temperature linearity determination unit, 61 and 64 are adders, 62 is a divider, and 63 is B
Register, 65 is an absolute value calculation circuit, 66 and 68 are memories (RAM1), 70 is a linearity (1) determination unit, 71
Is an OR circuit (OR), 72 is a memory (RAM1), 80
Is a replacement address generation / update section, 81 is a replacement address generation section, 82 and 83 are memories (RAM3, RAM4), 8
Reference numerals 4 and 85 denote switching units.

Capture (normal temperature) / sensitivity correction coefficient calculator 40
During the invalid scanning period, the normal temperature data obtained by detecting the infrared ray from the normal temperature reference heat source 10 and performing A / D conversion is fetched.
High-temperature data obtained by detecting the infrared rays from 1 and AD-converted is taken in. For example, the data is taken in 16 fields for 32 lines. For example, one infrared detector
Assuming a unitary array configuration of 80 elements, 180 × 32 × 16
= 92160 data.

The adders 31, 41 and the memories 32,
42 and 3 for every 1 to 16 fields, respectively.
The accumulated values for the two lines are obtained, and divided by 33 and 43.
Is calculated by 1/32 to calculate the line average value.
4 and 44, add the average value of 16 fields by adders 35 and 45, and 1/1 by dividers 36 and 46.
The field average value, that is, the high temperature average value and the normal temperature average value are calculated by dividing by 16, and the high temperature average value and the normal temperature average value are stored in the memories 37 and 47. The high-temperature average value is stored in the memory 37 and the normal-temperature average value is stored in the memory 47 for each of the first to 180th elements constituting the infrared detector 3.

The high-temperature linearity determining unit 50
7 is added by the adder 51 to the high temperature average value corresponding to the first element to the 180th element, and the divider 52 divides by 1/180, and the high temperature average value of the first element to the 180th element is stored in the B register 53. The difference between the high-temperature average value for all the elements and the high-temperature average value stored in the memory 37 corresponding to each element is obtained in an adder 54, and an absolute value calculation circuit 55
The absolute value of the positive / negative difference value is obtained and stored in the memory 56 as a high temperature linear value. The comparator 57 compares the high-temperature linear value with the threshold value, determines that the element exceeding the threshold value is a defective element (“1”), and determines that the element exceeds the threshold value as a normal element (“0”). To be stored.

The normal temperature linearity determining unit 60 includes an adder 51 and a divider 6 similarly to the high temperature linearity determining unit 50.
2, the average value of the room temperature of all the elements is calculated and stored in the B register 63, the difference between the average value of all the elements and the average value of each element is calculated, and the absolute value is compared with the threshold value in the comparator 67. An element exceeding the threshold value is determined as a defective element (“1”), and an element exceeding the threshold value is determined as normal (“0”) and stored in the memory 68.

The linearity (1) judgment unit 70 stores the judgment values stored in the memories 58 and 68 in the memory 72 via the OR circuit 71 as the judgment values corresponding to the elements. That is,
An element determined as a defective element (“1”) in one or both of the high-temperature linearity determination and the normal-temperature linearity determination is stored in the memory 72 as a defective element. That is,
The logical sum of the determination result is stored in the memory 72.
Therefore, when the linearity characteristic of each element of the infrared detector 3 is largely deviated from the average, it is determined that the element is defective, and the element is replaced with a normal element.

The replacement address generating / updating unit 80 determines the normal element based on the element determination values (normal = “0”, defect = “1”) stored in the memory 72 in the replacement address generating unit 81. Without replacement, the defective element is replaced with, for example, the next adjacent element. For example, when the first element is a defective element, the detection signal of the first element is replaced with a detection signal of a normal second element adjacent to the first element. Therefore, the detection signal corresponding to one pixel by the second element is processed as a detection signal corresponding to two pixels. An address indicating this replacement is generated and stored in one of the memories 82 and 83. This memory (RAM
3, RAM 4) 82, 83 show a two-sided configuration, and for example, for one field, when one is updated, control to read from the other is performed.

In the normal operation, the detection output signal of the infrared detector 3 during the effective scanning period is subjected to real-time sensitivity correction and replacement address generation / update at one-field intervals. FIGS. 3, 4 and 5 are explanatory diagrams of functions during normal operation, in which the reference heat sources 10 and 11 are set to, for example, the aforementioned temperature 1
Fixed to. FIG. 4 shows the functions of the capturing (high-temperature) unit 30 and the high-temperature linearity determination unit 50, and the same reference numerals as those in FIGS. 2 and 3 denote the same parts. The capture (high-temperature) unit 30 obtains a high-temperature average value for all of the first to 180th elements over 16 fields for 32 lines in the invalid scanning period, and obtains a high-temperature linearity determination unit 50.
The comparator 57 determines whether the difference between the high-temperature average value and the high-temperature average value corresponding to each element exceeds a threshold value, and determines that the element exceeding the threshold value is a defective element (“1”). It is stored in the memory (RAM1) 58.

FIG. 5 shows the functions of the acquisition (normal temperature) / sensitivity correction coefficient calculating section 40 and the normal temperature linearity determining section 60, and the same reference numerals as those in FIGS. 2 and 3 indicate the same parts. The acquisition (normal temperature) / sensitivity correction coefficient calculation unit 40 calculates the normal temperature average value of all of the first to 180th elements of the infrared detector over 16 fields for 32 lines in the invalid scanning period, Linearity judgment unit 6
0 indicates that the comparator 67 determines whether or not the difference between the room temperature average value and the room temperature average value corresponding to each element exceeds a threshold value. The element exceeding the threshold value is determined as a defective element (“1”). And store it in the memory (RAM1) 68.

FIG. 6 shows the functions of a linearity determination section 70, a replacement address generation / update section 80, a sensitivity correction section 90, and a sensitivity correction coefficient calculation section 100. The element determination value from the unit 50 and the element determination value from the normal temperature linearity determination unit 60 are stored as element determination values in the memory 72 via the OR circuit 71.

The replacement address generating / updating unit 80 includes a replacement address generating unit 81 and memories 82, 83, 86, 87.
, Switching units 84 and 85, and an OR circuit 88 (OR), and stores the element determination value obtained at power-on in the memory 86. The memory 87 stores an element determination value of a logical sum of the element determination value stored in the memory 86 and the element determination value stored in the memory 72 of the linearity determination unit 70.
That is, the element determination value obtained during normal operation is added to the element determination value obtained at power-on, and the replacement address generation unit 81 indicates the defective element ("1") stored in the memory 87. A replacement address is generated according to the element determination value, and switching control of the switching units 84 and 85 is performed for each field, and stored in one of the memories 82 and 83.

The sensitivity correction section 90 includes an adder 91 and a multiplier 92, and the sensitivity correction coefficient calculation section 100 has a function including an adder 101, a divider 102, and a memory (RAM1) 103. , The difference between the high-temperature average value from the memory 37 (see FIG. 4) and the normal temperature average value from the memory 47 (see FIG. 5) is obtained by the adder 101, the reciprocal is obtained by the divider 102, and the memory is used as a sensitivity correction coefficient. 103.

