JP3053857B2 - Image reading processor - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Purpose of the Invention] (Industrial application field) The present invention is intended to create a photo card (or booklet) such as a driver's license, an identification card, an introductory card, and a bank passbook. The present invention relates to an image reading processing device that reads an image on an application form and performs a predetermined process.

(Prior Art) Generally, a driver's license or other certificate is affixed with a photo covered with a transparent sticker to identify the holder of the certificate, and the name, date of birth, gender, The issue date and issue number are printed.

Conventionally, the creation of such a certificate with a photograph is performed in the following procedure, taking a driver's license as an example. First, the applicant fills in a required form on a prescribed form, pastes a photograph, etc., and prepares an application form, and submits it to the center window. In the process of creating a certificate, first, the applicant's name, date of birth, gender, issuance date, and issuance number are typed on the mount of the certificate based on the submitted application form. Next, a separately prepared photograph is pasted on the mount. Finally, a transparent seal is adhered on the attached photograph to coat. Since each of these operations is performed in an independent process, it is inefficient, and even in a computer processing center, it takes about several hours from receiving an application form to issuing a certificate. In addition, in the past, there are many cases where the photo to be attached to the certificate and the photo to be attached to the application form at the time of application are different from each other, and a plurality of leaves must be prepared.

On the other hand, a more improved driver license issuance / renewal system is being considered. In this system, the photographing fields (name, date of birth, gender, issue date, issue number, etc. entered in a prescribed layout) provided on the application form are optically combined with the applicant's face image. And shoot on silver halide photographic film. Next, this film is developed, and the obtained photographic paper is cut into an appropriate size, and then subjected to lamination processing to complete a driving license.

Further, Japanese Patent Application Laid-Open No. 1-20609 describes a system for electronically synthesizing and laying out an image signal obtained by photographing the face of an applicant with a video camera, and printing the processing result by a printer. Have been.

In these known systems, since a series of license issuance operations includes a step of photographing the applicant's face, the time required for issuance cannot be sufficiently reduced due to the photographing time and the waiting time for the photographing time. In addition, a photographing space and a space for waiting for a turn are required for photographing a face. Furthermore, since the photographing of the face is generally performed once in terms of efficiency and cost, an image of the face with an expression that the applicant does not like may be displayed on the license. In such a case, the applicant will feel mental distress during the validity of the license for several years.

(Problems to be Solved by the Invention) As described above, the conventional system for issuing a certificate with a photograph such as a license is inefficient because it requires a plurality of steps, and is attached to the application form. There is a problem that it is necessary to prepare a photograph separately from the photograph or to secure a space for photographing the applicant's face, and it is not possible to shorten the time required for issuance due to the photographing process.

SUMMARY OF THE INVENTION An object of the present invention is to provide an image reading processing apparatus capable of easily and efficiently issuing a certificate with a photograph such as a driver's license from an application form in which necessary items are filled in and a photograph of a face is stuck. .

[Structure of the Invention] (Means for Solving the Problems) In order to solve the above problems, the image reading processing apparatus of the present invention converts an image in which each area of a color image, a black-and-white image, and a character image to be recognized is mixed. An image reading means for reading and outputting an image signal; an area dividing means for dividing the image signal output from the image reading means for each area; and an image signal of each area divided by the area dividing means in a predetermined recording format. And image processing means for performing processing so as to conform to the above.

In another aspect of the present invention, an image reading unit reads an image in which at least two types of image areas are mixed among a color image, a black and white image, and a character image to be recognized, and separately from the image reading unit. And image input means for inputting an image signal of at least one of these images. The image processing means performs a predetermined image processing on the image signal output from the image reading means and divided for each area and the image signal input by the image input means.

(Operation) In the device of the present invention, for example, a color image of a face photograph on an application created by an applicant applying for the issuance of a certificate such as a driver's license, a black and white image such as a signature, an address,
Character images to be subjected to character recognition, such as handwritten characters such as name, date of birth, and gender, or printed characters are collectively read. The image signal obtained by this reading is divided for each area and then subjected to image processing. The image processing is performed on the image signal of each area according to the recording format of the issued certificate, for example, the layout and size of each area. The processed image signal is sent to, for example, a printing unit, or once stored and sent to a printing unit, where a certificate of a predetermined recording format is printed and issued.

