JPH10224600A - Image input device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately and quickly recognize the direction of an original and to output image data in an appropriate direction in simple constitution. SOLUTION: From the color image data of an original read in an image read part, after replacing the color data of a chromatic color with white data in a color data cancellation circuit 520 beforehand, characters are segmented in a character segmentation circuit 530 and the feature amount of the segmented characters is extracted in a feature amount extraction circuit 540. The feature amount extraction circuit divides the feature amount into feature amount data depending on the direction of the characters and feature amount piece number information not depending on the direction of the characters, lists up the characters to be recognition candidates based on the feature amount piece number information first and outputs dictionary data relating to the candidate characters to a comparator circuit 580. In the comparator circuit 580, the dictionary data are rotated by 90 deg. and compared with the feature amount data of the segmented characters and the information of the rotation angle with a highest matching degree is sent to a CPU 501. The CPU 501 considers that the segmented character and the recognition candidate character match in the case that the matching degree is more than a prescribed threshold value, the rotation processing of the image data for the pertinent rotation angle is performed and they are outputted.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image input device used as an image reader or a document reading section of a copying machine.

[0002]

2. Description of the Related Art Conventionally, for example, in a copier equipped with an automatic document feeder, when a plurality of documents are continuously read and copied, a copy is discharged in accordance with the direction of the read document. When documents are set on the automatic document feeder, it is necessary to check whether each document is uniformly directed in the same direction. However, such a check operation is quite time-consuming, and if the number of documents is large,
The possibility of overlooking the document in the opposite direction increases accordingly.

[0003] If the operation of checking the direction of the original is neglected,
When a predetermined sort mode is executed in a copying machine having a sort unit that sorts and discharges a plurality of output bins, it is necessary to correct the copy direction for each copy bundle discharged to each output bin. Furthermore, in a copier equipped with a stapling unit for automatically stapling and a punch unit for punching holes, when those functions are executed, a copy bundle including copies in different directions is stapled as it is. In the worst case, the copying operation has to be performed again. In this case, the efficiency of the copying operation is greatly reduced, the copy paper is wasted, and the resource is not saved.

In order to avoid such inconvenience, the orientation of the original is first determined from the read image data (this determination of the orientation of the original is hereinafter referred to as “top and bottom recognition”), and the image output is in an appropriate direction. As described above, a method of rotating image data and outputting the image data has been considered. For example, JP
Japanese Unexamined Patent Publication No. 229763 discloses the following method of recognizing the top and bottom.

That is, a plurality of points (reference points) in a line portion of a predetermined reference character are previously extracted and stored as a pattern, and a character image is cut out from image data obtained by reading an original image. The cut-out character is compared with the reference point, and the presence or absence of a character image signal at each reference point is checked while rotating the cut-out character image by 90 ° to determine the degree of coincidence at each rotation angle.
The rotation angle having the highest degree of coincidence is recognized as the direction of the character, thereby performing top-bottom recognition.

More specifically, for example, as shown in FIG. 23 (a), six reference points A1 to A
6 is extracted and stored, and when the cut-out character image has “A” oriented 90 ° leftward as shown in FIG. 23B, the reference points A2 and A5 Since there is no character image signal, the degree of coincidence is low. This character image is rotated 90 ° at a time, and the rotation angle with the highest degree of coincidence with the reference point is determined to recognize this as the direction of the character. I have.

[0007]

However, in the above-described conventional top / bottom recognition method, since only the presence or absence of an image signal at a reference point is compared for a specific character, it is determined whether the character read in the first place is "A". Whether it is or not is unknown, and it is necessary to store data of many reference characters in an internal memory and sequentially switch and compare them. Also,
Even if the cut-out character is a character of "A", the stored data of "A" and the cut-out character image may be exactly the same size and the same font, but the character size may be different. If the font changes even a little, the image signal will not be located at the position of the reference point even though the directions of the characters match, and the recognition rate will decrease.

Such a decrease in the recognition rate becomes remarkable when the original is a color original and the cut out characters are chromatic characters. This is because, as seen in newspapers and magazines, chromatic characters are often used for article headings, etc., to attract readers' attention, intentionally tilting,
This may be due to the fact that there are many characters that are exaggerated with some decoration or deformation, such as extreme bolding.

SUMMARY OF THE INVENTION The present invention has been made in view of the above-described problems, and it is an object of the present invention to improve the accuracy of recognizing a document direction in a color document and to output an image by rotating color image data in an appropriate direction. An object is to provide an input device.

[0010]

In order to achieve the above object, an image input apparatus according to the present invention is an apparatus for inputting an image of a color document, and is an apparatus for reading a document and generating color image data. Reading means, image data storage means for storing the color image data, color data detection means for detecting chromatic image data in the color image data, and other black and white except for the detected chromatic color image data Document direction determining means for determining the direction of the document based on a character image in the image data, and image data rotating means for rotating the color image data according to the determined document direction, I do.

The present invention also provides a color space detecting means for converting the color image data into a color space represented by luminance and two kinds of chromaticity; And chromatic color discriminating means for discriminating the image data as chromatic color image data when any of the values exceeds a threshold value set for each.

Further, according to the present invention, the document direction discriminating means replaces the detected chromatic image data with white data to generate black and white image data; Character extraction means for extracting the relative position of the local shape in the character image or the adjacent direction thereof as a character characteristic amount, and a characteristic amount storage for storing the character characteristic amount of a predetermined character in advance Means, comparison means for comparing the character feature quantity extracted by the feature quantity extraction means with the character feature quantity stored in the feature quantity storage means, and recognition for recognizing the direction of the character image based on the comparison result. Means for distinguishing the direction of the character image recognized by the recognition means from the direction of the document.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A case where an image input apparatus according to the present invention is applied to a full-color digital copying machine will be described below. (1) Overall Configuration of Full-Color Digital Copier First, the overall configuration of a full-color digital copier (hereinafter, simply referred to as "copier") will be described with reference to FIG.

As shown in FIG. 1, the copying machine includes an automatic document feeder 100 that automatically feeds a document, an image reading unit 200 that reads an image of the fed document, and data of the read document image. And a printer unit 300 for forming an image on a recording sheet based on the image data. Automatic document feeder 1
00 is a document feed tray 11, a transport roller group 12,
It is composed of a registration roller 13, a conveyor belt 14, and the like. The bundle of documents set on the document feed tray 11 is fed downward one by one from the lowermost document by the conveying roller group 12, and is timed by registration rollers 13.
The document is conveyed to the document reading position on the document glass plate 21 of the image reading unit 200 by the conveyance belt 14.

After the image is scanned by the image reading unit 200, the transport belt 14 is driven again and is discharged onto the document discharge tray 16 via the discharge roller 15. When the back side of the original is also scanned, just before the leading end of the original reaches the discharge roller 15, the switching path 17 switches the original conveyance path to the reversing roller 18, and the reversing roller 18 turns the original. , And transported again toward the original glass plate 21.

The image reading section 200 includes the original glass plate 2
1 for optically reading an image of a document conveyed to the first document reading position, comprising a scanner 22, a condenser lens 23,
It comprises a CCD image sensor 24, a scanner motor M2 and the like. The scanner 22 is provided with an exposure lamp 25 and a mirror 26 for changing the optical path of light reflected from the original by irradiation of the exposure lamp 25 in a direction parallel to the original glass plate 21, and is driven by a scanner motor M2 to The original on the original glass plate 21 is scanned by moving in the direction of the arrow.

A pulse motor is used as the scanner motor M2, and the position of the scanner 22 is determined by the photoelectric scanner home sensor SE1 installed at the home position.
The amount of movement from that position,
That is, control is performed based on the number of steps of the pulse motor. The reflected light from the original is reflected by the mirror 26,
CCD via mirrors 27 and 28 and condenser lens 23
The light is guided to the sensor surface of a color image sensor (hereinafter, simply referred to as a “CCD sensor”) 24, where it is converted into an electric signal to generate image data for each color component (R, G, B).

The mirrors 27 and 28 are paired so as to move in the same direction at half the moving speed of the scanner 22, thereby reducing the optical path length from the scanner 22 to the condenser lens 23. The configuration is such that the original image is always formed on the sensor surface of the CCD sensor 24 while being kept constant. The image data generated by the CCD sensor 24 is
After being A / D-converted into a digital signal in an image signal processing unit 420 (see FIG. 2) in the control unit 400, and subjected to necessary processing such as shading correction, density conversion, and edge enhancement, the image signal is processed for each color component. Image memory 431
(FIG. 2).

The image data stored in the image memory 431 is subjected to rotation processing according to the result of the top / bottom recognition performed by the document discrimination unit 500 as described later, and finally, cyan (C), magenta (M), The data is converted into yellow (Y) and black (K) print data and output to the exposure head unit 310 of the printer unit 300, or transmitted to another copying machine or computer via the communication interface 451 via a telephone line or the like. Sent.

The printer unit 300 forms an image on a recording sheet such as a copy sheet or an OHP film sheet by a known electrophotographic method. The printer unit 300 includes an exposure head unit 310, an image forming process unit 320, , Paper feed unit 3
30 and a re-feed unit 340. Exposure head 3
Reference numeral 10 includes four exposure heads 311c, 311m, 311y, and 311k corresponding to the above-described cyan (C), magenta (M), yellow (Y), and black (K) print data. , A laser diode and a polygon mirror.

Similarly, the image forming process section 320 also has four image forming units 321c, 321m, 321y, and 321c corresponding to the C, M, Y, and K print data.
21k, and each of the photosensitive drums 322
c, 322m, 322y, and 322k, and a cleaner, a charging charger, a developing unit, and the like are arranged around the photosensitive drums 322c. , 3
At 22m, 322y and 322k, a toner image corresponding to the color component is formed.

For example, when cyan printing data is output to the laser diode 312c of the exposure head 311c, a modulated laser beam is emitted from the laser diode 312c, and this laser beam is rotated at a predetermined angular velocity by a polygon mirror. The surface of the photosensitive drum 322c, which is reflected and reflected by the mirror surface 313c and is driven to rotate clockwise in FIG. 1, performs exposure scanning, whereby an electrostatic latent image is formed on the surface of the photosensitive drum 322c. The electrostatic latent image formed on the photosensitive drum 322c is supplied with cyan toner particles by the developing device 323c, and is visualized as a toner image.