In the adder 91, the sensitivity correction unit 90
The normal temperature average value is subtracted from the image data obtained during the effective scanning period, multiplied by the sensitivity correction coefficient from the memory 103 in the multiplier 92, and input to the scan conversion circuit 7 as image data to generate a replacement address. In accordance with the replacement address from the updating unit 80, the defective element is replaced with a normal element adjacent to the defective element and converted into an analog signal by the DA converter 8, which is added to the monitor 9 to display an infrared image. Therefore, an infrared image accurately representing a temperature distribution or the like can be displayed. The functions shown in FIGS. 2 to 6 determine whether or not a defective element is present by the first linearity determination at the time of power-on / replacement command and during normal operation. Here is an example of the case.

FIGS. 7 and 8 are explanatory diagrams of the operation sequence. The fields at the time of power-on or replacement command, the operation of the optical system, the operation sequence, DSP1, DSP2 (DSP21, 22 in FIG. 1) Operation. In the invalid period during the scanning period of the operation of the optical system in the field N (hereinafter, the invalid scanning period and the effective scanning period are simply referred to as the invalid period and the valid period), the data is fetched by the DSP 1 and the data is acquired as described above. In addition, the normal temperature reference heat source 10
Assuming that the average value of the scene is A, L = A [° C.]
The high-temperature reference heat source 11 is H = A + 10, where α = 10.
Set to [° C]. In this case, the number of acquired data is as follows: high-temperature data and normal-temperature data are 1 in each field.
80 (elements) × 32 (lines) = 5760.

Then, in field N + 15, 1
With respect to the high-temperature data and the low-temperature data taken in the six fields, an average value is obtained for all the elements, and element determination, replacement address generation and update, and replacement of defective elements are performed based on a difference from the average value. Next field N + 16 to N + 31
In this case, similarly to the fields N to N + 15, high-temperature data and normal-temperature data are taken in, and an average value is obtained, thereby calculating a sensitivity correction coefficient corresponding to each element. Then, the heat source temperature is changed. For example, the high-temperature reference heat source 11
Assuming that T ₂ = + 20, H = A + 20 [° C.], the standard heat source 10 at normal temperature is ΔT ₁ = −10, and L = A−10.
Set to [° C]. The field required for this heat source temperature change is a.

Therefore, the next field N + 32 + a to N
At + 48 + a, the sensitivity correction processing is performed by the DSP2, and based on the data corrected for the sensitivity by the DSP1, the element is determined based on the linearity characteristic, and the replacement address of the defective element is generated and updated, and then the heat source temperature is changed. Do. That is, as in the first case, the normal-temperature reference heat source 10 is set to L = A [° C.], where A is the average value of the scene, and the high-temperature reference heat source 11 is set to H = A + 10 [° C.]. .
Then, the process proceeds to the normal operation of processing the captured image signal.

FIG. 9 shows an operation sequence after shifting to the normal operation, and shows only the case of the fields N to N + 3, but the same operation is performed in the subsequent fields. In the fields N + 2 and N + 3, the description of the same processing contents as those in the previous fields N and N + 1 is omitted. During normal operation, image data is processed during the valid period, high-temperature data and normal-temperature data are fetched by the DSP 1 during the invalid period, and element determination based on linearity characteristics (first linearity determination) and replacement address generation / update are performed. , Sensitivity correction coefficient calculation and sensitivity correction processing. The DSP 2 is in charge of the sensitivity correction processing.

As described above, the normal temperature reference heat source 10 at A [° C.] and the high temperature reference heat source 1 at A + 10 [° C.]
By taking the normal temperature data and high temperature data corresponding to 1
By judging the S / N and the DC offset, whether the element is normal /
Defects are determined, and then normal temperature data and high temperature data are taken in to calculate a sensitivity correction coefficient.
0 is A-10 [° C], and the temperature of the reference heat source 11 is A
+20 [° C], and for the data subjected to the sensitivity correction, calculate the average value of all the elements, find the difference from the average value of all the elements for each element, compare the average value with the preset threshold value, and set the threshold value. An element with a value exceeding the value is determined as a defective element by the linearity judgment,
The replacement process is performed on a defective element based on a logical sum with the defective element in the determination of the S / N and DC offset.

FIGS. 10 and 11 are flowcharts showing the operation of the DSP. The fields N to N + 4 in FIGS.
For 8 + a, the DSP (DSP1) 21 shown in FIG.
And the operation with the DSP (DSP2) 22. That is, D
Initial values of SP1 and DSP2 are set. DSP1 fetches high-temperature data and low-temperature data during the invalid period. In field N + 15, fields N to N +
Based on the data acquired during the period 15, the element determination regarding the S / N and DC offset is performed in the same manner as in the conventional example, a replacement address for the defective element is generated, and the memory (RA
M3) Stored in 25.

A sensitivity correction coefficient is calculated in the field N + 31 based on the data obtained in the next field N + 16 to N + 31. Then, the heat source temperature is changed. That is, the reference heat source 10 is changed from A [° C] to C [° C], and the reference heat source 11 is changed from B [° C] to D [° C]. A field required for changing the temperature of the reference heat source is defined as a.

In the field N + 32 + a, D
SP2 performs a sensitivity correction process in the fields N + 32 + a to N + 37 + a using the sensitivity correction coefficients calculated in the previous field N + 31. Also DS
P1 performs a linearity determination based on the high-temperature data and the normal-temperature data taken in the fields N + 32 + a to N + 48 + a, generates a replacement address for the defective element, and stores it in the memory (RAM3) 25. Then, the heat source temperature is changed. This heat source temperature change is to return the temperatures of the normal temperature reference heat source 10 and the high temperature reference heat source 11 to the initial values of A [° C] and B [° C], respectively.

FIGS. 12 and 13 are flow charts showing the operation of the DSP. The D in the fields N, N + 1 and N + 2 are shown in FIGS.
The detailed processing of SP1 and DSP2 will be described. That is, after the initial value is set, in the field N, the DSP 2 performs the sensitivity correction process using the sensitivity correction coefficient calculated in the previous field N-1, and the DSP 1 performs the high temperature data and the normal temperature Data is taken in, and during the valid period, a sensitivity correction coefficient is calculated using the data for 32 lines acquired in this field N and the data for 32 lines in the past 16 fields. , And a replacement address for the defective element is generated.

In the next fields N + 1 and N + 2, the DSP 1 fetches high-temperature data and normal-temperature data during the invalid period.
Performs sensitivity correction coefficient calculation, element determination, and replacement address generation / update. Also, the DSP 2 determines that the previous fields N, N
The sensitivity correction processing is performed using the sensitivity correction coefficient calculated at +2.

FIG. 14 is a flow chart of data acquisition, showing a case where the infrared detector has a one-dimensional array configuration of 180 elements. The DSP 1 at the time of acquiring high-temperature data by the reference heat source 11 and normal-temperature data by the reference heat source 10 is shown.
3 shows a processing flow. In addition, one screen of the infrared image has, for example, a 180 × 480 (pixel) configuration as shown in FIG. That is, a case is shown in which one screen is constituted by 480 lines by scanning an infrared detector having a one-dimensional arrangement of 180 elements in the scanning direction indicated by an arrow.