Apart from the image signals obtained by reading the images on the application form, for example, when image signals obtained by reading images recorded on various recording media are input, these image signals Is processed.

In this case, the image signal obtained by reading the face photograph on the application form or the image signal of the applicant's face image input separately is processed and printed on the certificate. Not only is the step of photographing the face image unnecessary, but also all the information to be printed on the certificate is processed simultaneously and collectively, so that the time required for issuing the certificate is greatly reduced. In addition, a face image selected by the applicant himself is printed on the certificate.

(Example) Hereinafter, an example of the present invention is described with reference to drawings.

FIG. 1 is a block diagram showing one embodiment in which the image reading processing apparatus of the present invention is applied to a certificate issuing system with a photograph. This system creates and issues a certificate 200 such as a driver's license from an application form 100, and includes a main control unit 101, a keyboard 102, and a display unit 10 configured using a CPU.
3, image reading unit 104, character recognition unit 105, input image processing unit 1
06, image processing control unit 107, output image processing unit 108, printing unit 10
9, a large-capacity storage unit 110, an image input unit 111, an image interface (IF) 112, and a communication interface 113.
The communication interface 113 is connected to a center host computer 115 via a communication line 114, for example. In FIG. 1, thick lines represent multi-bit data lines, and thin lines represent control lines.

As shown in FIG. 2, the application form 100 includes a face photograph area 20 for attaching a monochrome or color photograph of the applicant's face.
1. Signature area 202 for the applicant to sign (sign) and handwritten character area 20 for the applicant to write necessary information
Three types of image areas are mixed. A reference line 204 is provided in the signature area 202. Printed reference marks 205 and 206 for giving reference coordinates for performing frame detection described below are printed on the left end of the application form 100 in the figure. The mark 205 indicates the reference coordinates for detecting the frame of the photograph area 201, and the mark 206 indicates the reference coordinates for detecting the frame of the signature area 202. At the upper end in the drawing of the application form 100, there is formed a mark 207 indicating the presence or absence of printing and a mark 208 indicating the presence or absence of a notch for identifying the form of the application form. In the example shown in FIG. 2, each of the marks 207 and 208 has three unit mark areas, and therefore, each of them can express eight types of application forms by digital codes.

Depending on the purpose of the application (new issue, reissue, correction / addition, registration cancellation, urgent issue, etc.), the layout may differ,
Furthermore, not only the layout, but also the printing colors (characters subject to drop-out color processing in the handwritten character area 203) indicating character frames, etc., are changed so that different types of applications are devised so that there is no erroneous procedure in the applicant and counter operations. A book style is prepared in advance. The application form 100 is selected from the plurality of types according to the purpose of application. Is processing of the application form 100 performed collectively for each application content,
Alternatively, application forms of various application contents are mixed and continuously made. In any of the processing modes, it is possible to respond flexibly by identifying the format of the application form 100.
The identification of the form of the application form 100 may be performed by an operator to input and instruct with the keyboard 102, or may be automatically performed from the marks 207 and 208 provided on the application form 100.
The former is effective when applications of the same format are processed to some extent collectively, but the automatic identification of the latter according to the present embodiment is effective when applications of various formats are mixed.

The schematic operation of the system shown in FIG. 1 will be described. An application form 100 shown in FIG. 2 completed and prepared by the applicant is submitted. After receiving this application form 100, the image reading unit 1
When set to 04, the form of the application form 100 is
The application form identification code, which will be described later, is provided therein, and the application form identification code is sent to the main control unit 101. The type of application identification code that defines the identified application form is arbitrarily determined in system design. Thereafter, the image on the application form 100 is read by the image reading unit 104, and an image signal is output. This image signal is input to the input image processing unit 106. The input image processing unit 106 divides the input color image signal into each of the regions 201, 202, and 203 in FIG. 2 under the control of the image processing control unit 107, and performs a predetermined process on the image signal in each region.