On the other hand, the paper supply unit 330 has three paper cassettes 331 to 33 for storing recording sheets of different sizes.
3, a recording sheet of a predetermined size is fed from the paper cassette 331, for example, in synchronization with the image forming operation on the photosensitive drum 322c, and the transfer belt 334 transfers the recording sheet below the photosensitive drum 322c. And the toner image on the surface of the photosensitive drum 322c is transferred to the recording sheet surface by the electrostatic force of the transfer charger 324c.

Such an image forming operation is executed while shifting the timing in the other exposure heads 311m, 311y, 311k and the image forming units 321m, 321y, 321k, and the toner images of the respective colors are transferred onto the recording sheet in a superimposed manner. Thus, a full-color image is reproduced. Thereafter, the recording sheet is fixed by the fixing unit 327 and is discharged onto the discharge tray 342 via the transport path 341 in the re-feeding unit 340.

If printing is to be performed on the back side of the recording sheet (so-called double-sided copying), the re-feeding section 34
0, the recording sheet is conveyed by changing the direction of the switching claw 343.
After being guided in the 44 direction and once extruded into the reversing conveyance path 345, it is sent out to the intermediate tray 347 via the conveyance path 346 by a conveyance roller (not shown), and the recording sheet is stored in a state where the recording sheet is turned upside down. Then, the intermediate tray 347
By feeding the recording sheet therein to the transfer belt 334, an image is formed on the rear surface thereof, and then discharged onto the discharge tray 342 via the fixing unit 327.

Note that SE2 in the printer unit 300
Is a timing sensor, which is a transfer belt 33 for transfer.
4, a reference mark (not shown) is detected, and the conveyance timing of the recording sheet is adjusted. Further, each image forming unit 321c, 321m, 321y, 32
Each of the registration correction sensors 325c inside 1k,
m, y, and k are built in, and the transfer belt 334 for transfer is used.
By sequentially detecting the above-mentioned reference marks, the output timing of the print data of each color component is finely adjusted to prevent color shift in image formation due to each color component.

The roller 328 on the rightmost side in the drawing among a plurality of rollers around which the transfer belt 334 is transferred is
It is supported by a vertically movable frame (not shown), and when forming a black-only monochromatic image, the frame is moved downward by a vertically moving device (not shown) to transfer a portion on the right side of the auxiliary roller 334a. Conveyor belt 334
Are separated from the photosensitive band drums 322c, 322m, 322y of the image forming units 321c, 321m, 321y. At this time, the transfer charger 326
c, 326m, and 326y also move downward together with the frame.

As a result, during the formation of a single-color image, the transfer belt 334 is brought into non-contact with the photosensitive drums 322c, m, and y, and the corresponding image forming units 321c, 321c,
Since the driving of m and y can be stopped, the abrasion of each of the photosensitive drums 322c, m and y and the process units around them can be reduced. The image forming units 321c, m, y, and k are integrated in each process and can be attached to and detached from the main body, so that maintenance such as toner replacement can be easily performed. Further, an operation panel 70 (see FIG. 2) is provided at a position on the front of the image reading unit 200 where the operation is easy, whereby the operator performs a predetermined input operation such as start of copying and setting of the number of copies. . (2) Configuration of Control Unit 400 Next, the configuration of the control unit 400 installed inside the copying machine will be described with reference to the block diagram of FIG.

As shown in FIG. 3, the control unit 400 includes an image reading control unit 410, an image signal processing unit 420, a memory control unit 430, a printer control unit 440, an external communication control unit 450, a main control unit. 460 and the document discrimination unit 500
Etc. Each of the control units 410 to 460 and the document discrimination unit 500 are configured with a CPU as a center. The control units 410 to 460 transmit and receive mutually necessary data and commands via a command line CR, and further, via an image data bus GB. Image data is transferred.

The image reading control unit 410 controls the operations of the automatic document feeder 100 and the image reading unit 200 to execute reading of a document. That is, when a predetermined copy mode is set from the operation panel 70 and a copy start operation is performed, a document reading request is received via the main control unit 460, and the operation of the automatic document feeder 100 is controlled to copy the document. The document is conveyed to a predetermined position on the document glass plate 21 of the reading unit 200, and thereafter, the operation of each unit of the image reading unit 200 is controlled so that the document is scanned by the scanner 22 and the image data read by the CCD sensor 24 is converted into an image. The signal is transmitted to the signal processing unit 420.

The image signal processing section 420 includes an A / D converter, a shading correction section, an MTF correction section, a scaling section,
The image data of the input document is converted into a digital multi-level signal by an A / D converter, and the shading correction unit reduces unevenness in the illuminance of the exposure lamp 25 and unevenness in the sensitivity of the CCD sensor 24. After correction, MTF
After undergoing processing for improving image quality such as edge enhancement in the correction unit, and further performing scaling processing and γ correction processing in the scaling unit and the γ correction unit, respectively, Can be

The document discrimination section 500 determines the direction of the read document based on the image data and generates a predetermined rotation angle signal, or a specific document, such as a bill, for which copying is prohibited. In the case of a document, a copy prohibition signal is generated. Details will be described later. Memory control unit 430
Receives the read request from the main control unit 460 while binarizing or further encoding the image data and storing it in the image memory 431 in page units.
The image data of the target page is read out from the image memory 431 (after decompression processing is performed if compressed) and returned to multi-valued data. After rotating the image data as necessary, the image data is transferred to the printer control unit 440. This rotation processing is performed by a known technique for changing a memory address of image data (for example, see Japanese Patent Application Laid-Open No.
26769).

The printer control unit 440 operates the internal R for the image data output from the memory control unit 430.
The operation of each part of the printer unit 300 is controlled based on the control program stored in the OM to execute image formation on a recording sheet. The external communication control unit 450 controls transmission and reception of image data and the like to and from an external device such as another copying machine, a facsimile machine, or a computer. The external communication control unit 450 receives image data received from the external device via the communication interface 451. The image data of the document stored in the image memory 431 and read by the image reading unit 200 of the own device is transmitted to the external device via the communication interface 451.

The main control unit 460 receives various key inputs from the operation panel 70 and reports the set copy mode to each of the control units 410 to 450. Alternatively, the main control unit 460 transmits the set copy mode to each of the control units 410 to 450. In addition to displaying the necessary contents on the display unit of the operation panel 70 in response to the notification, each control unit 410 manages the time of the processing routine.
To 450, and instructs the original discrimination unit 500 of the timing of each operation, controls the whole unity, and realizes a smooth copying operation. When a copy prohibition signal is received from the document discrimination unit 500, the printer control unit 440
To not execute the image forming operation. (3) Configuration of Document Discrimination Unit 500 FIG. 3 is a block diagram showing a configuration of the document discrimination unit 500.

In the block diagram, normal solid arrows indicate the flow of normal data or control signals, thick solid arrows indicate the flow of image data, and thick broken arrows indicate the characteristic amounts of other characters. The flow of data relating to each is shown. The document determination unit 500 performs two processes based on the image data of the document read by the image reading unit 200. That is, a specific original discriminating process of discriminating whether or not the original is a specific original such as a bill, and outputting a copy prohibition signal for prohibiting copying to the main control unit 460 when the specific original is determined,
Document direction discrimination processing (that is, top / bottom recognition processing) for discriminating the direction of the document read by the image reading unit 200 and outputting a rotation angle signal to the memory control unit 430 so that image data is output in an appropriate direction. .

The former specific document discriminating process is executed by the CPU 501 based on the output from the specific color extracting circuit 510.
The latter upside-down recognition processing is performed by the color data canceling circuit 5.
20, character extraction circuit 530, feature amount extraction circuit 54
0, and after passing through the dictionary data generation circuit 560, C is finally determined based on the comparison result performed by the comparison circuit 580.
This is executed by the PU 501.

Hereinafter, the two processes will be described separately. (3-1) Specific Document Discrimination Processing Image read by image reading section 200 and image signal processing section 420
When the image data of each of the R, G, and B color components that have been subjected to the predetermined processing are input to the specific color extraction circuit 510 via the image data bus GB (FIG. 2), the specific color extraction circuit 510
Calculates the color area of a preset specific color based on these image data, and determines whether or not the original is a specific original.

The specific color extraction circuit 510 comprises a color space conversion circuit 511, three specific color counter units 512 to 514, and an OR circuit 515 as shown in the block diagram of FIG. The color space conversion circuit 511 is a circuit that converts input R, G, and B image data values into color signals in a color space including luminance (Y) and two chromaticities (Cr, Cb). Such conversion to a color space is generally a known technique for color reproduction in a television or the like, and a detailed description thereof will be omitted.

If the data values of R, G, and B remain as they are, the bandwidth becomes wide and the data processing becomes complicated, so that the "color" is converted by converting the color values into the Y, Cr, and Cb color signals as described above.
Can be specified by a color signal having a relatively narrow band, and specific color counter units 512 to 514 described below.
To facilitate color judgment. The Y, Cr, and Cb color signals obtained by the color space conversion circuit 511 are input to specific color counter units 512 to 514, respectively.

Of these, the specific color counter unit 512
Are the color determination units 5121 to 5123, the counter 5124 to
5126 and an AND circuit 5127. Each of the color determination units 5121 to 5123
Based on the values of the Y, Cr, and Cb color signals converted in step 1, it is determined whether the pixel is a specified specific color.
Therefore, for example, each of the color determination units 5121 to 5123
Each of the comparators has a plurality of comparators therein. In each of the comparators, a lower limit value and an upper limit value of Y, Cr, and Cb are set for a specific color designated by its own determination unit. If it is determined from the output result that all of the color signal values are within the predetermined range, the color of the pixel is determined to be the specified specific color and counted up to the next-stage counters 5124 to 5126, respectively. Send a signal.

Each of the counters 5124 to 5126 increments the count by "1" each time a count-up signal is obtained from the corresponding color judging section 5121 to 5123. When the count exceeds a predetermined threshold value set by the CPU 501, A color area matching signal is output to the AND circuit 5127. AND circuit 5127
When a color area matching signal is received from any of the counters 5124 to 5126, a logical signal “1” is output and the OR circuit 51
A specific color count-up signal is sent to the CPU 501 via the CPU 5.