FIG. 16 shows the RAM 1 (the memory 23 in FIG. 1).
FIG. 3 is an explanatory diagram of the infrared detector 3, which corresponds to 180 elements of the infrared detector 3;
Area for storing 7-bit data and areas H1 to H16 for storing high-temperature data for 16 fields
And an area L1 for storing room temperature data for 16 fields
To L16 are formed, and the illustrated data are stored.

DSP1 and DSP2 are arithmetic circuits and A,
The DSP 1 has a configuration including a plurality of registers such as B, C, and D registers. The DSP 1 writes high-temperature data from a bus line to an A register as shown in FIG. Transfer to and store. This is performed for one line for 180 elements. In this case, assuming that one cycle of the DSP is 40 ns, the processing time is 40 (ns) × 180 (elements) × 2 (cycles) = 14400 (ns).

High-temperature data is written from the bus line to the A register, written from the memory 23 (RAM1) to the B register, and the A register data and the B register data are added and stored in the A register. This is performed for 31 lines. Therefore, the processing time is 40 (ns) × 180.
(Element) × 31 (line) × 4 (cycle) = 8928
00 (ns). Next, the data of the A register is reduced to 1/3
2 and store the result in the A register. From the A register, fields H1 to H1 corresponding to fields in the memory 23 (RAM1) are stored.
Store in H16. In this case, it is 14.4 μs. Therefore, the line average value calculation processing time of the high temperature data is 92
1.6 μs.

Similarly, for the room temperature data, the room temperature data is written from the bus line to the A register, and is transferred from the A register to the memory 23 (RAM1) and stored. Note that, from the (A register) to the RAM1 (area (LB1 to LB3
2)) indicates processing steps performed only during normal operation in the embodiment described later. This is performed for one line for 180 elements, high-temperature data is written to the A register from the bus line, and written to the B register from the memory 23 (RAM1). Also in this case, (A register to RAM
1 (storage in areas (LB1 to LB32)) indicates processing steps performed only during normal operation in the embodiment described later.

Then, the data of the A register and the data of the B register are added and written into the A register, and this is
Perform for the line. Next, the data of the A register is
The result of the division is stored in the register A, and the room temperature average value of the result of the division is stored from the register A in the field corresponding areas L1 to L16 of the memory 23 (RAM1).

FIG. 17 is a flowchart for calculating the sensitivity correction coefficient. The average value of the high-temperature data is written to the A register from the areas H1 to H16 of the RAM 1, and the A register is
It is stored in the area of AM1 (see FIG. 16). And R
The average value of the high temperature data is read from the areas H1 to H16 of AM1 and written to the A register, and the average value of the high
The average value of the room temperature data is read from 16 and written to the B register, the data of the A register and the data of the B register are added and written to the A register, and the RAM is read from the A register.
The addition result is stored in the area of No. 1. This is performed over 15 fields, and then the data of the A register is
, And stores the result in the A register.
The division result is stored in the area of AM1. Thereby, a field average value of the high temperature data is obtained.

The same processing is performed for the normal temperature data.
The field average value of the room temperature data is obtained and stored in the area of the RAM 1. Then, the field average value of the high-temperature data is written into the A register from the area of the RAM 1 and the field average value of the normal temperature data is written into the B register from the area of the RAM 1, and the reciprocal of the data difference between the A register and the B register is obtained. The result is written into the register, and the calculation result is stored as a sensitivity correction coefficient from the A register to the area of the RAM 1.

FIG. 18 is a flow chart of the sensitivity correction processing, in which the sensitivity correction coefficient stored in the area of the RAM 1 is represented by A
Write to the register, and from this A register to RAM2 (FIG. 1)
In the memory 24) as a sensitivity correction coefficient value. The RAM 2 has, for example, 16 elements corresponding to 180 elements.
And a bit configuration area. And
The field average value is written from the area of the RAM 1 to the A register, and is written from this A register to the area of the RAM 1 as an offset correction coefficient value.

Next, image data is written from the bus line to the A register, a sensitivity correction coefficient value is written from the area of the RAM 1 to the B register, and an offset correction coefficient value is written from the area of the RAM 1 to the C register. The difference from the data in the C register is multiplied by the data in the B register, stored in the A register, and the sensitivity-corrected data from the A register is sent out to the bus line. The time required for this sensitivity correction is, for example, 10.40 ms.

FIGS. 19, 20 and 21 are flow charts of element determination. High-temperature data is written from the areas H1 to H16 of the RAM 1 to the A register,
1 and then, for 15 fields, write high-temperature data from the areas H1 to H16 of the RAM 1 to the A register, write an average value from the area of the RAM 1 to the B register, and write data of the A register and the B register. Add data and write to A register.
The addition result is stored in the area of AM1.

A of the addition result for 15 fields
The data in the register is divided by 1/16 and stored in the A register, and the result of the division is stored in the A register from the A register. That is, the field average value of the high temperature data is obtained.

The normal temperature data is written into the A register from the areas L1 to L16 of the RAM 1 and the field average value of the normal temperature data as a result of the division is set to A in the same manner as in the case of the high temperature data.
The data is stored in the area of the RAM 1 from the register.

Then, the content of the B register is set to 0 (FIG. 2)
0), the field average value of the high-temperature data from the area of the RAM 1 is written to the A register, the data of the A register and the data of the B register are added and stored in the A register. This is performed for 180 elements, and the average value of all elements of the high-temperature data is stored in the B register, with the addition result of the B register being 1/180.

Next, the field average value of the high temperature data is written into the A register from the area of the RAM 1, the difference between the data of the A register and the data of the B register is written into the A register, and the absolute value of the data of the A register is obtained. The data is stored in the A register, and stored in the area of the RAM 1 from the A register. That is, the difference from the average value corresponding to the element is stored in RAM1.
Is stored in the area.

The difference value is written to the A register from the area of the RAM 1 and compared with the judgment value (threshold). If the judgment value does not exceed the difference value, it is judged that the element is normal and the A register is set to “0”. If the difference value exceeds the determination value and is large, it is determined that the element is defective (abnormal element), and "1" is written to the A register. Then, the determination result is written from the A register to the area of the RAM1. That is, the determination result is written into the RAM 1 for each of the 180 elements.

The contents of the B register are set to 0, the field average value of the high temperature data is written from the area of the RAM 1 to the A register, and the result of adding the data of the A register and the data of the B register is stored in the B register. Register 1
The addition result for 80 elements is divided by 1/180, and the division result is stored in the B register. Also RAM1
, Data is written into the A register from the area, the difference between the data in the A register and the data in the B register is stored in the A register, the absolute value of the data in the A register is obtained and stored in the A register, and the RAM 1 To the area. Therefore, the absolute value of the difference between the high-temperature data for 180 elements is stored in the area.

The absolute value of the difference is written from the area of the RAM 1 to the A register, and the judgment value (threshold) is compared with the data of the A register. If the absolute value of the difference exceeds the judgment value and is larger, the defective element (abnormal) Element) and writes "1" into the A register. Otherwise, it is determined that the element is normal and A
Write "0" to the register. Then, from the A register to R
The determination result is written in the area of AM1. Therefore, the determination result for the high-temperature data is written in the area corresponding to 180 elements.