The processing result of the input image processing unit 106 is
Is transferred to the main control unit 101 via In the main control unit 101, of the processing results from the input image processing unit 106, the image data in the handwritten character area 203 in FIG.
Transfer to 05. The character recognizing unit 105 performs character recognition by a two-stage process using a known composite similarity method and a feature structure matching method, and transfers the recognition result to the main control unit 101 as character code data. In the main control unit 101, a character image specified by the character code data from the character recognition unit 105,
The display unit 103 displays the face photo image and the signature image in the face photo area 201 and the signature area 202 in FIG.

The operator compares the display content of the display unit 103 with the content of the application form 100 and confirms it. If the display content of the character image is different from the content in the handwritten character area 203 of the application form 100, the operator determines that there is a reading error in the image reading unit 104 or that there is a recognition error in the character recognition unit 105. Then, a command to make a correction using the keyboard 102 or to read again in the image reading unit 104 is input. When the certificate 200 is issued by re-issuance or renewal, in addition to the face photograph image and the signature image in the face photograph area 201 and the signature area 202, the certificate 200 is registered in the host computer 115 and the large-capacity storage unit 110 as a database. The displayed content is displayed on the display unit 103. The operator confirms the contents of the application from the displayed contents and performs an appropriate change operation.
At this time, the operator may communicate with the host computer 115 on-line as necessary via the main control unit 101, perform various confirmation operations including a credit check, and receive an issue number. After that, the operator once inputs the work classifications such as new issue, reissue, correction / addition, deletion of registration, urgent issue, etc., to the main control unit 101 via the keyboard 102 to inform the host computer 115, and continue the work. Receive necessary instructions. If the registration is deleted, the work ends here.

The image data that has been checked and corrected in this way is temporarily transferred from the main control unit 101 to the large-capacity storage unit 110, stored, and then transferred to the output image processing unit 108 via the image processing control unit 107. Here, for the image data stored in the large-capacity storage unit 110, a preset recording format is set for the work category recognized from the application form identification code. In the output image processing unit 108, the large-capacity storage unit 110
The image data transferred via the main control unit 101 and the image processing control unit 107 is subjected to processing for matching the recording format in the certificate 200 and the characteristic matching with the printing unit 109. In parallel with the transfer to the large capacity storage unit 110,
It may be transferred to the main image processing unit 108. The printing unit 109 receives the processing result of the output image processing unit 108, performs printing on predetermined base paper, and creates a certificate 200. Certificate 2 created
00 is passed to the applicant.

 Next, details of each unit in FIG. 1 will be described.

FIG. 3 shows the configuration of the image reading unit 104. The image reading unit 104 is for reading an image of each of the areas 202, 202, and 203 on the application form 100, and a sufficient reading resolution for a photograph of the applicant's face on the photograph area 201 where the highest resolution is required, and It is configured to be able to read not only monochrome photographs but also color photographs. In FIG. 3, the application form 100 is transported by the transport mechanism 301 at a constant speed in the arrow Y direction (sub-scanning direction). This transport operation corresponds to sub-scanning. The surface of the application form 100 is illuminated by a light source 303 driven by a light source driving circuit 302, and an image on the surface is formed on a color image sensor 305 by an imaging lens 304. The color image sensor 305 reads an image on the application form 100 at a density of, for example, 16 dots / mm.

The color image sensor 305 includes three CCD line image sensors 306R each including a photoelectric conversion element array to which R (red), G (green), and B (blue) color filters are applied, and a transfer CCD (charge coupled device). , 306G, 306B. The photoelectric conversion element arrays in each of the line image sensors 306R, 306G, and 306B are arranged in a line in a direction perpendicular to the plane of the drawing (main scanning direction). Alternatively, a CCD color image sensor in which R, G, and B three-color filters are formed in a dot-sequential manner may be used as the color imaging element 305. Color image sensor 30
Reference numeral 5 denotes an electric signal which is driven by a CCD driving circuit 307 and is separated into R, G, and B components according to an image formed by electric scanning (main scanning) in the arrangement direction of the photoelectric conversion element array. That is, a line-sequential RGB color image signal is output. The color image signal is converted into a digital signal of, for example, about 8 bits by the A / D converter 309 and then input to the image correction circuit 310.

