JP2000261617A - Image reader - Google Patents

Abstract

PROBLEM TO BE SOLVED: To attain accurate image reading by combining sets of reliable information according to two ranging methods so as to recognize shapes which are not affected by persons and background. SOLUTION: A shape is recognized by 1st and 2nd range finding methods of different reference with respect to a height of an object, and in the case that the reliability of shape data by the 1st ranging method is discriminated as being low, the difference from the shape data between the 1st and 2nd ranging methods is detected (#32). The shape data by the 1st or 2nd range finding method is corrected on the basis thereof.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image reading apparatus for reading a three-dimensional object such as a book from above, and more particularly to recognition of the shape of the object.

[0002]

2. Description of the Related Art Conventionally, in an image reading apparatus such as a face-up book scanner, a first detection for recognizing a shape from an upper side of a subject by a distance measuring mirror for detecting a height of the subject, and a subject from above. The second detection for recognizing the shape based on the protruding amount from the reference line when the image is viewed is performed, and the subsequent image reading process is performed by using either shape data.

More specifically, a CCD (Charge Coupled De) is used in accordance with the shape of a book as a subject.
Automatic focus adjustment to the image pickup means using Au
To perform focus, finger detection, and bend / shrinkage correction, the shape of the object is recognized using a part of the effective pixels of the CCD (corresponding to the image of the distance measuring mirror), and the original is scanned using the distance measuring mirror. It has first detection means for recognizing the shape from the upper side surface, and second detection means for recognizing the protruding shape when viewed obliquely from above, and employs either one of the detection results. Was.

[0004]

However, in the conventional image reading apparatus as described above, in the first detection using the distance measuring mirror, the detection accuracy of the height component is high, but the distance is measured depending on the light amount. There is a case where the accuracy of shape recognition becomes a problem due to the reflection of the clothes, arms and background of the person on the mirror, and it is determined that the mirror cannot be used as height data. On the other hand, in the second detection, the detection accuracy of the shape component is high, but the criterion for detecting the height is strictly defined by the detection algorithm. In the case of a book that is slightly larger, there is a problem that the height of the book contents (the inside of the cover) is calculated to be lower than the actual height if the book is separated from the reference position. For this reason, in either distance measurement alone, accurate “shape / height”
Was not obtained in some cases.

The present invention has been made to solve the above-mentioned problem, and synthesizes reliable information by the two distance measuring methods as described above to perform accurate shape recognition without being affected by a person or a background. It is an object of the present invention to provide an image reading apparatus which can read an image easily and has good operability.

[0006]

According to a first aspect of the present invention, there is provided a first shape recognizing means for recognizing a shape of a subject from read information of the subject, and a first recognizing means for recognizing the shape of the subject. In the case where it is determined that the reliability of the shape data by the second shape recognition means different from 1 and the first shape recognition means is low, the difference between the shape data by the first and second shape recognition means is determined. And based on it, the first
Alternatively, a correction means is provided for correcting the shape data of the second shape recognition means to recognize the shape of the subject.

In the first aspect of the present invention, when it is determined that the reliability of the shape data by the first shape recognition means is low, a difference between the shape data by the first and second shape recognition means is detected, Based on the correction, the shape data of the first or second shape recognition means is corrected.
Here, the shape recognition of the subject by the first shape recognition means is different from that by the second shape recognition means in terms of the height of the subject (for example, the direction in which the subject is viewed is different), and the first shape. When the shape data obtained by the recognition unit is not reliable, the two types of read information are combined to recognize the shape of the subject, and the reading accuracy can be improved.

According to a second aspect of the present invention, in the image reading apparatus according to the first aspect, the first shape recognizing means converts the information read from the image reflected on the distance measuring mirror for detecting the height of the subject. The second shape recognizing means recognizes the shape based on information on the amount of protrusion of the read image of the subject from the reference position.