The field average value of the room temperature data is written from the area of the RAM 1 into the A register, the addition is performed for 180 elements using the B register, and the result of the addition is set to 1/180 and stored in the B register. Then R
The field average value of room temperature data from the area of AM1 is A
Write to the register, find the difference between the data in the A register and the data in the B register, store the absolute value of the difference in the A register,
The absolute value of the difference between the room temperature data for 80 elements is stored.

An absolute value is written from the area of the RAM 1 to the A register, and is compared with a determination value (threshold) to determine an element. As described above, when the determination value is exceeded, “1” is determined as a defective element (abnormal element), and when the determination value is not exceeded, “0” is determined as a normal element.
It is stored in the area of M1.

Then, the logical sum of the high-temperature data linearity determination result from the RAM1 area to the A register and the normal temperature data linearity determination result from the RAM1 area to the A register, and the linearity (1) from the A register to the RAM1 area. (The first linearity determination).

FIG. 22 is a flowchart of replacement address generation / update, showing the operation at power-on / replacement command. The D register and the A register are each set to 0, and the A register corresponds to 180 elements. It is determined whether or not the value is 180. If the value is not 180, linearity determination data is written from the area of the RAM 1 to the B register, and whether or not the value is "1" is determined. , A register is 0 or not. If 0, the C register is set to 1. If the D register is even, it is determined whether the D register is even or not. If the D register is even, the area from the C register to the RAM 3 (the memory 25 in FIG. 1) is determined. Write the judgment result of normal and defect to
+1 the contents of the register.

If the B register is not "1", the contents of the A register are written to the C register because the element is a normal element. If the A register is not 0, the content of the A register is decremented by 1 and written to the C register. Then, it is determined whether or not the above-mentioned D register is an even number. If the D register is even,
The normal / defective judgment result is written from the C register to the area of the RAM 4 (the memory 26 in FIG. 1), and the content of the A register is incremented by one. By doing this for 180 elements,
The content of the A register becomes 180. Then, the content of the D register is incremented by one. When it is determined whether or not the content of the D register is an even number, if the content is an odd number, the RAM 3 is switched to the RAM 4 and data is written from the C register to the RAM 4 area.

FIG. 23 is a flow chart of the replacement address generation / update, showing the case of normal operation.
Data is sequentially written from the RAM 1 to the A register for the elements, the linearity determination result is written from the area of the RAM 1 to the B register, and the logical sum of the A register and the B register is written to the A register. Then, the D register and the A register are set to 0, and it is determined whether or not the content of the A register is 180. Subsequent steps are the same as the corresponding steps in FIG. 22, and duplicate description will be omitted.

FIG. 24 is a time chart at power-on / replacement command, showing trigger signals 1 to 3 by a synchronizing signal of the scanning system, DSP operation, and DSP instructions. As described above, the trigger signals 1 to 3 are signals synchronized with the scanning of the optical system, and include the scanning start timing (high-temperature data acquisition start timing) in the invalid scanning period, the normal temperature data acquisition start timing, and the effective scanning period. 5 shows the scanning start timing. Then, assuming that the invalid scanning period of each field of the 1/60 cycle is 3.33 ms and the valid scanning period is 13.33 ms, the rising edge of the trigger signal 1 causes a D of 1.65 ms, which is half of the invalid scanning period.
In accordance with the SP instruction INST1, the DSP starts acquiring high-temperature data for 32 lines.

At the rising edge of the trigger signal 2 in the middle of the invalid scanning period, the DSP takes in room temperature data for 32 lines in accordance with the DSP instruction INST2. In this case, 921.6 μs is required to acquire the high temperature data,
This shows a case where 1.1520 ms is required to capture room temperature data. Then, by taking in the high temperature data and the normal temperature data over the fields N to N + 15, the field N + 1
5 during the effective scanning period due to the rising edge of the trigger signal 3 in accordance with the DSP instruction INST8.
/ N element determination and DC offset determination are performed, and replacement address generation / update is performed based on the element determination. In this case, it is assumed that it takes 5.91 ms for element determination and 166 μs for replacement address generation / update.

Although the illustration is omitted by repeating the same operation, high-temperature data and normal-temperature data are fetched by the same processing in the next field N + 16 to field N + 31, and the effective scanning of the field N + 31 is performed. The sensitivity correction coefficient is calculated in the period. Then, the temperature of the reference heat source is changed, and high-temperature data and normal-temperature data are fetched in fields N + 32 + a to N + 48 + a, where a is the number of fields at that time, and field N + 48
In the + a effective scanning period, the element is determined by the linearity determination, the replacement address is generated and updated, and then the temperature of the reference heat source is returned to the initial state, and the operation shifts to the normal operation.

FIG. 25 is a time chart at the time of normal operation, showing trigger signals 1 to 3, DSP1 operation and DSP1 instruction, DSP2 operation and DSP2 instruction. In the invalid scanning period, DSP1
According to the instructions INST1 and INST2,
With the temperature of the reference heat source kept constant, high-temperature data and normal-temperature data are fetched, and during the effective scanning period, the DSP1 operation of sensitivity correction coefficient calculation, element determination, and replacement address generation / update is performed according to the DSP1 instruction INST7. Also, D2 of sensitivity correction by DSP2 instruction
The operation becomes SP2.

In this case, the sensitivity correction coefficient was calculated to be 2.39 m.
s, 7.09 ms for element determination, and 195 μs for replacement address generation / update. The DSP 2 does not operate during the invalid scanning period, but does not operate during the effective scanning period.
2) A sensitivity correction process is performed according to instruction INST4. In this case, for example, in the case of 90 elements and 480 lines, 10.40 ms is required.

FIG. 26 shows the memories 24 to 26 in FIG.
FIG. 7 is an explanatory diagram of (RAM2 to RAM4), where RAM2 has an area as described above, and an offset correction coefficient and an address 181 to area of addresses 0 to 180 corresponding to 180 elements for storing 16-bit data. 360
The sensitivity correction coefficient is stored in the area. RAM3, R
The AM 4 has areas of addresses 0 to 180 for respectively storing 16-bit replacement address data.

In the first embodiment, the linearity (1) is set at the time of power-on / replacement command and at the time of normal operation.
A case where element determination by (first linearity determination) is used will be described. As a second embodiment, when power is turned on /
At the time of a replacement command, element determination based on linearity (1) is performed. At the time of normal operation, a logical sum of the determination results of the element determination based on the linearity (1), the element determination based on S / N, and the element determination based on the DC offset is performed. Can be used. In this case, the operation at the time of power-on / at the time of the replacement command is the same as that of the first embodiment described above, and the duplicated description will be omitted.

FIG. 27, FIG. 28, and FIG. 29 are explanatory diagrams of the function in the normal operation in the second embodiment. In the normal operation, the case where the sensitivity correction, the element determination and the S / N are performed, and the replacement is performed. In FIG. 27, reference numeral 180 denotes a capture (high temperature) unit, 110 denotes a sensitivity correction coefficient calculation unit, and
1,185,111 are adders, 182,184,18
7, 113 denotes a memory (RAM1), 183, 186 denotes a divider, and 112 denotes a reciprocal part.