The image correction circuit 310 is a shading (SHD) correction circuit 31
1. It comprises a space correction circuit 312 and a rearrangement circuit 313. The shading correction circuit 311 uses the light source 303
The level correction and normalization of the image signal are performed for the above uneven illumination, uneven light amount of the imaging optical system including the imaging lens 304, and uneven sensitivity of the color image sensor 305. Spatial correction circuit 312
Corrects a spatial shift between the CCD line image sensors 306R, 306G, and 306B in the main scanning direction. Rearrangement circuit 31
3 is the dot-sequential RGB color image signal
Rearrange to color image signal. Each of these correction circuits 311,
Since the configurations of 312 and 313 are all publicly known, detailed description is omitted here. The configuration of the image correction circuit 310 is appropriately changed according to the configuration of the color imaging element 305.

The application form identifier 314 identifies the format of the application form 100 by reading the marks 207 and 208 in FIG. This application form discriminator 314 is composed of, for example, six non-contact optical sensors installed in positions corresponding to a total of six unit mark areas of the mark 207 and the mark 208, and an electronic circuit for logically processing the output of these sensors. Is done. Each sensor detects the status of each unit mark area (whether there is printing or notch) using the difference in the amount of reflected light between the background and the mark, and generates a logical level signal according to the state of the unit mark area. I do. By performing logical processing (basically logical product) on the output signals of the three sensors corresponding to the mark 207 and the output signals of the three sensors corresponding to the mark 208, the application form An identification code is generated. This application form identification code is supplied to the main control unit 101 in FIG. Mark 207,208
Can increase the number of application forms that can be identified. Mark if the number of identifiable styles is small
Only one of 207 and 208 may be used.

The timing generator 315 operates with the basic clock supplied from the clock generator 316 under the control of the main controller 101 in FIG. 1, and controls the light source driving circuit 302 and the CCD driving circuit 30.
7, A / D converter 309, image correction circuit 310 and application form discriminator
Various timing signals required for 314 are generated.

In this manner, the image reading unit 104 reads the image of the entire required area on the application form 100 at a density of 16 dots / mm by one image reading operation including the main scanning and the sub-scanning.
A dot-sequential RGB color image signal with 2 ⁸ = 256 tones is output. This color image signal is supplied to the input image processing unit 10 shown in FIG.
Supplied to 6. The input image processing unit 106 performs the above-described processing for confirming and correcting the operator, whether or not permission to issue from the host computer 115, and designation of various management numbers such as an issue number, for an input color image signal. Do.

As shown in FIG. 4, the input image processing unit 106 includes a frame detection circuit 401, a region division circuit 402, and line density conversion circuits 403, 404, and 40.
5. A monochrome processing circuit 406, a density detection circuit 407, a color determination circuit 408, a dropout color processing circuit 409, a quantization circuit 410, and a buffer memory 411. The frame detection circuit 401 detects the range (frame) of each area necessary for the area division by the area division circuit 402 and generates a frame signal. The region dividing circuit 402 determines each region of the application form 100 according to the above frame signal.
The image signals 201, 202, and 203 are divided and transmitted on different signal lines. The region dividing circuit 402 can be realized by a switching circuit that performs a switching operation using a frame signal as a control input.

FIG. 5 shows the configuration of the frame detection circuit 401. Frame detection circuit 401
The pixel control unit receives a dot-sequential RGB image signal from the image reading unit 104, a pixel clock GCLK synchronized with pixel data of the RGB image signal from the main control unit 101 via the image processing control unit 107, and an image reading unit. Synchronization signals HSYNC and PSYNC synchronized with the main scan and sub-scan of the unit 104, respectively, and the application form identification code are input. The synchronization signal PSYNC is generated based on a leading end detection signal of the application form 100 generated from an application form sensor (not shown) provided in the image reading section 104, and is used as a processing start signal of the input image processing section 106. The XY reference counter 501 includes an X counter for obtaining a coordinate value in the main scanning direction by counting the pixel clock GCLK using the main scanning synchronization signal HSYNC as a count start signal, and a main scanning synchronization signal using the sub-scanning synchronization signal PSYNC as a count start signal. It is constituted by a Y counter for obtaining a coordinate value in the sub-scanning direction by counting HSYNC, and generates a coordinate value corresponding to each pixel data of the input RGB image signal. The data of the X and Y coordinate values are respectively represented by, for example, 12-bit digital data.