According to a second aspect of the present invention, the first shape recognizing means detects height data from the side of the subject by using a distance measuring mirror for recognizing the shape of the subject. Detects the subject in an oblique direction from above, and detects height data from the amount by which the subject protrudes from the reference position. The difference of the effective area is extracted from the height data recognized by the second shape recognition means and the height data recognized by the first shape recognition means, and the height obtained by the first or second shape recognition means is extracted. The shape of the subject is recognized by correcting the difference with respect to the data.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment in which the image reading apparatus of the present invention is embodied in a book scanner will be described below with reference to the drawings. FIG. 1 shows the overall configuration of the book scanner. The book scanner 1 includes a main body housing 10 for housing an electric circuit and the like, and a dark platen for placing a document (subject).
20, an imaging unit 30 for converting a document image into an electric signal, and a lamp unit 40 for illuminating the document.
The document table 20 is arranged on the front side of the main body housing 10. The imaging unit 30 is arranged above the document table 20, and is supported in a cantilever manner by a support column 12 extending upward from the upper surface of the main body housing 10.

The lamp unit 40 is arranged on the lower surface side of the image pickup unit 30 and is fixed to the column 12. A space 80 between the document table 20 and the imaging unit 30 is open to a free space outside the apparatus, and has a sufficient size for setting a book document. The distance between the document table 20 and the imaging unit 30 is 30 cm or more.

An operation panel OP is provided at the upper end of the front surface of the main body housing 10, and a distance measuring mirror 16 for detecting the height of the document surface is fixed to the lower end. Distance mirror
The front surface of the document 16 is a glossy flat surface which is inclined at an angle of 45 ° with respect to the upper surface of the document table 20, and projects the side end surface of the document. The upper end surface of the distance measuring mirror 16 functions as a white plate 18 for shading correction. On the right side of the body housing 10 toward the operation panel OP,
A main switch 51 is provided. The stopper 7 is used for positioning the upper end of the document, and serves as a reference position for height detection. Between the stopper 7 and the distance measuring mirror 16, a second distance measuring area 8 described later is provided.

A start key 52 for instructing the user to start reading is provided on both sides of the document table 20 in the left-right direction.
Are provided one by one. Also, on the front side of the platen 20,
An armrest 25 is provided.

The imaging unit 30 has a line sensor 31, which is a CCD array, an imaging lens 32, and a mirror 33. The original screen is projected onto the light receiving surface of the line sensor 31 by the mirror 33 and the imaging lens 32. Imaging lens 32
Is provided so as to be movable in the front-rear direction, and is not shown.
Positioned by the AF mechanism. The line sensor 31 is attached to a movable body of a scanning mechanism (not shown),
The parallel movement is performed along the left-right direction (sub-scanning direction) M2 while keeping the arrangement direction of the elements up and down. By this parallel movement, a two-dimensional document image is captured.

The book scanner 1 having the above configuration is an image input unit suitable for reading a book document. By combining the book scanner 1 and the digital copying machine, a comprehensive copying system suitable for various originals can be configured.

Next, control during image reading will be described. FIG. 2 is an image read from above with the book document BD opened on the document table 20. When using the book scanner 1, the user operates the distance measuring mirror 16 on the platen 20.
The book document BD is pressed against a stopper 7 provided near the lower end of the book document to perform positioning, and the edges of the left and right pages are pressed by hand so as to keep the book document BD open. Since the book document BD is set in this way, the boundary with the document table 20 becomes the reference line for the document set. The center of the reference line is the reference position MB (see FIG. 7). The document table 20 is configured to be vertically movable independently of left and right. Thereby, the heights of the left and right pages when the pages are opened can be made substantially the same. FIG.
In the figure, regions for performing first and second shape detections described later are indicated by dotted frames.

In the book scanner 1, the same document is scanned twice (document reading). Book manuscript BD
Then, unlike the sheet document, the document surface is curved,
It is necessary to adjust the focus of imaging according to the bending state.
Processing for compensating for the difference in luminance is also required. Therefore, the bending state is detected in the first scan (hereinafter, referred to as a preliminary scan), and necessary processing is performed in the second scan (hereinafter, referred to as a main scan) based on the detection result. Image output to the external device is performed during the main scan. The reading mode includes a mode in which both the left and right pages are read at a time (scanning mode similar to a sheet document), and a mode in which the left and right pages are read separately. In both modes,
A preliminary scan and a main scan are performed on each page.

FIG. 3 is a plan view showing an example of reading a book document BD. FIG. 3A shows a book document on the document table 20.
FIG. 3B shows a state in which a BD is placed, and FIG.
31 shows a read image G0 by 31.