Referring back to FIG. 3, when receiving the specific color count-up signal, CPU 501 determines that the document is a specific document and transmits a copy prohibition signal to main control unit 460. Upon receiving the copy prohibition signal, the main control unit 460 refuses to accept the operation from the operation panel 70 and prohibits the printer control unit 440 from performing an image forming operation based on the image data.

As described above, the specific color counter unit 512
Counts the number of pixels (color area) of three specific colors in the read image data of a color original, and generates a specific color count-up signal only when the number of pixels of each specific color is all larger than a predetermined threshold value Accordingly, copying of the original is prohibited. For example, if the copy-protected original is a 10,000-yen bill, three characteristic colors used for this bill are selected, and the color signals Y and Cr are used so that each specific color can be determined. , Cb, the upper limit value and the lower limit value of each color signal value for each of three characteristic colors.
In addition to the setting of the internal comparators 121 to 5123, the pixels having the specific color are counted, and the total value is set as a threshold value in each of the counters 5124 to 5126, and the image data obtained by reading the original is specified. If the number of color pixels is greater than the threshold value for all three selected colors, the original is regarded as a 10,000-yen bill and a copy prohibition signal is generated, thereby preventing the color digital copier from being misused in advance. be able to.

Characteristic colors that are not used in normal printing are used for banknotes, and it is unlikely that all of the three characteristic colors of a document other than a banknote will exceed a predetermined color area. In addition, the above configuration makes it possible to almost certainly determine that the original is a 10,000-yen bill. In order to avoid such a copy prohibition operation, there may be a case where the copy magnification is slightly changed to cause an error in counting the number of pixels of a specific color. In such a case, C
The threshold value given to the counters 5124 to 5126 may be changed by the PU 501 in conjunction with the setting of the copy magnification.

The other specific color counter unit 51
Reference numerals 3 and 514 also have the same configuration as the above-described specific color counter unit 512. The upper and lower limits of the color signal in each color determination unit and the number of pixels in each counter are determined in accordance with the color document to be copy-prohibited. The description is omitted because only the thresholds are set differently. As a result, in addition to the 10,000-yen bill, a 5,000-yen bill and a 1,000-yen bill can be determined, and copying thereof can be prohibited.

According to the specific document discriminating method as described above,
The specific document can be determined very easily and easily. That is, in a conventional copying machine, a pattern of a specific document is extracted, and it is determined whether or not the specific document is a specific document by matching the pattern with a pattern set in advance. When a new banknote is issued and its pattern is changed, it is necessary to store the new banknote pattern in a new manner. However, in the present embodiment, since the determination is made only based on the color areas of a plurality of specific colors, a complicated control program is not required at all, and the simple A specific document can be identified in real time with only a simple hardware circuit.
Only by changing the upper and lower limit values of each color signal and the threshold value of the number of pixels of the corresponding specific color by U501, it is possible to easily cope with not only domestic bills but also foreign bills.

Further, even if there is a slight error in the value of each color signal or a variation in the read value due to the variation peculiar to the apparatus, the CP value is not changed.
By correcting the above values by U501, correction can be easily performed. Further, each specific color counter unit 51
If the number of specific colors that can be determined is increased by increasing the number of color determination units and counters in 2 to 514, the accuracy of determining a specific document can be further improved, and the specific color counter units 512 to 514 themselves can be used. By increasing the number of documents, it is possible to determine more types of specific documents.

It should be noted that the CPU 501 also generates a copy prohibition signal when a character or symbol indicating copy prohibition is found in a document, which will be described later. (3-2) Top / bottom recognition process The top / bottom recognition process is roughly divided into a character cutout process of cutting out a character image from a document image one by one, a feature amount extraction process of extracting a feature amount from the cutout character, A dictionary data generation process of selecting dictionary data of candidate characters for top-bottom recognition from data that does not depend on the direction of the character in the feature amount; and data that depends on the direction of the character in the feature amount and the dictionary data. And a rotation angle determination process for determining a rotation angle by comparison.

[Character Extraction Processing] In FIG. 3, the R, G, B image data read from the image memory 431 is input to a color data cancel circuit 520, where the chromatic color image data is converted into white data. Replacement processing is performed. Specifically, for example, R, G, and B image data are converted using the same color space conversion circuit as the color space conversion circuit 511 (FIG. 4) in the specific color extraction circuit 510.
The color signals are converted into Y, Cr and Cb color signals. Generally, Cr, C
Since the smaller the value of b, the closer to achromatic color, a predetermined threshold value close to “0” is set in advance, and the values of Cr and Cb are compared with the threshold value for each pixel, and both values are smaller than the threshold value. In this case, it is regarded as achromatic image data and the value of Y of the pixel is output as it is, and at least one of Cr and Cb is
If the value is larger than the threshold value, the pixel value is regarded as color image data, and the value of Y of the pixel is replaced with a value corresponding to “white” and output.

Canceling the chromatic color image data in advance as described above is unnecessary data for performing the top-bottom recognition when the read original is a color photographic image. This is because color characters used for headings or the like have been subjected to design processing or are intentionally placed at an angle, and are considered to be unsuitable for recognizing the direction of the characters. By excluding such data having a possibility of erroneous recognition in advance, it is possible to further ensure the subsequent top and bottom recognition processing.

Since the image data (luminance signal Y) output from the color data canceling circuit 520 includes gradations of density, unnecessary binarization information is removed by binarization by the next binarization circuit 521. Makes data processing easier and more reliable. The binarizing circuit 521 binarizes the image data into “0” and “1” based on predetermined parameters given from the CPU 501,
22 and the character extraction circuit 530.

The histogram generation circuit 522 generates a histogram in which the density values of the binarized image data are integrated in the main scanning direction and the sub scanning direction. FIG. 5 is a diagram showing an example of this histogram generation. When the binary image data D1 for one document is integrated in the main scanning direction, the result becomes a first histogram H1, and when integrated in the sub-scanning direction, the result becomes a second histogram H2. These first and second histograms H1 and H2 are stored in the histogram memory 523 as histogram data.

The character extracting circuit 530 extracts character images one by one from the binary image data based on the histogram data. Such character extraction processing itself is a known technique, and the first histogram H1 shown in FIG.
As can be seen from the distribution, the valley portion of the histogram indicates the line spacing, whereby image data is first extracted for each row, and the extracted rows are integrated in the sub-scanning direction to obtain a third histogram. obtain. In the third histogram, the valleys of the integral values occurring at substantially equal intervals should indicate the space between characters, and image data for one character is cut out as a character image based on this position.

Based on the result of the character extracting process, the character extracting circuit 530 sends the CPU 501 an "XY signal" indicating the size (number of pixels) of the extracted character image and a "B / W signal" indicating the black and white ratio. Output. The XY signal is expressed as the product of the number X of pixels in the horizontal direction and the number Y of pixels in the vertical direction of the cut-out character image, and the B / W signal is a pixel (binarized) indicating a line portion in the character image. The number B of "1", which is hereinafter referred to as "black pixel", and the number of pixels other than the line portion (the binary "0"portion; hereinafter, referred to as "white pixel"). It is shown as a value divided by W.

The CPU 501 compares this information with a predetermined threshold value to determine whether the cut-out character is suitable for recognizing the direction of the cut-out character. When there are many characters, a RETRY signal is generated and transmitted to the character cutout circuit 530 to prompt cutout of the next character image. Details will be described later. [Feature Extraction Processing] The character extraction circuit 530 uses the R
Unless an ETRY signal is received, the image data of the cut-out character is output to the feature extraction circuit 540 at the next stage.

Based on the input image data, the characteristic amount extraction circuit 540 generates a plurality of local shapes that characterize the character, that is, a closed loop, a cross intersection, a T-shaped intersection,
This is a circuit that extracts end points (hereinafter simply referred to as “feature shape”), the number of elements, and the like to generate predetermined data. As shown in FIG.
Element number extraction circuit 542, closed loop extraction circuit 543, cross extraction circuit 544, T-shaped extraction circuit 545, end point extraction circuit 5
46, thinning processing circuit 547, error check circuit 54
8, a feature data rearranging circuit 549.

The circumscribed rectangle ratio extraction circuit 541 calculates the vertical / horizontal ratio (X / Y) of the cut-out XY dot character image, and expresses this as 8-bit data as ratio information and outputs it. The number-of-elements extraction circuit 532 determines the number of elements constituting a character (for example, "ku" is one, "i" is two,
"B" is extracted as 4), and the number of elements is represented as 3-bit (up to 8) as number information and output. Such a number of elements can be obtained by a known image processing technique.
For example, a raster scan is performed in the sub-scanning direction with the upper left pixel of the character image as an initial point, and when a black pixel is first found, the pixel of the outline outside the line portion of the character is traced in a predetermined direction using this as a tracking start point. (This processing is hereinafter referred to as “contour line tracking processing”), and when it returns to the black pixel at the tracking start point again, it is determined that the area surrounded by the contour line is one element. Therefore, by repeating such tracking scan, the total number of elements of the character can be extracted.

A dictionary size switching signal from the CPU 501 is supplied to the closed loop extraction circuit 543, the cross extraction circuit 544, the T-shaped extraction circuit 545, and the end point extraction circuit 546 other than those described above, and the extraction circuits 543 to 546 The resolution for extracting the feature amount is switched according to the content of the switching signal. Specifically, the character image cut out by the character cutout circuit 530 is divided into 3 × 3 9 blocks or 5 × 5 25 blocks, and a characteristic shape is extracted from each block. At this time, for example, if the closed loop extraction circuit 543 scans one block and finds at least one closed loop, it does not scan the closed loop in the other block, but proceeds to scan the closed loop in the next block. Therefore, the extraction speed of 3 × 3 is faster than that of 5 × 5, but even if there are two closed loops in one block, only one is counted. The resolution will be lower than in the case. Since the dictionary data stored in the dictionary data generation circuit 560, which will be described later, is also created in accordance with this resolution, the following distinction of the resolution is made based on the number of divided blocks in a 3 × 3 dictionary size, 5 × 5 I will call it the dictionary size.