In FIG. 28, reference numeral 120 denotes an acquisition (normal temperature) / sensitivity correction coefficient calculation unit, 130 denotes a noise value calculation determination unit, 121, 125, and 135 denote adders, and 122 and 12
4, 127, 131, 133, 136, and 138 are memories (RAM1); 123 and 126 are dividers;
Reference numeral 4,137 denotes a comparator.

In FIG. 29, reference numeral 140 denotes a difference signal judgment unit, 150 denotes an element judgment unit, and 160 denotes a replacement address generation /
Updating section, 170 is sensitivity correction section, 141, 143, 15
2, 162 and 163 are memories (RAM1), 142 is a comparator, 151 and 161 are OR circuits (OR), 164 is a replacement address generator, and 165 and 166 are memories (RAM).
3, RAM 4), 167 and 168 are switching units, 171 is an adder, 172 is a multiplier, 7 is a scan conversion circuit, 8 is a DA converter (D / A), and 9 is a monitor.

The acquisition (high temperature) section 180 and the sensitivity correction coefficient calculation section 110 in FIG.
6 has the same function as the sensitivity correction coefficient calculation unit 100 in FIG. 6, and the capture (normal temperature) / sensitivity correction coefficient calculation unit 120 in FIG. 28 has the capture (normal temperature) / sensitivity in FIG. It has the same function as the correction integer calculation unit 40, and also has the replacement address generation / update unit 160 and the sensitivity correction unit 170 shown in FIG.
Has the same functions as the replacement address generating / updating unit 80 and the sensitivity correcting unit 90 in FIG. 6, and a duplicate description will be omitted.

FIGS. 30 and 31 are explanatory diagrams of the RAM 1 at the time of element determination based on S / N, showing a case where an area of 17 × 16380 is used, in which respective data on the right side are stored. FIG. 32 is a graph showing the relationship between R and R when the element is determined based on the DC offset.
FIG. 4 is an explanatory diagram of AM1, showing a case where a 17 × 7200 area is used, and stores respective data on the right side. The same or different area is used for the RAM 1 in each process. The RAM 3 and the RAM 4 can have the same configuration as that of the above-described embodiment shown in FIG.

The noise value calculation judging section 130 in FIG.
Stores high-temperature data for 32 lines corresponding to 16 fields in the memory 131 (the areas LB1 to LB32 in FIG. 31), extracts the maximum value by the comparator 132, and
(The area of RAM 1) as the maximum DC value,
The minimum value is extracted by the comparator 134 and stored in the memory 134.
(Stored in the area of the RAM 1 as the minimum DC value.
Then, in the adder 133, (the maximum value of the DC value) −
By calculating (the minimum value of the DC value), a noise value is calculated and stored in the memory 136 (the area of the RAM 1). Comparator 1
The memory 37 compares the noise value with the threshold value, and determines that an element having a noise value exceeding the threshold value is an abnormal (defective) element ("1") and an element having a noise value not exceeding the threshold value is normal ("0").
(Area of the RAM 1).

The difference signal judging section 140 in FIG.
The difference between the high temperature average value and the normal temperature average value of the output of the adder 111 of the sensitivity correction coefficient calculation unit 110 is stored in the memory 141,
The comparator 142 compares the threshold value with the threshold value. If the threshold value is exceeded, the defective element (“1”) is determined. If the threshold value is not exceeded, a normal element (“0”) is stored in the memory 43. The element determination unit 150 includes the noise value calculation determination unit 1
The logical OR of the element determination value based on the noise from the memory 138 of the memory 30 and the element determination value based on the difference signal from the memory 143 of the difference signal determination unit 140 is stored in the memory 152 (the area of the RAM 1). Stored as

FIGS. 33 to 36 are flowcharts of element determination. FIGS. 33 to 35 show element determination by S / N, and FIG. 36 shows element determination by DC offset. Note that the flow of calculating the high-temperature field average value and the flow of calculating the normal-temperature field average value in FIG. 33 are the same as the flow illustrated in FIG. 17, and redundant description will be omitted.

In FIG. 34, as a difference signal calculation, a difference between a high-temperature field average value and a normal-temperature field average value is obtained, and as a device determination, the difference signal is compared with a determination value (threshold). When the difference signal is small, it means that the difference between the high-temperature data and the normal-temperature data is small. Therefore, if the difference signal is smaller than the judgment value, it is judged as an abnormal element and it is set to “1”. To “0”. This is stored in the RAM 1 as a determination value of the signal S.

Next, a noise value is calculated. That is, RAM1
The normal temperature data is written from the area LB1 (see FIG. 31) to the A register, and the normal temperature data is stored in the area of the RAM1. When the 180 elements are completed, the DC maximum value is stored in the RAM1 area by sequentially comparing the magnitudes of the 31 lines using the A register, the B register, and the C register, and the DC minimum value is stored in the RAM1 area. Is stored. Then, as shown in FIG. 35, the difference between the maximum value and the minimum value
It is stored in the areas PP1 to PP16 of M1 (see FIG. 31). That is, noise values for 16 fields are stored for each of the 180 elements.

Then, data is written from the area PP1 of the RAM1 to the A register, and the areas PP2 to PP1 of the RAM1 are written.
6, data is sequentially written to the B register, the A register and the B register are compared, and if A <B, the RAM
The noise value is stored in the area of No. 11, and if A <B, the process proceeds to the next step.

Next, data is written from the area of the RAM 1 to the A register and compared with a judgment value (threshold). If the data in the A register exceeds this determination value, "1" is written to the A register as an abnormal element, and if not, "0" is written to the A register as a normal element. The determination values for the noise N of “1” and “0” are stored in the area. Then, the logical sum of the determination value of the signal S and the determination value of the noise N is stored in the area of the RAM 1.
It is stored as an element determination result by the S / N determination.

In FIG. 36, the RAM 1 shown in FIG.
Using the areas L1 to L16 and the A and B registers, the average value of the normal temperature data for 15 fields is calculated, and the average value is stored in the area of the RAM 1. Next, a difference between the average value in the m-th field and the average value up to the (m-1) -th field is obtained, and the absolute value is determined as a DC determination value.
The data is stored in the area of the RAM 1 and compared with a predetermined judgment value (threshold). If the value is larger than the determination value, “1” is set as an abnormal element, and if not, “0” is set as a normal element in the area of the RAM 1 via the A register.
Store as offset judgment value.

The first embodiment uses the element determination based on linearity (1) (first linearity determination) at the time of power-on / replacement command and during normal operation. The embodiment performs element determination based on linearity (1) at power-on / replacement command, and performs element determination based on linearity (1), element determination based on S / N, and element determination based on DC offset during normal operation. In the third embodiment, the element is determined by the linearity (2) (second linearity determination) at power-on / replacement command, and the linearity (1) (first linearity) is determined during normal operation. Judgement).