The read image G0 includes an image G1 of the reading surface of the book document BD, an image G20 of the document table 20, and the distance measuring mirror 1
It is composed of the image G18 reflected in 6. Image G18 within image G18
1, G182 indicates the shape of the end face (the part called “heaven” in the book) of the book document BD in the opened state. The portion of the image G18 other than the images G181 and G182 is a background image reflected on the distance measuring mirror 16. Since the boundary between the image G18 and the image G20 corresponds to the above-described reference line and is known, the boundary and the image
The height of the document surface can be calculated from the distance (the number of pixels) from the contour lines of G181 and G182. The boundary between the image G1 and the image G20 is
The determination can be made by utilizing the difference between the background color of the document surface and the color of the document table 20.

The upper and lower edges of the image G1 are curved because the height of the document surface is not constant.
That is, a subject near the imaging surface is imaged larger than a distant subject. At the time of the main scan, image processing for correcting the curved image G1 to an image when the height of the document surface is constant is performed based on the height information of the document surface obtained at the time of the preliminary scan.

Next, a block configuration of the signal processing will be described. FIG. 4 is a block diagram of the signal processing system 100 of the book scanner 1. The output of the line sensor 31 is output from the A / D converter 1
02 converts, for example, 8-bit image data D102.

At the time of the preliminary scan, the image data D102 is
Once stored in the temporary memory 104, it is stored in the RAM 10 via the CPU 101.
Transferred to 6. A detection unit that detects the shape of the document, that is, a luminance detection unit 112, a document height detection unit 113, and a document size detection unit 114 perform detection using the sampling data transferred to the RAM 106. However, since this data is used only for each detection system, the data may be thinned to a resolution lower than that of the main scan reading (for example, a CCD 8-pixel pitch). The luminance detecting unit 112 counts the luminance of each pixel for each predetermined line in the sub-scanning direction, creates a luminance histogram, and detects the background luminance for each predetermined line in the sub-scanning direction. The document height detector 113 detects the mirror background of the distance measuring mirror 16 and the edge (boundary) position of the document upper end surface. The document size detection unit 114 detects a document area on the document table 20. The CPU 101 calculates a background removal amount / image distortion correction amount / image masking position and the like based on the above detection results. Also, the CPU 101 determines the height based on the height data.
It also generates data for AF control.

At the time of the main scan, the image data D 102 is sent from the AD converter 102 to the image processor 103. Image processing unit 103
Is for image distortion correction, density correction,
Responsible for processing such as masking to erase the image of the user's hand holding the document. Image data that has undergone predetermined image processing is output to an external device such as a copying machine or a printer.

The CPU 101 controls a drive system including a lamp control unit 45 and a sensor drive unit 35 according to a program in the ROM 105. The sensor driver 35 supplies a predetermined clock signal to the line sensor 31.

Next, the reading operation will be described. FIG. 5 is a flowchart of the reading operation. In response to the start key 52 being turned on (# 1), the book scanner 1 turns on the light source of the lamp unit 40 and starts the preliminary scan.
(# 2). In the preliminary scan, while transferring the sampling thinned data to the RAM 106 (# 3), the original size detection /
Original surface luminance detection / original surface height detection / original base level detection are performed (# 5, # 6, # 7, # 8). After the completion of the preliminary scan, correction of the curvature of the document shape, creation of AF control information, and the like are performed, and then, the main scan is started (# 9).

In the main scan, image processing such as document distortion correction based on the document height detection result, edge emphasis and scaling, etc.
(# 10) and an AF operation for keeping the focus on the document surface in accordance with the scanning of the line sensor is performed (# 11), and the processed image data is sequentially output to the external device. When the main scan is completed, the light source of the lamp unit 40 is turned off, and a standby state is waited for a new start instruction.

Next, an outline of the document height detecting process will be described. FIG. 6 shows a “mirror image of the upper end surface of the original reflected on the distance measuring mirror 16 and a sampling image of the original surface on the original platen 20” obtained when the book-type original is set on the book scanner 1 and a preliminary scan is performed. Some parts are shown. The purpose of the image acquisition at the time of the preliminary scan is to perform sampling of the read image data necessary for performing various shape correction / image processing in the main scan. Execute.

As described above, in the image sample data of the distance measuring mirror 16 obtained by the preliminary scan, since the upper end surface of the original is illuminated by the lamp unit 40, the luminance data is close to the white side. On the other hand, the background portion where the document end face is not reflected has a small amount of incident light, and conversely has luminance data close to the black side. Basically, using the change in luminance, the edge of the upper end surface of the document can be detected by comparing the threshold value in the main scanning direction.