The mark cutout frame generation circuit 502 knows the form of the application form from the application form identification code, and outputs a frame signal for cutting out the signal from the reference marks 205 and 206 in FIG. , And is supplied to the trimming gate circuit 503 as a gate signal. The mark cutout frame generating circuit 502 is constituted by a PLD (programmable logic device, for example, GAL16V8 manufactured by Lattice) or a ROM (read only memory), and a mark cutout frame set in advance corresponding to various forms of the application 100 is a PLD. Or programmed in ROM. The position and size of the mark extraction frame
With sufficient consideration of the positional reading accuracy of the image reading unit 104, particularly the transfer accuracy of the transfer mechanism 301 including skew, the setting is made such that only the marks 205 and 206 are surely included inside this frame. Specifically, the size of the mark cutout frame is, for example, approximately 8 mm square in actual size on the application form 100, which is equivalent to 128 dots × 128 dots of the color image signal output from the image reading unit 104.

The trimming gate circuit 503 receives such a mark cutout frame signal as a gate signal and gates the RGB image signal. The gated RGB image signal is converted by the monochrome processing circuit 504 into a monochrome image signal. The monochrome image signal is averaged by, for example, a 2 × 2 matrix by an averaging circuit 505 to remove noise and isolated point components, and then a binary signal is used by a binarization circuit 506 using a predetermined threshold. After that, it is input to the mark detection circuit 507. The mark detection circuit 507 determines the coordinate values of the marks 205 and 206 from the input binary signal, and supplies the coordinate values to the latch / comparator 508 as frame reference coordinate values. The latch / comparator 508 latches the frame reference coordinate value, compares this with the coordinate value sequentially updated from the XY reference counter 501, and supplies a start signal to the XY frame counter 509 when they match. The XY frame counter 509 counts the pixel clock GCLK and the main scanning synchronization signal HSYNC from the time when the start signal is input. That is, from the XY frame counter 509,
The read coordinate value based on the frame reference coordinate value is output.

The read coordinate value from the XY frame counter 509 is input to the signature frame generation circuit 511 and the photo frame generation circuit 512.
Each of the signature frame generation circuit 511 and the photo frame generation circuit 512 knows the form of the application form 100 by the application form identification code,
The signature frame signal and the photo frame signal corresponding to the format are sequentially generated according to the read coordinate values from the XY frame counter 509. Famous frame generation circuit 511 and photo frame generation circuit 512
Is composed of, for example, PLD or ROM, and the application form 100
Of the signature area 202 and the photo area 201 having different positions and sizes corresponding to the various styles (signature frame and photo frame)
Is previously programmed in the PLD or written in the ROM.

The handwritten character frame generation circuit 513 basically regards a portion other than the region surrounded by the signature frame and the photo frame in the entire region of the application form 100 as the handwritten character region 203, and outputs a handwritten character frame signal indicating the frame of the region. Occurs as The handwritten character frame generation circuit 513 is similarly constituted by a PLD or a ROM, and data of frames of all regions, which are different for each format of the application form 100, are previously programmed or stored. The program or written data is read out according to the application identification code, and data of a frame other than the region surrounded by the signature frame and the photo frame is generated as a character frame signal. To this end, frame signals from the signature frame generating circuit 511 and the photo frame generating circuit 512 are given to the handwritten character frame generating circuit 513.

Frame signals from the signature frame generation circuit 511, the photo frame generation circuit 512, and the handwritten character frame generation circuit 513 are input to the decoder 514, and are converted into 2-bit code signals for distinguishing these three types of regions. Since the RGB image signals from the image reading unit 104 are time-series signals, image signals of two or more regions are not included at the same time. Therefore, the decoder 51
The code signal output from 4 only needs to indicate to which region the RGB image signal of the pixel input at that time belongs.