In the element determination of the linearity (2) (second linearity determination), the temperature of the reference heat source 10 at normal temperature is set to A [° C.], and the temperature of the reference heat source 11 at high temperature is set to A + 10.
[° C] (α = 10), each reference heat source 1
A sensitivity correction coefficient is calculated using the normal temperature data and the high temperature data according to 0 and 11, and is set as a coefficient 1. Next, the temperature of the reference heat source 10 is set to A-10 [° C] (ΔT ₁ = -10), and the temperature of the reference heat source 11 is set to A + 20 [° C] (ΔT ₂ = + 20).
And the sensitivity correction coefficient is calculated using the normal temperature data and the high temperature data from the respective reference heat sources 10 and 11 to obtain a coefficient 2. The difference between these coefficients 1 and 2 is calculated, the difference is compared with a threshold, and an element exceeding the threshold is determined as a defective element.

FIGS. 37 and 38 are explanatory diagrams of functions at the time of power-on / replacement command. In FIG. 37, reference numeral 200 denotes a take-in (high temperature) section, 210 denotes a take-in (normal temperature) section, and 201 and 201.
05, 211 and 215 are adders, 202, 204 and 20
7, 212, 214, and 217 are memories (RAM1), 2
Numerals 03, 206, 213 and 216 denote dividers.

In FIG. 38, 220 and 230 are sensitivity correction coefficient calculation units (sensitivity) units, 240 is a linearity (2) determination unit, 250 is a replacement address generation / update unit, 7
Is a scan conversion circuit, 8 is a DA converter (D / A), 9 is a monitor, 221, 231, 241 are adders, 222, 232
Is a reciprocal part, 223, 233, 242 and 244 are memories (RAM1), 243 is a comparator, 251 is a replacement address generator, and 252 and 253 are memories (RAM3 and RAM3).
4), 254 and 255 are switching units.

Uptake (high temperature) section 200 and uptake (normal temperature)
The unit 210 has the same functions as the acquisition (high temperature) unit 30 and the acquisition (normal temperature) / sensitivity correction coefficient calculation unit 40 in FIG. 2, and performs a high temperature average value calculation process and a normal temperature average value calculation process. Is a similar process, and a duplicate description will be omitted.

The memory (RAM1) is, for example, as shown in FIG.
As shown in FIG. 7, a 17 × 6840 area is used to store data as shown on the right side. The sensitivity correction coefficient calculation (sensitivity) section 220 in FIG.
When the temperatures of 0 and 11 are A [° C] and B [° C], the difference between the normal temperature average value and the high temperature average value is obtained, and the reciprocal of this difference value is obtained by the reciprocal part 222 to correct the sensitivity. The coefficients are stored in the memory 223 (the area of the RAM 1 shown in FIG. 39). This is set as a coefficient 1.

The sensitivity correction coefficient calculation (sensitivity) section 230
The difference between the normal temperature average value and the high temperature average value when the temperatures of the reference heat sources 10 and 11 are C [° C] and D [° C] is determined, and the reciprocal of the difference value is determined by the reciprocal part 232 to obtain the sensitivity. The correction coefficient is stored in the memory 233 (the area of the RAM 1 shown in FIG. 39). This is referred to as coefficient 2.

The linearity (2) determination unit 240 calculates the difference between the coefficient 1 and the coefficient 2 and determines the difference between the coefficients 1 and 2 in the memory 24.
2 (area of RAM 1), and the comparator 243 compares the threshold value with the threshold value. If the threshold value is exceeded, it is determined that the element is abnormal.
A determination value of “1” for an abnormal element and “0” for a normal element is stored in 44 (area of RAM1).

The replacement address generation section 251
Based on the judgment value from 4, the replacement address of the element of the infrared detector is generated, stored in the memories 252 and 253, and the scanning conversion circuit 7 performs the replacement process on the image data.

FIGS. 40 and 41 are operation flowcharts of the DSP. DSP1 and DSP2 (see FIG. 1) perform initial value setting, and DSP1 captures high-temperature data and normal-temperature data during an invalid period. Do not work. In field N + 15, high-temperature data and normal-temperature data are acquired for 32 lines each, a sensitivity correction coefficient (1) is calculated in the next effective period, and this is set to coefficient 1 as described above, and then a heat source By changing the temperature and performing the same processing, the sensitivity correction coefficient (2) is calculated in the effective period of the field N + 31 + a, and this is set as the coefficient 2.

In the next field N + 32 + a, as shown in FIG. 41, element determination based on linearity (2) determination is performed using coefficient 1 and coefficient 2, and replacement address generation / update is performed according to the determination result. , The replacement address is stored in the memory 252 (RAM3), and is input to the scan conversion circuit 7 as the replacement address in the next field.

FIG. 42 is a flow chart of the linearity (2) determination. The sensitivity correction coefficient, that is, the coefficient 1 is written into the A register from the area of the RAM 1 (see FIG. 39).
The sensitivity correction coefficient, that is, the coefficient 2, is written from the area of AM1 to the B register into the B register, a difference between the A and B registers is obtained, and the difference is stored in the area of RAM1 as the coefficient difference. Then, the difference between the coefficients in this area is written into the A register, and is compared with a determination value (threshold). When the data of the A register is larger than the judgment value (threshold), “1” is stored as an abnormal element, and when the data is not larger, “0” is stored as a normal element from the A register to the area of the RAM 1 as a linearity (2) judgment value. .

FIG. 43 and FIG. 44 are explanatory diagrams of the operation sequence, showing the field, the optical system operation, the operation sequence, the DSP1 operation and the DSP2 operation. In the fields N to N + 15 in FIG. Assuming that the temperature of the heat source is H = L + 10 [° C.] (A = L, α = 10), and L = the average value of the scene (A), the DSP 1 takes in data during the ineffective period and the sensitivity correction coefficient during the effective period. The calculation is performed, the heat source temperature is changed, and then, as shown in FIG.
= L + 20 [° C] (L = A, ΔT ₂ = + 20), L =
Assuming that the average value of the scene is −10 [° C.] (ΔT ₁ = −10), data acquisition and sensitivity correction coefficient calculation are similarly performed in the fields N + 16 + a to N + 31 + a, and linearity is calculated in the validity period of the field N + 32 + a. (2) Element determination by determination, replacement address generation / update,
Perform the replacement process.

In the normal operation, as in the first embodiment, the element determination is performed by the linearity (1) determination, and the replacement processing is performed on the defective element in real time using the detection signal of the normal element. , Overlapping description will be omitted.

As a fourth embodiment, at the time of power-on / replacement command, linearity (2) determination (second linearity determination) is performed, and during normal operation, linearity (1) determination (first linearity determination) is performed. Linearity determination), S / N determination, and DC offset determination. That is, at the time of power-on / replacement command, as described with reference to FIGS. 37 to 44, the temperature of the reference heat source is changed to obtain the respective sensitivity correction coefficients, and the linearity (linearity) is determined based on the difference between the sensitivity correction coefficients. 2) A determination is made, and during normal operation, the linearity (1) determination is performed as described with reference to FIGS.

At power-on / replacement command, the linearity (2) is determined. During normal operation, sensitivity correction and DC
The functions of the embodiment for performing the offset determination are shown in FIGS.
As shown in FIG. In FIG. 45, reference numeral 300 denotes a take-in (high temperature) portion, 301 and 305 denote adders, and 302 and 30.
Reference numerals 4 and 307 denote memories (RAM1), and 303 and 306 denote dividers.