Based on the specification of the dictionary size from the CPU 501, the closed loop extraction circuit 543 detects the closed loops present in the character image for each divided block, and detects the number of the closed loops (for example, “く” is 0). , "Ha" is 1
And the position of the closed loop is obtained, and the number information and the position information of the closed loop are output.

The extraction of the closed loop can also be performed by a known image processing technique. For example, by extracting the character element by the above-described contour line tracing processing, the inside of the element is scanned, and the white pixel area is extracted. This can be achieved by extracting the number in the same manner as in the case of extracting the number of elements. Then, the number of the extracted closed loops is represented by 3 bits (up to 8) and output as number information.

On the other hand, position information is generated from the relative position of the closed loop in the character image. This position information is generated by determining the presence or absence of a closed loop for each divided block. With reference to FIG. 7, generation of closed-loop position information when the extracted characters are “Δ” and “nu” will be specifically described.

FIG. 7A shows that a character image of “ぱ” is 3 × 3
It is the figure divided | segmented into nine blocks. Bits corresponding to blocks 11 to 33 are prepared, and the position information is set to “1” for the bit corresponding to the block having the closed loop.
, And other bits are set to “0”. If "ぱ", blocks 13 and 3
2, there are closed loops L1 and L2, so that the bit corresponding to the block is set to "1". Even if a closed loop is located in the block 22, the position information does not depend on the rotation of the character, so that it is unnecessary as top-down recognition data.
The position information of the closed loop is indicated by 8 bits corresponding to the number of blocks 8 excluding 2 (similarly, in the case of a 5 × 5 dictionary size, 24 bits are used).

On the other hand, in the case of “ne”, as shown in FIG. 7B, one closed loop L 3 is provided across the blocks 32 and 33. The position information in this case is a closed loop L3
The block containing more white pixels in FIG.
In the example of (b), it is determined that the closed loop L3 is located in the block 32, and the bit corresponding to the block is set to “1”.
Set to generate location information.

In the circumscribed rectangle ratio extraction circuit 541, the element number extraction circuit 542, and the closed loop extraction circuit 543,
Each extraction process was performed without processing the image data output from the character extraction circuit 530. However, the cross extraction circuit 544, the T-character extraction circuit 545, and the end point extraction circuit 5 described below.
In step 46, in order to easily and surely perform the extraction processing, the thinning processing circuit 547 performs a thinning processing for narrowing the width of the line portion (line figure) of the character in advance.

In the thinning process, it is necessary to mainly satisfy the following requirements. (I) The line width is 1 (one pixel unit). (Ii) The position of the line is approximately at the center of the original line figure. (Iii) The connectivity of the figures is preserved. (Iv) The end of the line figure does not shrink more than necessary.

In order to perform such thinning processing, various thinning processing methods have been conventionally proposed in the field of image processing, and a detailed description thereof will be omitted.
The data of the character image thinned by the thinning processing circuit 547 is output to the cross extraction circuit 544, the T-shaped extraction circuit 545, and the end point extraction circuit 546, respectively. Cross extraction circuit 54
In step 4, a cross intersection is detected for each of the divided blocks based on the thinned image data, and the number information and the position information are generated.

The cross intersection is extracted by, for example, scanning a neighboring black pixel (near 8) around a specific black pixel (hereinafter referred to as a “pixel of interest”) and scanning a black pixel adjacent to the pixel of interest. Detect number M. Since the image data is thinned by the thinning processing circuit 547, no black pixels are adjacent to each other in the vicinity of eight points of interest, so when M = 4, the pixel of interest is a cross intersection. Can be determined.

If this scanning is performed for each block, the number of cross intersections in the character image is determined. For example, "ku" has 0, "sa" has 1, and "ki" has 2
Individual. Since the position of the cross intersection changes according to the direction of the character, the position is useful information for top and bottom recognition. This position information is similar to the case of the above-described closed loop position information, and corresponds to the block where the cross intersection exists.
By setting it to “1”, 9 bits for 3 × 3 dictionary size and 25b for 5 × 5 dictionary size
It is represented by it.

The T-shaped extraction circuit 545 detects a T-shaped intersection for each divided block in the thinned image data, and generates information on the number, position, and adjacent direction. This T-shaped intersection is detected when the number of black pixels near the target pixel is M = 3 in the above-mentioned crossed intersection extraction, and by performing this scan for each block, the T-shaped intersection of the character image is detected. The number is determined. For example, "ku" has 0, "to" has 1, and "e" has 2
Individual.

In accordance with this, the position information of the T-shaped intersection is generated. This position information is generated in the same manner as in the case of the cross intersection described above. Since the direction adjacent to the T-shaped intersection also differs depending on the direction of the character, it is useful information for top and bottom recognition. The adjacent direction of this T-shaped intersection is (up / down / left /
There are a total of four types of adjacent directions (right).
The direction adjacent to the T-shaped intersection in is “up”, and these are expressed as a 2-bit signal for the block where the T-shaped intersection exists, and the number of divided blocks × 2 bits
Expressed as a number.

The end point extraction circuit 546 detects the character end points of the thinned image data for each divided block, and generates information on the number, position, and adjacent direction. An end point is detected when the number of black pixels M = 1 in the vicinity of 8 of the pixel of interest in the above-described cross intersection detection,
By performing this detection operation for each block, the number of end points in the character image is determined. For example,
There are one "no", two "ku", and four "e".

In accordance with this, position information of the end point is generated. This position information is, as in the case of the above-described closed loop and cross intersection, b corresponding to the block in which the end point exists.
It is represented by setting it to “1”. The direction adjacent to the end point also differs depending on the direction of the character, and thus is useful information for top-bottom recognition. There are eight types of adjacent directions of the T-shaped intersection at this end point (up / down / left / right / upper left / upper right / lower left / lower right). Since these 8 types of adjacent directions are expressed as 3-bit signals, The direction adjacent to each end point is expressed by the number of divided blocks × 3 bits.

The CPU 501 includes the extraction circuits 543-5
A dictionary size switching signal is transmitted to 46 to instruct to switch the dictionary size and extract each feature amount at a resolution required for top and bottom recognition. The details of this switching operation will be described later. As described above, each of the extraction circuits 541 to 541
The 12 types of information generated in 546 are data independent of the direction of the character, that is, each of the extraction circuits 542 to 546.
And the data depending on the direction of the character, that is, the ratio information from the circumscribed rectangle ratio extraction circuit 541 and the position information and the adjacent direction information from the other extraction circuits 542 to 546 can be divided into two. Hereinafter, the former information will be collectively referred to as “feature amount number information”, and the latter information will be collectively referred to as “feature amount data”.

The above-mentioned feature quantity number information and feature quantity data are input to an error check circuit 548 at the next stage, and the values of the feature quantity number information and the ratio information in the feature quantity data are determined in advance. It is determined whether the threshold value is equal to or less than the threshold value. If at least one of the threshold values exceeds a predetermined threshold value, it is determined that the extracted feature amount is not appropriate for performing the top-bottom recognition.

For example, if the extracted character is the character "purchase", the number of closed loops output from the closed loop extraction circuit 543 is considered to be 9 to 7 in a 5 × 5 dictionary size. Since the dictionary data described later does not store such complicated character data,
The cut-out character is not a character suitable for top and bottom recognition. Therefore, an upper limit value of the number information of closed loops is obtained in advance from the dictionary data, and this is set as a threshold value in a comparator or the like in the error check circuit 548. This is a feature when the input number information of closed loops exceeds the threshold value. An amount extraction error signal is generated and transmitted to the CPU 501.

When the CPU 501 receives the error signal, the CPU 501 executes RETRY to cut out the next character image.
A signal is transmitted to the character extraction circuit 530 to execute a new character extraction, and the information of the current extracted feature amount is discarded. The extraction circuits 541 to 546 described above are used.
, And the feature extraction check in the error check circuit 548 are all performed by a hardware circuit, and the CPU 501 only checks the feature extraction error signal from the error check circuit 548, so that high-speed processing is possible. .

The threshold set in the error check circuit 548 is set smaller in the 3 × 3 dictionary size than in the 5 × 5 dictionary size. This is because, in the case of a 3 × 3 dictionary size, one block size is large, and even if two or more similar types of feature amounts are included in the dictionary size, the number of feature amounts is one maximum per block as described above. This is due to the fact that the count of the feature amount is smaller than in the case of the 5 × 5 dictionary size.

When both the feature quantity information and the ratio information extracted by the extraction circuits 541 to 546 pass the check by the error check circuit 548, the feature quantity information and the feature data are arranged in the next stage of the feature data list. It is sent to the change circuit 549. The characteristic data rearranging circuit 549
The information of these feature amounts is rearranged and output in an order that is easily processed by the dictionary data generation circuit 560 and the comparison circuit 580, respectively.

In particular, the feature amount data is rearranged in the same order as dictionary data output from a dictionary data generation circuit 560 described later, and receives a dictionary data switching signal from the CPU 501 and is output on a common bus in 8-bit units. With such a configuration, a plurality of feature amount data can be effectively handled by a bus having a small bit width, the circuit can be simplified, and the dictionary in the comparison circuit 580 can be obtained. The comparison processing with the data can be facilitated.

[Dictionary Data Generation Processing] Next, the dictionary data generation circuit 560 will be described. Dictionary data generation circuit 5
When receiving the feature quantity number information from the feature quantity extraction circuit 540, the list of characters matching the feature quantity number information among a plurality of characters stored therein is listed as recognition candidate characters. Is output to the comparison circuit 580 in a predetermined order as feature data stored in advance as dictionary data.

As shown in the block diagram of FIG. 8, this dictionary data generation circuit 560
1, RETRY counter 562, dictionary data ROM 56
3, a switching signal counter 564, a dictionary size switching signal is supplied to the dictionary address generating circuit 561 and the dictionary data ROM 563, and a feature data switching signal is supplied to the switching signal counter 564 by the CPU 501.

In the dictionary data ROM 563, 256 characters or symbols for the top and bottom recognition shown in the following (Table 1) (hereinafter, simply referred to as “reference characters”), 3 × 3.
Data (dictionary data) corresponding to the above-described feature amount data is previously obtained and stored for each of two types of 5 × 5 dictionary sizes.