The height detected by the distance measuring mirror 16 is
Here, it is called a first distance measurement, and a configuration for that is a first distance detection unit (a component of the first shape recognition means in the claims).
Call. In addition to the “mirror image of the upper end surface of the original”, luminance data of the surface of the original placed on the original table 20 is simultaneously acquired from the sampled image data.

Another height can be detected from the luminance data of the original surface by measuring the amount of protrusion from the reference MS. Obtaining the height detection in this manner is referred to as a second ranging, and a configuration therefor is referred to as a second ranging detecting unit (a component of the second shape recognition unit in the claims).

FIGS. 7A and 7B each show a distance measuring mirror.
FIG. 16 is a model diagram showing a basic principle of document height detection by first and second distance measurements at 16. In FIG. 2B, the height direction is shown larger than that in FIG. If the document height detection unit 113 can detect the boundary address between the “distance measuring mirror background position” and the “document upper end surface position” at a certain sub-scanning position by a method described later, this address is determined by the main scanning direction line sensor 31. Can be obtained as the distance PXLEN from the start address MT. On the other hand, the reference position MB of the original height of 0 mm is also
By performing initial machine adjustment and the like, the distance PXSTD from the start address MT can be recognized as known data. Therefore, the document height pixel number HT at the sub-scanning position can be calculated by the following equation.

[Equation 1] HT = PXSTD-PXLEN

In the initial machine adjustment processing, how many pixels are projected on the line sensor 31 when the object having the specified height is projected on the distance measuring mirror 16 is measured / stored. By using this adjustment value, the detected document height pixel number HT can be converted to the actual document height (for example, in mm).

Similarly, in the second distance measuring section, the number HT2 of the original height pixels at the sub-scanning position can be calculated by the following equation.

[Equation 2] HT2 = PXSTD2-PXLEN2

The relationship between HT2 and the actual height data is determined by the mechanical configuration, and can be obtained by a simple coefficient. By using this coefficient, the detected document height pixel number HT2 can be converted to the actual document height (for example, in mm).

Hereinafter, the details of the document height detection processing at the time of the preliminary scan will be described with reference to the drawings. FIG.
9 shows a flowchart of document height detection, FIG. 9 shows first distance measurement data from the first distance measurement area, FIG. 10 shows second distance measurement data from the second distance measurement area, and FIG. Distance and second
FIG. 7 is a diagram illustrating a method of obtaining corrected second distance measurement data from differences in distance measurement. For document height detection, first, a distance measurement image is read (# 21 in FIG. 8A). Based on this, as shown in FIG. 9, the height / shape data is detected from the area where the first distance measurement is performed, that is, the side image of the upper part of the subject projected on the distance measurement mirror 16 (# 22). This is set as first distance detection data.

Next, as shown in FIG. 10, from the area where the second distance measurement is to be performed, the protruding amount from the stopper 7 (reference position) shown in FIG. 1 is extracted and converted into height data. Is detected (# 23). This is used as second distance detection data. Then, the shape recognition of the present invention is performed (# 24). For details, as shown in FIG. 8B, valid data determination is performed on the first distance measurement data and the second distance measurement data (# 31), and only the difference between the height data of the two is extracted (# 31). 32), an average value of the differences is obtained, and this is used as a correction amount (# 33). This correction amount is added to the second distance measurement data, and this is used as corrected second distance measurement data (#
34). Thus, it is determined which shape data can be used (# 25), and processing according to each case is performed (# 26, # 27, # 28). That is, if the first distance measurement data is usable, the first distance measurement data is set valid (# 26), and if the corrected second distance measurement data is usable, the corrected second distance measurement data is used. Is set valid (# 27), and if all the shape data cannot be used, a warning message is displayed (# 28).

Whether the first distance measurement data can be used is determined in the book document, for example, in the central area of the first distance measurement data shown in FIG. If this is not the case, it can be done by assuming that the reliability is low. The reason why the detection values (heights) of the left and right portions are large in the first distance measurement detection data in FIG. 9 is that an image of a hand holding down the document is reflected.

In the prior art, either one of the first distance detection data and the second distance detection data is used.
However, in the first distance measurement, as shown in FIG. 2, since the stopper 7 is used as a reference in the second distance measurement, if the subject moves away from the stopper 7, there is a problem that accurate height data cannot be obtained. .