In FIG. 46, reference numeral 310 denotes an acquisition (normal temperature) / sensitivity correction coefficient calculator, 320 denotes a DC offset value calculator, 311, 315, 321 and 325 denote adders,
12,314,317,322,324,327,32
9 is a memory (RAM1), 313, 316 and 323 are dividers, 326 is an absolute value circuit, and 328 is a comparator.

In FIG. 47, 330 is a sensitivity correction coefficient calculation (sensitivity) section, 340 is a sensitivity correction section, 350 is a replacement address generation / update section, 331 and 341 are adders, 332
Is a reciprocal part, and 333, 352 and 353 are memories (RAM
1) and 342 are multipliers, 351 is an OR circuit (OR), 3
54 is a replacement address generation unit, 355 and 356 are memories (RAM3, RAM4), 357 and 358 are switching units, 7
Denotes a scan conversion circuit, 8 denotes a DA converter (D / A), and 9 denotes a monitor.

Capture (high temperature) section 300 and capture (normal temperature)
The sensitivity correction coefficient calculation unit 310 has, for example, the same function as the capture (high temperature) unit 30 and the capture (high temperature) sensitivity correction coefficient calculation unit 40 in FIG.
The high temperature average value is stored in the area corresponding to each element of the memory 37, and the normal temperature average value is stored in the area corresponding to each element of the memory 47, and a duplicate description will be omitted.

The DC offset value calculation section 320 adds the line normal temperature average values of 1 to 15 fields stored in the memory 314 and stores the sum in the memory 322. The divider 323 multiplies the field normal temperature average value by 1/15. Is stored in the memory 324, the difference between the field normal temperature average value and the line normal temperature average value is obtained in the adder 325, the absolute value circuit 326 obtains the absolute value of the difference, and
7 is stored as a DC offset value, is compared with a threshold value in a comparator 328, an element having a DC offset value exceeding the threshold value is determined as an abnormal element, and “1” of the determination value is stored in a memo 32
9 and the element having a DC offset value not exceeding the threshold value is determined as a normal element, and the determination value “0” is stored in the memory 329.

The sensitivity correction coefficient calculation (sensitivity) section 330 uses the reciprocal of the difference between the high temperature average value from the memory 307 and the normal temperature average value in the memory 317 as the sensitivity correction coefficient for the memory 333.
And the sensitivity correction coefficient is input to the multiplier 342 of the sensitivity correction unit 340. This sensitivity correction unit 340 corrects the sensitivity of each element of the infrared detector by subtracting the normal temperature average value from the memory 317 from the image data and multiplying the image data by the sensitivity correction coefficient from the memory 333 in the multiplier 342. .

The replacement address generating / updating section 350 obtains the logical sum of the element determination value from the memory 329 and the element determination value obtained when the power is turned on from the memory 352, and stores the logical sum as a new element determination value. 353. The replacement address generation unit 354 generates a replacement address based on the element determination value and stores the replacement address in the memory 355 or 356. This replacement address is input to the scan conversion circuit 7, and signal processing is performed in a state where a defective element having an element determination value of "1" is replaced by a normal element having an element determination value of "0".

The present invention is not limited to the above embodiments. For example, the number of elements of the infrared detector 3 can be set to a number other than 180. The number of lines and the number of fields for taking in the normal temperature data and the high temperature data can be arbitrarily selected in accordance with the scanning efficiency, the operation capability of the DSP, and the like. Also, various combinations of the first linearity determination, the second linearity determination, and other determinations such as the S / N determination can be applied. When the power is turned on, the first linearity determination or the second linearity determination includes a determination as to whether the element is a defective element or a normal element. Therefore, when the operation shifts to the normal operation, at least the linearity characteristics are substantially uniform. An infrared image by the element can be obtained. Then, when a predetermined time has elapsed from the normal operation, an operation based on the replacement command may be automatically performed to determine the linearity.

[0115]

As described above, according to the present invention, the linearity judgment is added to the conventional S / N judgment and DC offset judgment as means for judging each element of the infrared detector 3. By the element determination by the linearity determination, it is possible to guarantee the uniformity of the characteristics of each element of the infrared detector 3 composed of a plurality of elements, and to cope with the characteristic change due to the change of the ambient temperature. it can. Therefore, there is an advantage that the accuracy of the infrared image displayed on the monitor can be significantly improved. Further, even during normal operation, it is possible to perform real-time determination of presence / absence of a defective element and replacement processing of the defective element, including linearity determination together with S / N determination, DC offset determination, etc. There is an advantage that the reliability of the device can be improved.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram of an embodiment of the present invention.

FIG. 2 is an explanatory diagram of functions at the time of power-on / when a replacement command is issued.

FIG. 3 is an explanatory diagram of functions at the time of power-on / when a replacement command is issued.

FIG. 4 is an explanatory diagram of functions during a normal operation.

FIG. 5 is an explanatory diagram of functions during a normal operation.

FIG. 6 is an explanatory diagram of functions during a normal operation.

FIG. 7 is an explanatory diagram of an operation sequence.

FIG. 8 is an explanatory diagram of an operation sequence.

FIG. 9 is an explanatory diagram of an operation sequence.

FIG. 10 is an operation flowchart of a DSP.

FIG. 11 is an operation flowchart of a DSP.

FIG. 12 is an operation flowchart of a DSP.

FIG. 13 is an operation flowchart of the DSP.

FIG. 14 is a flowchart of data acquisition.

FIG. 15 is an explanatory diagram of a screen configuration.

FIG. 16 is an explanatory diagram of a RAM 1;

FIG. 17 is a flowchart of sensitivity correction coefficient calculation.

FIG. 18 is a flowchart of a sensitivity correction process.

FIG. 19 is a flowchart of element determination.

FIG. 20 is a flowchart of element determination.

FIG. 21 is a flowchart of element determination.

FIG. 22 is a flowchart of replacement address generation / update.

FIG. 23 is a flowchart of replacement address generation / update.

FIG. 24 is a time chart at power-on / replacement command.

FIG. 25 is a time chart during normal operation.

FIG. 26 is an explanatory diagram of RAM2 to RAM4.

FIG. 27 is an explanatory diagram of functions during a normal operation.

FIG. 28 is an explanatory diagram of functions during a normal operation.

FIG. 29 is an explanatory diagram of functions during a normal operation.

FIG. 30 is an explanatory diagram of the RAM 1 at the time of element determination based on S / N.

FIG. 31 is an explanatory diagram of the RAM 1 at the time of element determination based on S / N.

FIG. 32: RAM1 at the time of element determination by DC offset
FIG.

FIG. 33 is a flowchart of element determination.

FIG. 34 is a flowchart of element determination.

FIG. 35 is a flowchart of element determination.

FIG. 36 is a flowchart of element determination.

FIG. 37 is an explanatory diagram of functions at power-on / replacement command.

FIG. 38 is an explanatory diagram of functions at the time of power-on / when a replacement command is issued.

FIG. 39 is an explanatory diagram of a RAM 1;

FIG. 40 is an operation flowchart of the DSP.

FIG. 41 is an operation flowchart of a DSP.