[0083]

[Table 1]

On the other hand, the dictionary address generation circuit 561 has the feature quantity number information (total 15 bits) and the RETRY counter 5
3 and the dictionary size switching signal (1
R) of 256 Kbytes that outputs the address of the corresponding dictionary data in the dictionary data ROM 563 as 8-bit data, with a total of 19 bits as an address.
OM table (hereinafter "address generation ROM table")
That. ).

The procedure for creating the address generation ROM table will be described specifically below. That is, first, for each of the 256 reference characters, the number information is obtained in advance for each dictionary size of 3 × 3, 5 × 5 by the above-described procedure, and the 256 characters are classified for each combination of the number. FIG. 9 shows the number of the four characteristic shapes (closed loop, cross intersection, T-shaped intersection, character end point) and the number of elements of 46 characters of the hiragana character among the reference characters of the above (Table 1) in a 5 × 5 dictionary size. Listed in the search.

From this table, for example, (the number of closed loops, the number of crosses, the number of T-shaped intersections, the number of characters, the number of elements) = (0,
It can be seen that the hiragana corresponding to the combination of the number information of 0, 0, 4, 2) includes four characters of “i”, “u”, “ko”, and “ri”. Therefore, for the combination of the number information, the four characters in Hiragana will
At 80, the list is listed as recognition candidate characters to be compared, and stored in the address generation ROM table in association with the feature quantity number information. Similarly, other reference characters are classified for each combination of the number information, and stored in the address generation ROM table.

In this embodiment, a maximum of eight recognition candidate characters can be stored in the ROM table for the same feature quantity number information.
When the number N of recognition candidate characters corresponding to the input feature amount information is found from the OM table, the number is transmitted to the CPU 501 as a dictionary number signal (3 bits).

The CPU 501 uses the dictionary number signal to
Number N of recognition candidate characters corresponding to predetermined feature quantity number information
Is transmitted, a dictionary RETRY signal is transmitted to the RETRY counter 562, and the count value is updated to the maximum N, thereby selecting one of the plurality of recognition candidate characters one by one. Therefore, for example, “ko” of the recognition candidate character is (closed loop number, cross intersection number, T-shaped intersection number, character fraction, element number) = (0, 0, 0, 4, 2)
And the count number (3bi) from the RETRY counter 562 indicating the third.
t).

Actually, each character is stored in the address generation ROM table in a character code represented by 8 bits. When "ko" is selected as a recognition candidate character as described above, the character code is stored in the dictionary. The address is sent to the dictionary data ROM 563 as an address. Dictionary data ROM563
Is a dictionary address (8 bits), a dictionary size switching signal (1 bit), and a switching signal counter (5bi).
t) is the address, and the dictionary data of the corresponding character is 8b
It is configured as a 16-Kbyte ROM table (hereinafter, referred to as a “dictionary generation ROM table”) that outputs each it.

FIG. 10 is a diagram showing the structure of dictionary data stored for each character code in the dictionary generation ROM table.
The dictionary size and the right side store dictionary data of 5 × 5 dictionary size, and ratio information, closed loop information and feature point data information are stored in units of 8 bits for each address.

In the data column of FIG.
Represents the aspect ratio of a rectangle circumscribing the character in 8 bits. “Closed loop information” is information corresponding to the above-described closed loop position information. This information is obtained by a method similar to that described with reference to FIG. 7. For example, the position of the closed loop of Hiragana in a 3 × 3 dictionary size is represented by a circle in FIG.
It becomes the position marked with a bit, and the bit corresponding to the block
Is set to “1”. According to this, 9 bits are required for the 3 × 3 dictionary size and 25 bits are required for the 5 × 5 dictionary size, but the presence or absence of the central closed loop does not affect the direction of the character as described above. Therefore, it is unnecessary data for top and bottom recognition, reducing 1 bit each,
In the case of a 3 × 3 dictionary size, 8 bits (1 Byte
e) In the case of a 5 × 5 dictionary size, 24 bits (3 By
te) data, the former in one address,
The latter is stored separately in three addresses (closed loop information 1, 2, 3).

"Feature point data" is a general term for position information or adjacent direction information for three types of feature points, such as a cross intersection, a T-shaped intersection, and an end point. The adjacent direction is expressed as 8-bit data for each corresponding block. FIG. 12 is a diagram illustrating a storage state of the feature point data in a 3 × 3 dictionary size in a bit plane.

As shown in the figure, each block (MAT3_
Eight types of bit planes indicating the presence or absence of each feature point or information on adjacent directions in 11 to 33) are prepared from the front to the back, and the data of these bit planes for each block are collected in 8 bits (1 byte). )
Feature point data MAT3_1-11-33 (left side in FIG. 10)
It is treated as. Therefore, in the case of a 3 × 3 dictionary size, the data is divided into nine blocks, and this feature point data is represented as 9 bytes of data for one character.

The creation of such feature point data is basically performed in the same manner as that described in the case of extracting the feature amount data. First, each reference character in (Table 1) is converted to the reference character in FIG. As shown in an example, it is divided into 3 × 3 9 blocks. The cross position information is the MAT3_11 of the bit plane.
B of the block where the cross intersection of the character exists
It is set to “1”, and the bits of the other blocks are set to “0”. Similarly, in the T-shaped position information and the end point position information, the bit of the block where the feature amount is located is set to “1”, and the bits of the other blocks are set to “0”.

The T-character adjacent direction information is such that the T-character adjacent direction is indicated by 2-bit data in the block where the T-character intersection exists. That is, the adjacent direction is
“00”, “01”, “1” according to (up / left / right / down)
0 "and" 11 ", and therefore, two bit planes (T-shaped adjacent direction information bits 1 and 2) are prepared.

The end point adjacent direction information indicates the adjacent direction in the block where the end point exists by 3-bit data. That is, the adjacent directions are “000”, “001”, “01” according to (up / down / left / right / upper left / upper right / lower left / lower right).
0 "," 011 "," 100 "," 101 "," 11 "
0 "and" 111 ". Therefore, three bit planes (end point adjacent direction information bits 1, 2, and 3) are prepared.

FIG. 14 is a diagram showing bit planes of feature point data in the case of a dictionary size of 5.times.5. As shown in an example of FIG. 15, each reference character in (Table 1) is divided into 25 blocks. , Bits are allocated to each block, and feature point data is created in the same manner as in the case of the above 3 × 3 dictionary size. The feature point data in each block is 8 bits
(1 Bytes), the feature point data is 1
Each character is represented as 1 × 25 = 25 Byte data.

The above dictionary data is created in advance for all 256 reference characters shown in (Table 1), and stored in the dictionary generation ROM table of the dictionary data ROM 563 (FIG. 8) for each character code. The data required according to the dictionary address (character code) output from the dictionary address generation circuit 561 and the reception state of the dictionary size switching signal and the characteristic data switching signal from the CPU 501 are represented by the address number xx01 (3) in FIG. Eight bits are output to the comparison circuit 580 (FIG. 3) in order of their address numbers from the × 3 dictionary size) or × 21 (5 × 5 dictionary size).

As described above, the feature data switching signal from the CPU 501 is transmitted to the switching signal counter 564.
In addition, the feature data rearranging circuit 549 of the feature amount extracting circuit 540 (FIG. 6) is also provided. The feature data rearranging circuit 549 converts the feature amount data extracted from the extracted characters according to the dictionary size. In this case, the data is rearranged in the same order as the storage state of FIG. 10 described above, and every time the characteristic data switching signal is received, the characteristic amount data after the rearrangement processing is output to the comparison circuit 580 in units of 8 bits. As described above, the data can be efficiently transmitted through a bus having a small bit width as described in the feature amount extraction circuit 540, and the comparison processing in the comparison circuit 580 becomes easy, and the hardware circuit thereof can be simplified. There is.

[Rotation Angle Determination Processing] The comparison circuit 580 converts the dictionary data output from the dictionary data generation circuit 560 into 9
The data rotated by 0 ° is compared with the feature data output from the feature extraction circuit 540 to generate information on the degree of coincidence between the feature data and the dictionary data at each rotation angle. In comparison, if the cut-out character is recognized as a specific character meaning copy prohibition, copy prohibition information is generated.

FIG. 16 is a block diagram showing the structure of the comparison circuit 580. As shown in FIG.
Are selectors 58 provided at the input unit and the output unit, respectively.
1, 582, a specific character dictionary 583, a rotation angle evaluation circuit 584 and a specific character evaluation circuit 585 corresponding to a 3 × 3 dictionary size, and a rotation angle evaluation circuit 586 and a specific character evaluation circuit corresponding to a 5 × 5 dictionary size 587.

The specific character dictionary 583 stores dictionary data related to specific characters included in a normal copy-inhibited document, for example, so-called “secret” characters in which “secret” characters are circled. In order to distinguish the dictionary data from the top / bottom recognition dictionary data output from the data generation circuit 560, hereinafter referred to as “forbidden dictionary data”), the dictionary sizes 3 × 3 and 5 × 5 are stored in advance and given to the selector 581. Can be

When the same kind of feature amount data and dictionary data are input from the feature amount extraction circuit 540 and the dictionary data generation circuit 560 to the selector 581, respectively,
Switches the destination of the feature amount data and the dictionary data according to the contents of the dictionary size switching signal sent from the CPU 501. Hereinafter, 3 × from the CPU 501
A description will be given mainly of a case where a switching signal of three dictionary sizes is transmitted. However, the same processing is performed for a case of a 5 × 5 dictionary size.

Upon receiving the 3 × 3 dictionary switching signal, the selector 581 converts the feature data into the rotation angle evaluation circuit 5
The dictionary data is sent to the rotation angle evaluation circuit 584 and the specific character evaluation circuit 585. On the other hand, the dictionary size switching signal is also given to the specific document dictionary 583, whereby the 3 × 3 dictionary size prohibited dictionary data is selected, and the selector 581 sends this to the specific character evaluation circuit 585.