In the case where it is determined that the first distance measurement cannot be used and the second distance measurement can be used when the subject is away from the stopper 7, the correction is performed with the wrong height data and the focus is adjusted. There was no image, or reading was prohibited due to a warning, and a required image could not be read.

On the other hand, in the present invention, the first distance measurement and the second distance measurement have different standards for height, the former is mainly based on height data, the latter is mainly based on shape data, In the first ranging, the reliability of the height component is high,
In the second ranging, the reliability of the shape component is high. It can be said that the present invention focuses on this point. Therefore, as described above, when it is determined that the first distance measurement cannot be used, a reliable data area of the height data of the first distance detection data and the second distance detection data is compared and calculated, and the difference is calculated. By calculating as a correction value of the second distance measurement data, the height detection accuracy of the second distance measurement data is improved. This makes it possible to recognize highly accurate height and shape data,
A high quality read image can be provided.

In implementing the present invention, height detection is performed by setting the reference of the second distance measurement from "absolute reference" to "relative reference". As shown in FIG. 7, the reference MS is used in the absolute reference, but the absolute reference is given by giving a height correction value using MS ′ in the relative reference. By using the relative reference MS ', even if the original set position is shifted with respect to the stopper, the original is corrected. The method of giving the correction value can be manually given from the operation panel OP in addition to the method of calculating from the first distance measurement and the second distance measurement by the above method.

[0042]

As described above, according to the present invention, the first shape recognizing means mainly performs shape recognition based on height data, and the second shape recognizing means mainly extracts shape data. If you determine that your data is unreliable,
The difference between the shape data by the two means is detected, and based on the difference, the shape data of the first or second shape recognition means is corrected to perform the shape recognition of the subject. This makes it possible to read a subject image with high accuracy and efficiency.

[Brief description of the drawings]

FIG. 1 is an external view of a face-up book scanner according to an embodiment of the present invention.

FIG. 2 is a diagram showing an image obtained by reading a book document placed on a document table of the face-up book scanner from above.

FIG. 3A is a plan view showing a state where a book document is placed on a document table, and FIG. 3B is a diagram showing an image read by a line sensor.

FIG. 4 is a block diagram of a signal processing system of the book scanner.

FIG. 5 is a flowchart of a reading operation.

FIG. 6 is a partial view of a sampling image.

FIG. 7 is a basic model diagram of document height detection.

FIG. 8 is a flowchart of a process of shape recognition by height detection according to the present invention.

FIG. 9 is a diagram showing shape recognition data based on first distance measurement.

FIG. 10 is a diagram showing shape recognition data obtained by a second distance measurement.

FIG. 11 is a diagram showing a correction method for shape recognition according to the present invention.

[Explanation of symbols]

 1: Book scanner 7: Stopper (reference position) 8: Second distance measurement detection area 16: Distance measurement mirror 20: Document table 31: Line sensor 101: CPU (first and second shape recognition means, correction means) 112 : Luminance detection unit 113: Document height detection unit BD: Book document G18: Image reflected on distance measuring mirror 16 G181: Shape of book document end surface G182: Book document end surface shape MT: Distance measuring mirror top MB: Measurement Lower end of distance mirror (based on lower end of first distance measurement) MS: Lower end of second distance measurement (based on lower end of second distance measurement) MS ': Lower end of second distance measurement according to the present invention (based on lower end of second distance measurement)

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 5B047 AA01 AB02 BA02 BB02 BC14 CB11 CB21 5C072 AA01 BA04 DA04 EA05 LA03 LA12 RA04 RA11 UA20 VA06 WA04 XA01

Claims (2)

[Claims] 1. A first shape recognizing means for recognizing a shape of a subject from read information of the subject, a second shape recognizing means different from the first for recognizing the shape of the subject, and the first shape recognizing means. When it is determined that the reliability of the shape data is low, the difference between the shape data by the first and second shape recognition means is detected, and the shape of the first or second shape recognition means is determined based on the difference. An image reading device comprising: a correction unit that performs correction on data to recognize a shape of a subject. 2. The method according to claim 1, wherein the first shape recognizing means recognizes a shape based on read information of an image reflected on a distance measuring mirror for detecting a height of the subject. 2. The image reading apparatus according to claim 1, wherein the shape is recognized based on information on an amount of protrusion of the read image of the subject from a reference position.