FIG. 42 is a flowchart of linearity (2) determination.

FIG. 43 is an explanatory diagram of an operation sequence.

FIG. 44 is an explanatory diagram of an operation sequence.

FIG. 45 is an explanatory diagram of functions during a normal operation.

FIG. 46 is an explanatory diagram of functions during a normal operation.

FIG. 47 is an explanatory diagram of functions during a normal operation.

FIG. 48 is an explanatory diagram of a conventional infrared imaging device.

FIG. 49 is an explanatory diagram of an optical system.

[Explanation of symbols]

 1, 2 optical system 3 infrared detector 4 amplifier 5 AD converter (A / D) 6 signal processing circuit 7 scanning conversion circuit 8 DA converter (D / A) 10, 11 reference heat source 12 reference heat source control circuit 13 control circuit 21, 22 DSP1, DSP2 23 to 26 memory (RAM1 to RAM4) 27, 28 memory (ROM1, ROM2)

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 2G065 AB02 AB03 BA02 BA15 BA34 BB06 BB11 BB49 BC07 BC14 BC16 BC28 BC33 BD01 DA01 DA18 2G066 AA06 BA13 BA14 BA22 BA25 BA26 BA27 BC02 BC07 BC09 BC21 CA08 CB01 5C024 AA06 AA11 CA09 FA01 FA HA08 HA12 HA18

Claims (4)

[Claims] An infrared detector comprising a plurality of elements; a high-temperature reference heat source and a normal temperature reference heat source; a reference heat source control circuit for controlling the temperatures of the high-temperature and normal temperature reference heat sources;
An optical system that scans infrared rays in a target direction during the effective scanning period and enters the infrared detector, and scans infrared rays from the high-temperature and normal-temperature reference heat sources and enters the infrared detector during the invalid scanning period. Based on high-temperature data from the high-temperature reference heat source taken during the invalid scanning period and normal-temperature data from the normal-temperature reference heat source, perform sensitivity correction corresponding to the element, and each element constituting the infrared detector And a signal processing circuit for performing a process of determining a defect of the defective element and replacing a signal of the defective element with a signal of the normal element. The reference heat source control circuit, when power-on / replacement command The temperature of the reference heat source between the high temperature and the normal temperature is switched in two stages for each predetermined number of fields, and has a configuration in which the temperature is controlled to a predetermined constant temperature during a normal operation. A sensitivity correction coefficient calculation, a sensitivity correction, an element determination based on at least one of S / N determination, DC offset determination or linearity determination, and a defective element An infrared imaging apparatus having a processing function of generating and updating a replacement address based on an element determination result and performing a replacement process on a defective element based on each determination. 2. The reference heat source control circuit sets the temperature of the normal temperature reference heat source to A [° C.] and sets the temperature of the high temperature reference heat source to A + α at the time of the power-on / replacement command.
[° C], and after a predetermined field, the temperature of the standard heat source at normal temperature is controlled to A + ΔT ₁ [° C], and the temperature of the high-temperature reference heat source is controlled to A + α + ΔT ₂ [° C], and the normal operation is performed. A temperature control unit that controls the temperature of the reference heat source between the normal temperature and the high temperature to the initial state, wherein the signal processing circuit is configured to control the normal temperature during the power-on / replacement command and during the normal operation. Based on the normal temperature data and the high temperature data acquired during the invalid scanning period, corresponding to the temperature of the reference heat source at the high temperature, the S / N determination, the DC offset determination, and the difference of each element with respect to the average value of all the elements are performed. 2. The infrared imaging apparatus according to claim 1, further comprising a processing function of performing element determination by switching a combination with the linearity determination by the above, and generating and updating a replacement address for a defective element by each determination. 3. An infrared detector constituted by a plurality of elements, a high-temperature reference heat source, a normal-temperature reference heat source, and an infrared ray directed to a target object during an effective scanning period, and the infrared ray is incident on the infrared ray detector. An optical system that scans infrared rays from the high-temperature and normal-temperature reference heat sources during the invalid scanning period and enters the infrared detector, and a signal processing circuit, and the high-temperature reference heat source output from the infrared detector In the defective element compensation method of determining each element of the infrared detector based on the high temperature data obtained by the method and the normal temperature data obtained by the normal temperature reference heat source and replacing the signal of the defective element with the signal of the normal element, At the time of raising / replacement command, the temperature of the normal temperature reference heat source is set to A [° C], and the temperature of the high temperature reference heat source is set to A +
The temperature is controlled to α (° C.), the normal temperature data and the high temperature data are fetched during the invalid scanning period, and the sensitivity correction coefficient is determined after the element determination as to whether or not the element is defective by one or both of the S / N determination and the DC offset determination. The temperature of the reference heat source at room temperature is controlled to A ± ΔT ₁ [° C], and the temperature of the reference heat source at high temperature is controlled to A + α ± ΔT ₂ [° C]. And the high-temperature data, the difference of each element from the average value of all the elements is compared as a first linearity determination with a threshold, and an element exceeding the threshold is determined as a defective element. A [° C]
The temperature of the high-temperature reference heat source is controlled to A + α [° C.], and after the sensitivity is corrected, the normal temperature data and the high-temperature data are acquired during the invalid scanning period, and the S / N determination or the DC offset determination or the linearity determination is performed. Determining whether or not the element is a defective element by at least one of the following determinations; and performing a replacement process of the defective element based on a logical sum of the determination results of the respective determinations. . 4. An infrared detector constituted by a plurality of elements, a high-temperature reference heat source, a normal-temperature reference heat source, and an infrared ray in a direction toward a target object during an effective scanning period, and incident on the infrared ray detector, An optical system that scans infrared rays from the high-temperature and normal-temperature reference heat sources during the invalid scanning period and enters the infrared detector, and a signal processing circuit, and includes high-temperature data obtained from the high-temperature reference heat sources and the normal temperature. In the defective element compensation method of determining each element of the infrared detector based on the normal temperature data from the reference heat source and replacing the signal of the defective element with the signal of the normal element, The temperature of the normal temperature reference heat source is set to A [° C], and the temperature of the high temperature reference heat source is set to A +
α [° C.], taking in room temperature data and high temperature data during the invalid scanning period to calculate a first sensitivity correction coefficient. Then, the temperature of the reference heat source at room temperature is calculated as A + ΔT ₁ [° C.]
And the temperature of the hot reference heat source is A + α + ΔT ₂ [°
C], the normal temperature data and the high temperature data are taken in the invalid scanning period to calculate a second sensitivity correction coefficient, and a difference between the first and second sensitivity correction coefficients corresponding to the elements is calculated as a second sensitivity correction coefficient.
In comparison with a threshold value, the element exceeding the threshold value is determined as a defective element. During normal operation, the temperature of the normal temperature reference heat source is set to A [° C].
Further, the temperature of the high-temperature reference heat source is controlled to A + α [° C.] to acquire normal temperature data and high temperature data during the invalid scanning period, and at least one of S / N judgment, DC offset judgment, and linearity judgment A defective element compensation method, comprising: performing element determination as to whether or not the element is a defective element by one determination, and performing a replacement process of the defective element based on a logical sum of the determination results of each determination.