As shown in the block diagram of FIG. 17, the rotation angle evaluation circuit 584 includes a closed loop evaluation circuit 590, an aspect ratio evaluation circuit 591, a feature point evaluation circuit 592, and each evaluation circuit 5
The output values from 90 to 592 have predetermined coefficients K1 to K1 respectively.
Weighting circuits 593 to 59 for weighting by multiplying by K3
5, an addition circuit 596 for adding the output value from the weighting circuit for each rotation angle, and a maximum value extraction circuit 597 for extracting the maximum value as a result of the weighting. Compare the quantity data with the dictionary data to determine the degree of coincidence, and extract the highest one,
This is converted into a percentage for the case of complete coincidence to obtain coincidence information, and the rotation angle at that time is output to the CPU 501 as rotation angle information.

That is, the feature data and the dictionary data are
The contents are input to the closed loop evaluation circuit 590, the aspect ratio evaluation circuit 591 and the feature point evaluation circuit 592, and each evaluation circuit 590 to 592 converts the input dictionary data to 90 °.
It has a rotation processing circuit for rotating the data in units, and calculates the degree of coincidence by comparing the contents of the dictionary data and the feature data rotated by each rotation angle for each divided block.

The degree of coincidence calculated here is a negative form of exclusive OR of the coincidence of the set state of the corresponding bit (except for the coincidence, except for the aspect ratio evaluation circuit 591, which is set to “1” only when the bit coincides). It is obtained by adding and using the evaluation points obtained based on the judgment. For example, the closed loop evaluation circuit 590 determines whether or not a closed loop exists in one of the central blocks among 9 bits (25 bits in a 5 × 5 dictionary size) corresponding to the number of divided blocks as described above.
The other 8 bits (except 24 bits) excluding the bits are compared with the dictionary data for each corresponding block, and all b bits are compared.
If it matches, 8 points (24 points) are given.

Since the vertical / horizontal ratio information is irrelevant to the number of divided blocks, the vertical / horizontal ratio evaluating circuit 591 evaluates the degree of coincidence of the ratio information as follows. That is, the ratio A / B of the aspect ratio A of the dictionary data and the aspect ratio B of the feature data is determined, and the higher the degree of coincidence, that is, the closer the A / B is to 1, the higher the score. For example, | A / B-1 | <
At 0.1, 8 points, O.D. 1 ≦ | A / B-1 | <
0 to 8 points are given according to the magnitude of the error between the A / B value and 1, such as 7 points when 0.2.

In this case, 0 ° and 1
The evaluation point of 80 ° matches the evaluation point of 90 ° and 270 °, and since this evaluation point is obtained irrespective of the number of block divisions as described above, it naturally depends on the dictionary size. do not do. The feature point evaluation circuit 592 compares feature point data (see FIG. 10) regarding three feature points (cross intersection, T-shaped intersection, and end point) in the divided blocks.

As shown in FIG. 12 (FIG. 14), each feature point data expresses eight types of bit plane information in 8 bits for each block, and the degree of coincidence with dictionary data is as follows. Required. That is, as for the cross position information, if the presence or absence in the block matches the dictionary data, one point is given as it is. Regarding the position information of the T-shaped intersection and the end point, even if the presence or absence of the T-shaped intersection or the end point in the block matches the dictionary data, if “Yes”, the information of the adjacent direction is further compared, and One point is given only when the adjacent information also matches. In the case of a match with "absence", one point is given as it is without comparing adjacent information.

Therefore, if the information in one block matches the information in all blocks, 9 points (25 points in the case of a 5 × 5 dictionary size) are given, and all three types of feature points completely match. In this case, 27 points (75 points) are given as the highest points. The above evaluation processing is performed by the closed loop evaluation circuit 590, the aspect ratio evaluation circuit 591 and the feature point evaluation circuit 592.
, The corresponding dictionary data is 0 ゜, 90 ゜, 180
{270}, the evaluation points at each rotation angle are output to the next-stage weighting circuits 593 to 594, and the weighting coefficients K1 and K are assigned to the respective evaluation points.
2, and K3 are multiplied and output to the addition circuit 596.

Here, the reason why the evaluation points of each feature data are weighted by weighting is that the result of the evaluation obtained for each feature data does not necessarily affect the recognition in the character direction (ie, top and bottom recognition) equally. In other words, depending on the contents of the document, it may be considered that changing the specific gravity of the evaluation may allow more accurate top-bottom recognition. For example, if the manuscript is an ordinary Japanese sentence, the hiragana accounts for 60 to 70% of the whole manuscript and plays a very important role in recognizing the top and bottom. In this case, the closed loop containing many hiragana is used. If the evaluation point is increased and its influence is increased, the top and bottom recognition can be performed more effectively than by referring to the evaluation points in other aspect ratios. Therefore, the coefficient K2 in the weighting circuit 594 is
Coefficients K1, K in other weighting circuits 593, 594
By making it larger than 3, the accuracy of the top and bottom recognition can be improved.

When the original contains many letters of the alphabet, it is considered that the evaluation point of the feature point data has a large influence for the top-bottom recognition. Therefore, the value of the coefficient K3 is changed to other coefficients K1 and K2. What is necessary is just to make it larger. The values of these coefficients K1 to K3 can be obtained experimentally or empirically, and they are
1 is stored in an internal memory, and a switch for inputting the type of the document is provided on the operation panel 70 as needed, for example. What is necessary is just to make it set to 593-595.

Thus, each of the weighting circuits 593-5
The evaluation points for which the specific gravity of the evaluation has been changed by 95 are added by the number of the evaluation items for each rotation angle in the addition circuit 596 and output to the maximum value extraction circuit 597 in the next stage. The maximum value extraction circuit 597 extracts the maximum one of the addition points at each rotation angle, converts the addition point into a percentage with respect to the maximum point in the case of complete coincidence to obtain “matching degree information”, The angle at that time is transmitted to the CPU 501 as “rotation angle information”.

Returning to FIG. 16, the specific character evaluation circuit 585 compares the feature data and the prohibited dictionary data for each rotation angle. The configuration of the specific character evaluation circuit 585 is basically the same as the operation of the rotation angle evaluation circuit 584. The evaluation point is calculated by comparing the feature data and the prohibited dictionary data at each rotation angle. If the maximum evaluation point exceeds a predetermined threshold set internally,
It is determined that the character is a copy prohibited character such as “confidential”, and a copy prohibited signal is transmitted to the CPU 501 via the selector 582.

When CPU 501 receives copy prohibition information from comparison circuit 580, CPU 501 transmits a copy prohibition signal to main control section 460 (FIG. 2),
Further includes a memory control unit 430 and a printer control unit 440
Of the original, so as not to execute the copy operation of the original. Usually, the copy prohibition phrase is often described at the top of the document, but it may be described at the bottom of the document, or even if the document is upside down by chance even at the top of the document, The above-described specific character evaluation process needs to be performed for all the characters of one page of the document. For this reason, it may be considered that the evaluation time of a specific character becomes longer and the copy operation is delayed. For example, a lockable switch box is provided inside the apparatus, and a mode changeover switch is provided inside the apparatus, or a security code is provided from the operation panel. It is convenient if only a specific person can switch to a mode in which recognition of the specific character is not performed as necessary by inputting.

On the other hand, the CPU 501 compares the value of the coincidence information output from the comparison circuit 580 with a predetermined threshold value set therein, and when the value of the coincidence information is larger than the threshold value, Then, it is determined that the extracted character and the reference character match, and a signal indicating the rotation angle of the rotation angle information is transmitted to the memory control unit 430 as a rotation angle signal.

The memory control section 430 rotates the image data of the document by the rotation angle designated by the CPU 501. Then, even if the time required for evaluating a specific character for one page of the document has elapsed, the CPU 501 causes the comparison circuit 580 to operate.
If the copy prohibition information is not received from the printer and the specific color count-up signal is not received from the specific color extraction circuit 510, it is determined that the original is not a copy-inhibited original, and the rotation processing is performed. The output image data is output to the printer control unit 440, whereby the original image is reproduced on the recording sheet in the correct direction.

If the value of the degree of coincidence information is lower than the threshold value set therein, the dictionary size is switched and re-evaluated, or the next character is cut out, and the feature amount extraction circuit 540, The processing in the dictionary data generation circuit 560 and the comparison circuit 580 is repeated until the top and bottom recognition is completed.
Details will be described later. (4) Operation in Control Unit 400 FIG. 18 is a flowchart showing a main routine of a control operation executed in control unit 400.

When the apparatus is turned on, first, each CP
Initial settings of registers inside U are performed (steps S1, S2). At this time, the dictionary size switching signal is also set to the 3 × 3 dictionary size as an initial value. After that, the fixing unit 32 not directly involved in the copy operation as a waiting process
7 (FIG. 1), the display control of the operation panel, and the like are performed (step S3). When the start key of the operation panel 70 is pressed and a copy start instruction is issued (step S3).
4) The originals are sequentially conveyed to the original glass plate 21 by the automatic original conveyance unit 100, scanned by the scanner 22, the read image data is subjected to image processing by the image signal processing unit 420, and one page of the original is stored in the image memory 431. An image input process for writing each time is performed (step S5).

Then, image data for one page of a predetermined document is read from the image memory 431, and the document discrimination section 500
In step S6, the orientation of the character is recognized, and the top and bottom of the document is recognized (step S6). As a result of the top and bottom recognition, if the image data of the document needs to be rotated in an appropriate direction, the rotation processing unit 432 Then, a rotation process is performed by a necessary rotation angle (step S7).

Next, it is determined whether or not the original is copy-prohibited (step S8). That is, as described above, the document discrimination unit 500 determines that the document is a banknote or a copy-prohibited document that is subject to copy prohibition,
When the main control unit 460 receives the copy prohibition signal, the main control unit 460 sends an instruction not to perform the image forming process to the printer control unit 440. If not, the image forming process is executed (step S9). The process returns to the waiting process 3.

FIGS. 19 and 20 are flow charts showing a subroutine of the top-bottom recognition process in step S6. First, in step S601 in FIG.
501 is a character extraction circuit 53 which generates a RETRY signal.
0 is instructed to cut out a character image from the image data in the image memory 431.