As described above, the frame detection circuit 401 combines the trimming using the fixed mark extraction frame set by the mark extraction frame generation circuit 502 with the detection of the frame detection marks 205 and 206 from the trimmed RGB image signal. And each area 2
The frame reference coordinate values for generating the frame signals of 01, 202 and 203 are obtained. Prior to mark detection, the averaging circuit 505 performs noise and isolated point removal processing. By these processes, there is an advantage that a mark is detected that is strong against disturbances such as skew and noise when the image on the application form 100 is read by the image reading unit 104 and influence of dirt on the application form 100.
The frame signal output from the frame detection circuit 401 is also supplied to the image processing control unit 107 in addition to the region division circuit 402 in FIG.

The mark detection circuit 507 in FIG. 5 is configured, for example, as shown in FIG. The binary signal from the binarization circuit 505 is serially input to a 512-bit shift register 600 configured by cascading eight 64-bit shift registers 601 to 608. The output of the last stage of each of the 64-bit shift registers 601 to 608 is input to an AND circuit 609, the output of the AND circuit 609 is serially input to an 8-bit shift register 610, and the 8-bit output of the shift register 610 is ANDed in parallel. Input to the circuit 611. When the output of the last stage of all the 64-bit shift registers 601 to 608 becomes “H”, the output of the AND circuit 609 becomes “H”. AND circuit 609
When the state of the "H" level output continues for eight consecutive times, the output of all the bits of the 8-bit shift register 610 becomes "H" simultaneously.
As a result, the output of the AND circuit 611 becomes "H", and the mark detection signal is output.

That is, the mark detection circuit 507 has a window of 8 dots × 8 dots of the input binary signal. The input signal of the mark detection circuit 507 is a signal obtained by binarizing a monochrome image signal (luminance signal) having a dot density of 8 dots / mm by averaging 2 × 2 in the averaging circuit 505. the window is the 1mm ^□ in the actual dimension on the application form 100. If all the binary signals in this window are “H” (black level), a mark detection signal is output assuming that the reference mark 205 or 206 is in the window.

Returning to FIG. 4, the area 2 from the area dividing circuit 402
The image signals of 01, 202 and 203 are converted into linear density conversion circuits 403, 404 and 405.
As a result, linear density conversion, that is, enlargement / reduction processing, is performed at a conversion magnification suitable for each area. The conversion magnification for the image signal in the photographic area 201 and the signature area 202 is calculated in the
The sizes of the upper areas 201 and 202, the reading dot density of the image reading unit 104 (16 dots / mm in this example), and the certificate 200
It is determined by the above recording format (such as the size of each area).
The conversion magnification for the image signal of the handwritten character area 203 is determined by the read dot density requested by the character recognition unit 105. In order to support a plurality of recording formats, a plurality of conversion magnifications are required.

Image recording density of printing unit 109 is 12 dots / mm, application form 100
It is assumed that the size is the same as that of the corresponding area of the photograph area 201 and the signature area 203 and the certificate 200, and that the read dot density required by the character recognition unit 105 is 8 dots / mm.
In this case, the conversion magnification for the image signal in the photo area 201 and the signature area 202 is 12/16 = 3/4, and the conversion magnification for the image signal in the handwritten character area 203 is 8/16 = 1/2. Simultaneously realizing such two or more types of conversion magnifications in real-time processing is difficult in general line density conversion processing performed in combination with mechanical sub-scanning. For this reason, in the present embodiment, the line density conversion is performed by electronic processing by the line density conversion circuits 403, 404, and 405 in both the main scanning direction and the sub-scanning direction, as described later.