The character extracting circuit 530 extracts a character image for one character based on this signal as described above,
The character size indicated by the product of the number of horizontal pixels X and the number of vertical pixels Y is transmitted to the CPU 501 as an XY signal and the ratio of black pixels to white pixels as a B / W signal. CPU5
First, the XY signal is used to determine whether or not the cut-out character is a large-size character of a predetermined size or more (step S602). The large-sized character referred to here has a vertical and horizontal character size of, for example, 10 mm or more (400 d
pi indicates that X or Y exceeds 157 dots).

The character size is determined as described above.
Such large-sized characters are rarely used as characters used in ordinary manuscripts, and even if large-sized characters are used as headlines in newspapers etc., emphasized characters and inclined characters are considerably deformed. This is because the possibility is high and it is considered that the character is not suitable as a character to be subjected to top-bottom recognition.

Next, the CPU 501 determines from the B / W signal whether the ratio B / W of black pixels to white pixels in the cut-out character image is not a predetermined value, for example, 1 or more (step S603). ). When the B / W ratio is 1 or more, that is, when the number of black pixels is 50% or more of the total pixels, the character is considered to be an extremely deformed bold or inverted character, and this is also recognized as upside down. Is not appropriate as a target character, and it is desirable to eliminate it in advance.

If the character size or the B / W ratio exceeds the respective predetermined values as a result of the above-described determination, the CPU 501 regards that a character cutout error has occurred (Yes in step S604), and returns to step S604. A cutout circuit 530 issues a RETRY signal to cause the next character to be cut out. On the other hand, if it is not determined in step S604 that there is a character cutout error, the cutout image data is sent to the feature amount extraction circuit 540, and the CPU 501 processes the image data in a 3 × 3 dictionary size. The dictionary size switching signal is transmitted to each unit of the feature amount extraction circuit 540, the dictionary data generation circuit 560, and the comparison circuit 580 (step S605).

Then, the feature amount of the character is extracted by each of the extraction circuits 541 to 546 (FIG. 6) in the above-described feature amount extraction circuit 540 (step S606). It is determined whether the number information of each of the number, the closed loop, the cross intersection, the T-shaped intersection, and the end point is equal to or larger than the threshold set for the 3 × 3 dictionary size (step S607).

If at least one piece of information exceeds the threshold value corresponding to the information, it is determined in step S608 that a feature quantity extraction error has occurred, and a signal (feature quantity extraction error signal) is sent to the CPU 501. It is determined whether or not the current dictionary size is 3 × 3 (step S611).
Since the size is set to the dictionary size, 5
In order to execute more accurate extraction of the feature amount in the × 5 dictionary size, a dictionary switching signal to the 5 × 5 dictionary size is transmitted to the feature amount extraction circuit 540 and the like (step S612).

As a result, the feature quantity extraction circuit 540 has a 5 ×
After the feature amount is extracted again with five dictionary sizes, an error check is performed on the number information of each feature amount (steps S606 and S607). If it is still determined that there is a feature amount extraction error (step S608) Ye
s) Then, the process proceeds to step S611. At this time, since the size of the dictionary is already 5 × 5, it is determined to be “No” and the process returns to step S601 to instruct the extraction of the next character by the RETRY signal. The above-described processing of step S602 and subsequent steps is performed on the cut-out character.

FIG. 21 is a diagram showing the relationship between the error check of the feature amount extraction in step S608 and the dictionary switching operation (steps S611 and S612). In the figure, the number of feature values increases as going upwards. However, as described above, the aspect ratio information and the number of elements do not depend on the dictionary size. Except for the four feature shapes (closed loop, cross intersection, T
Character intersections, end points).

First, feature quantities are extracted in the 3 × 3 dictionary size. If the number of feature quantities at this time exceeds the feature quantity extraction error value (threshold) in the 3 × 3 dictionary size, then 5 The size is switched to the dictionary size of × 5, and the feature amount is extracted. As described above, each piece of information excluding the aspect ratio information and the number of elements counts only one even if the same block includes a plurality of the same feature points and the like, so the number of divided blocks is increased to increase the resolution. The more accurate the number information can be obtained.

Therefore, usually, processing is performed with a 3 × 3 dictionary size that enables high-speed comparison processing, and processing is performed with a 5 × 5 dictionary size that can perform accurate comparison processing on complex characters having many feature values. If a feature extraction error occurs, the extracted character is determined to be a character that is not included in the reference characters such as kanji with a large number of strokes, so erroneous determination is prevented by switching the extracted character itself. is there.

FIG. 22 is a table showing feature value extraction error values in each feature value for each dictionary size. For each error value, the maximum value of each piece of feature quantity information obtained for each of the 3 × 3 dictionary and the 5 × 5 dictionary for the reference characters shown in Table 1 above is set. As for the feature shape, the higher the resolution, the greater the number of pieces of information to be extracted, so that the error value is also set higher.

It is to be noted that the error value of the number-of-elements information and the number-of-endpoints information in the table is large because the above-mentioned “confidential” character is compared with the 256 reference characters built in the dictionary data generation circuit 560. The number of elements and the number of end points are large, and unless the error value is increased in this way, the error check circuit 548 of the feature extraction circuit 540 before the feature data of the character is input to the comparison circuit 580. This is because it is excluded by the check in FIG. The numerical value in parentheses in the error value indicates an error value to be adopted when the copy prohibition mode based on the determination of the secret document is released.

Returning to FIG. 19, if it is determined in step S608 that there is no feature amount extraction error, the dictionary address generation circuit 561 of the dictionary data generation circuit 560 (FIG. 8) based on the characteristic quantity number information of the cut-out character. , A recognition candidate character is selected from the reference characters, and the number thereof is transmitted to the CPU 501 as a dictionary number signal N. CPU 501
Confirms this dictionary number signal N (step S609),
The dictionary RETRY signal is transmitted from the dictionary data generation circuit 560 to instruct to output the dictionary data.

Next, the rotation angle evaluation circuit 5 of the comparison circuit 580
Weighting coefficients K1 to K3 in each of the weighting circuits 84 and 586 (see FIG. 17) are set (step S61).
0). This weighting factor is, for example,
Each type of Japanese or English sentence used in the document is obtained in advance and stored in the CPU 501. When the operator inputs the type of the document from the operation panel, the CPU
501.

After such preparation is completed, the flow shifts to step S613 in FIG. 20, and the dictionary data is read. This is because the dictionary data generation circuit 560 uses the CPU 501.
, A RETRY counter 562 (FIG. 8) in the dictionary data generating circuit 560 counts up each time the dictionary RETRY signal is received, and generates a dictionary address. The signal is transmitted to the circuit 561. The dictionary address generation circuit 561 selects a character corresponding to the count value from the N recognition candidate characters, and transmits the character code to the dictionary data ROM 563 as a dictionary address. The dictionary data ROM 563 reads out the dictionary data as described in FIG. 10 based on the dictionary address and the dictionary size switching signal from the CPU 501, and counts up by the switching signal counter 564 which counts up by receiving the characteristic data switching signal from the CPU 501. , The dictionary data of the corresponding address is stored in 8 bits (1B
yte) to the comparison circuit 580 at a time.

At this time, the feature data switching signal from the CPU 501 is 9 times when the dictionary size is 3 × 3 because the feature point data is 9 bytes, and when the dictionary size is 5 × 5, the feature point data is 25 bytes. Since it is output 25 times, considering only the feature point data, when the dictionary size is 5 × 5, it is about 2.
This would require five times the processing time.

The comparison circuit 580 includes a feature amount extraction circuit 540
Is compared with the contents of the dictionary data from the dictionary data generation circuit 560 for each rotation angle as described above, and the respective evaluation points are calculated to obtain the degree of coincidence. Is converted to a percentage with respect to the case where it matches, and this is transmitted as matching degree information to the CPU 501 together with information about the rotation angle (step S614).

The CPU 501 compares the value of the coincidence information with a preset threshold TH1 (for example, 95%).
(Step S615), and when the value of the matching degree information exceeds the threshold value TH1, it is considered that the cut character and the character of the dictionary data completely match at the rotation angle, and the rotation at that time is determined. The angle of the angle information is transmitted as a rotation angle signal to the memory control unit 430 (FIG. 2), and the process returns (step S619).

On the other hand, in step S615, the value indicated by the coincidence information transmitted from comparison circuit 580 is equal to threshold value T.
If the value is equal to or less than H1, the dictionary data of another recognition candidate character is read and compared with the feature amount data, and the obtained coincidence information is compared with each other to extract the maximum value (step S61).
6). Therefore, in step S617, it is determined whether or not the comparison has been made for all of the dictionary numbers N of the recognition candidate characters.
A TRY signal is transmitted, whereby the dictionary data generating circuit 560 switches and outputs dictionary data (step S6).
13) If the coincidence information newly obtained by the comparison circuit 580 based on the dictionary data exceeds the above TH1, the rotation angle is used as a rotation certain signal as the memory control unit 430.
Otherwise, the process returns to step S616 to extract the maximum value of the matching degree information.

The extraction processing in step S616 is repeated up to (N-1) times until comparison with the dictionary data of all N recognition candidate characters (step S617). If any of the characters exceeds TH1, the rotation angle information at that time is used as a rotation angle signal (No in step S615, S619).
If none exceeds H1 (step S61)
(Yes in 7), there is no help for it, and the maximum matching degree information extracted in step S616 is compared with a threshold TH2 (for example, about 90%) smaller than the threshold TH1 (step S6).
18) If the matching degree information is equal to or greater than the threshold value TH2, the rotation angle at that time is used as a rotation angle signal (step S).
619).

If the maximum value of the coincidence information is smaller than the threshold value TH2 in step S618, the process proceeds to step S620 to check whether the current dictionary size is 3 × 3. If the dictionary size is 3 × 3, the process proceeds to step S612 in FIG.
And the resolution is increased, and the processing of steps S606 to S619 is executed again with this dictionary size.

If it is determined in step S620 in FIG. 20 that the dictionary size has already been changed to 5 × 5, it is determined that the cut-out character is inappropriate for the top-bottom recognition, and the process proceeds to step S620 in FIG. Returning to step S601, the cut-out character is switched, and the above-described operation of the top and bottom recognition processing is repeated. As described above, in the present embodiment, dictionary data relating to the 256 reference characters to be recognized is stored in advance, and the characters extracted by the character extraction circuit 530 are stored in the feature extraction circuit 540. Generates feature quantity number information that does not depend on the direction of the character and feature quantity data that depends on the direction of the letter. First, recognition candidate characters are selected from the 256 characters based on the feature quantity number information. Since a limited number of dictionary data is compared with the cut-out character feature data, the number of comparison processes is extremely reduced, and the processing time for top-bottom recognition can be greatly reduced. In this case, the direction of the character can be determined by comparing the matching degree information once.

In addition, since the processes such as the extraction of the feature amount and the comparison with the dictionary data can all be achieved by a hardware circuit, it is not necessary to execute a complicated pattern recognition algorithm, and the process can be performed more quickly. Also, regarding the extracted characters, for the characters for which it is difficult to determine the upside-down recognition in advance, the above-described character extraction error determination (FIG. 1)
9, step S604) and feature amount extraction error determination (step S608 in the figure), so that comparison is made for characters different from the dictionary data to prevent erroneous recognition when the comparison result is good by chance. Becomes possible. (5) Modifications While the embodiments of the image input device according to the present invention have been described above, the present invention is not limited to the above-described embodiments, and the following modifications are considered. be able to.

(5-1) In the above embodiment, a closed loop, a cross intersection,
Position information or adjacent direction information of a character-specific local shape such as a T-shaped intersection or an end point was extracted and used for determining the character direction. It goes without saying that the greater the amount of such information, the more certainty of top and bottom recognition increases, but depending on the character to be recognized, it is also possible to determine the direction based on some of the information, and not necessarily all of the characters. No information is needed.

On the other hand, in the above embodiment, when the matching degree information of one cut-out character is larger than the threshold value TH1, the rotation angle is immediately regarded as the document direction and the rotation processing of the image data is performed. However, if another cut-out character is also checked and the rotation angle is recognized as the document direction only when the result matches, the accuracy of the top-bottom recognition increases.

(5-2) In the above embodiment, both the feature quantity number information independent of the character direction and the feature quantity data dependent on the character direction are extracted from the cut-out character image. First, the recognition candidate characters to be compared are listed up based on the feature quantity number information, and the dictionary data of the recognition candidate characters and the feature data are compared in order. However, in some cases, all dictionary data and feature amount data may be compared in order without performing a process of listing recognition candidate characters based on feature amount number information. In this case, although the processing time is slightly longer than in the above-described embodiment, quicker and more accurate top-and-bottom recognition can be performed as compared with the conventional processing using pattern matching.

(5-3) In the above embodiment, the chromatic image data is detected by the color data cancel circuit 520, and the chromatic image data is replaced by white data. Although the cutout process is performed, the character cutout process may be performed first with the color image data unchanged, and thereafter, a chromatic character may be detected, and the feature amount may be extracted based on other black and white characters excluding this. . In this case, a histogram is generated even for unnecessary chromatic image data, and character extraction processing is performed. Therefore, it is considered that each processing circuit is complicated and processing time is slightly longer. The recognition accuracy in the document direction based on the data can be maintained.

(5-4) The rotation angle evaluation circuit 584 (5
86), the dictionary data is rotated by 90 ° and compared with the feature data. However, it is needless to say that the feature data may be rotated and compared. In this case, the direction in which the image data is rotated by the rotation processing unit 432 is reversed. In addition, most of ordinary documents have a portrait document in which characters are printed in horizontal writing.
Omitting the comparison at the rotation angle of ゜ would not be too inconvenient. Conversely, by making the width of the rotation angle of 90 ° smaller and making a comparison while rotating the document little by little, the document is moved to the document reading position of the document glass plate 21 due to the inconvenience of the conveyance belt 14 of the automatic document conveyance device 100. Even when the image data is installed obliquely, the image data can be rotated and output in a correct direction.

(5-5) In the above embodiment, an example in which the image input apparatus according to the present invention is applied to a copying machine has been described. It is also applied as a device.

[0153]

As described above, according to the image input apparatus of the present invention, the chromatic image data in the color image data is detected by the color data detecting means.
Since the direction of the document is determined based on the character image in the other black-and-white image data excluding the detected chromatic image data, the chromatic character image that is not suitable for character direction recognition is excluded in advance. It is possible to reliably recognize the direction of the document based on a character image consisting of only black and white image data that has almost no possibility of being decorated or deformed.

Further, according to the present invention, the color data detecting means converts the color image data into a color space represented by luminance and two kinds of chromaticity, and the two kinds of color space data. If any one of the chromaticity values exceeds a threshold value set for each of the chromaticity, the chromaticity determination unit determines that the image data is chromatic color image data. Can be detected.

Further, according to the present invention, the original direction discriminating means generates monochrome image data by replacing the detected chromatic image data with white data, and cuts out a character image from the monochrome image data. Then, the relative position of the local shape in the character image or the adjacent direction is extracted as a character feature amount, and this is compared with the character feature amount of a predetermined character stored in advance to determine the direction of the document. Since the chromatic image data is replaced in advance with white data, the subsequent character segmentation can be easily performed.In addition, the recognition of the character direction is not based on the pattern recognition but on the relative local shape of the character image. Since only the comparison of the target position or the adjacent direction is performed, no complicated algorithm for pattern recognition is required at all, and quick processing by a simple hardware circuit is performed. It is possible.

[Brief description of the drawings]

FIG. 1 is a diagram showing an overall configuration of a copying machine to which an image input device according to the present invention is applied.

FIG. 2 is a block diagram showing a configuration of a control unit in the copying machine.

FIG. 3 is a block diagram showing a configuration of a document discrimination unit in the control unit.

FIG. 4 is a block diagram illustrating a configuration of a specific color extraction circuit of the document discrimination unit.

FIG. 5 is a diagram for explaining the contents of a histogram generated by a histogram generation circuit of the document determination unit.

FIG. 6 is a block diagram illustrating a configuration of a feature amount extraction circuit of the document determination unit.

FIG. 7 is a diagram for explaining contents of closed loop position information extracted by a closed loop extraction circuit in the feature amount extraction circuit.

FIG. 8 is a block diagram showing a configuration of a dictionary data generation circuit in the document discrimination section.

FIG. 9 is a diagram showing the value of each feature quantity number information in a 5 × 5 dictionary size for 46 hiragana characters among reference characters.

FIG. 10 is a dictionary data ROM of the dictionary data generation circuit.
FIG. 4 is a diagram showing the contents of dictionary data for each reference character in FIG.

FIG. 11 is a diagram for explaining generation of closed loop position information stored in the dictionary data in the case of a 3 × 3 dictionary size.

FIG. 12 is a diagram schematically showing a storage state of feature point data of a 3 × 3 dictionary size in the dictionary data by a bit plane.

FIG. 13 is a diagram for explaining a case in which a hiragana character is divided into nine blocks to obtain position information of each feature point in a 3 × 3 dictionary size.

FIG. 14 is a diagram schematically showing a storage state of feature point data of 5 × 5 dictionary size in the dictionary data by a bit plane.

FIG. 15 is a diagram for describing a case in which a hiragana character is divided into 25 blocks to obtain position information of each feature point in a dictionary size of 5 × 5.

FIG. 16 is a block diagram illustrating a configuration of a comparison circuit in the control unit.

FIG. 17 is a block diagram showing a configuration of a rotation angle evaluation circuit in the comparison circuit.

FIG. 18 is a flowchart showing a main routine of a control operation in the control unit.

FIG. 19 is a flowchart illustrating a subroutine of a top-bottom recognition process in step S6 of FIG. 18;

FIG. 20 is a flowchart continued from FIG. 19;

FIG. 21 is a diagram showing the relationship between dictionary size switching and the number of feature quantities.

FIG. 22 is a diagram illustrating an example of an extraction error value of the number of feature values in each dictionary size.

FIG. 23 is a diagram for explaining a conventional character direction recognition method.

[Explanation of symbols]

 REFERENCE SIGNS LIST 100 automatic document feeder 200 image reader 300 printer 400 controller 410 image reader controller 420 image signal processor 430 memory controller 431 image memory 432 rotation processor 440 printer controller 450 external communication controller 460 main controller 500 Document discriminating unit 501 CPU 510 Specific color extraction circuit 520 Color data cancellation circuit 530 Character extraction circuit 532 Element number extraction circuit 540 Feature quantity extraction circuit 560 Dictionary data generation circuit 580 Comparison circuit

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Toru Iida 2-3-1-13 Azuchicho, Chuo-ku, Osaka-shi, Osaka Inside Osaka International Building Minolta Co., Ltd. (72) Tomoya Yoshimura, Azuchi-cho, Chuo-ku, Osaka-shi, Osaka 2-3-13 Osaka International Building Minolta Co., Ltd.

Claims (3)

[Claims] 1. An apparatus for inputting an image of a color document, comprising: image reading means for reading a document to generate color image data; image data storage means for storing the color image data; Color data detection means for detecting chromatic image data in the data, and document direction discrimination means for discriminating the direction of the document based on a character image in black and white image data other than the detected chromatic color image data And an image data rotating means for rotating the color image data in accordance with the determined document direction. 2. The color data detecting unit, wherein the color image detecting unit converts the color image data into a color space represented by luminance and two types of chromaticity, and one of the two types of chromaticity. 2. The image input device according to claim 1, further comprising: a chromatic color determining unit that determines that the image data is chromatic image data when the value of the image data exceeds a threshold value set for each of the image data. 3. The document direction discriminating means includes: color data replacing means for generating black and white image data by replacing the detected chromatic image data with white data; and a character for extracting a character image from the black and white image data. A cutout unit; a feature amount extracting unit that extracts a relative position of a local shape in the character image or an adjacent direction thereof as a character feature amount; a feature amount storage unit that stores a character feature amount of a predetermined character in advance; Comparison means for comparing the character feature quantity extracted by the feature quantity extraction means with the character feature quantity stored in the feature quantity storage means; and recognition means for recognizing the direction of the character image based on the comparison result. 3. The image input device according to claim 2, wherein the direction of the character image recognized by the recognition unit is determined as the direction of the document.