FIG. 7 is a diagram showing a configuration of a line density conversion circuit for performing two-dimensional conversion in such a main scanning direction and a sub-scanning direction, and includes a selector 701, two line memories 702, 703,
Memory control unit 704, latches 705,706, coefficient control unit 707, coefficient multipliers 708,709,714,715, buffer 710, adders 711,71
It consists of six and two three-stage latches 712,713. Coefficient multiplier
Reference numerals 708, 709, 714, and 715 each include a coefficient ROM storing coefficient data and a multiplier for multiplying an input signal by a coefficient. The principle of this line density conversion circuit will be described with reference to FIG.

The original image signal input to the line density conversion circuit is X, and the sample interval Px is 1. Consider that the density of the original image signal X in the main scanning direction is 0.75 times, that is, the sample interval Py of the image signal Y subjected to the linear density conversion is 0.75Px. It is assumed that the j-th (j ≧ 0) pixel of the image signal Y subjected to the linear density conversion represents the information of the position of 0.75j of the original image signal X. Position is below 0.75 J, and the value X _i of the sample to the nearest original image signal to 0.75 J, respectively the value of the signal X _{i + 1} of the right side (i + 1-0.75j), becomes (0.75 J-i) the results of combined addition by a coefficient, respectively, and the value Y _j of the j-th pixel of the line density converted image signal Y. For example, in FIG. 8, the value Y ₀ of the 0th pixel of the linearly converted image signal is
It is the same as the value of the first pixel of the original image signal. The value Y ₁ of the first pixel of the line density converted image signal, 0.25 times the value X ₀ of 0 pixel of the original image signal, a value obtained by adding 0.75 times the value X ₁ of the first pixel. When the operation of such a linear density conversion process is generalized and expressed, Y _j = (1−a) X _i + aX _{i + 1} a = frac (j / R). Here, frac represents a decimal part of the numerical value of (j / R), and R represents a magnification of linear density conversion. In the above description of the principle, the line density conversion in the main scanning direction has been described. However, the line density conversion in the sub-scanning direction can be performed according to the same principle.

The linear density conversion circuit of FIG. 7 is configured based on such a principle. The input image signal is supplied to the selector 701.
The line memory 70 alternately by one line in the main scanning direction.
2, 703, and are written to the line memories 702, 703 under the control of the memory control unit 704. Memory control unit 704
Is supplied with the pixel clock GCLK from the main control 101 via the image processing control unit 107. Image signals read one by one from the line memories 702 and 703 under the control of the memory control unit 704 are input to coefficient multipliers 708 and 709 via latches 705 and 706, respectively, and are multiplied by coefficients. The image signal weighted by the coefficient multipliers 708 and 709 is added to the adder 711.
, The linear density conversion in the sub-scanning direction is performed. The output of the latch 705 is input to the first three-stage latch 712 via the buffer 710, and the output of the adder 711 is the second three-stage latch 712.
Input to the stage latch 713. Buffer for coefficient multiplier 714
The output of either 710 or adder 711 is input,
The output of one of the three-stage latches 712 and 713 is input to the coefficient multiplier 715. The image signals weighted by being multiplied by the coefficients by the coefficient multipliers 714 and 715 are added by the adder 716, so that the linear density conversion in the main scanning direction is performed. The coefficient data stored in the coefficient ROMs in the coefficient multipliers 708, 709, 714, and 715 are read at a predetermined repetition cycle in synchronization with the pixel clock GCLK by the ROM address data from the coefficient control unit 707. The coefficient control unit 707 stores various ROMs in advance.
It has address data and information on the above-mentioned repetition period, and these are selected according to the application form identification code input from the main control unit 101 via the image processing control unit 107 and the frame signal. As a result, for example, as described above, the conversion magnification for the image signal of the photograph area 202 and the signature area 202 is increased.
3/4, the conversion magnification for the image signal of the handwritten character area 203 is
It is controlled to be 1/2.

Returning to FIG. The image signal of the signature area 202 converted by the linear density conversion circuit 403 is converted by the monochrome processing circuit 406 from an RGB image signal to a monochrome image signal. No matter what color ink is used to sign on the signature area 202 in the application form 100, the signature area on the certificate 200 is printed with a predetermined single color (for example, black or red) ink. Therefore, since it is useless to process the image signal of the signature area 202 as an RGB image signal, the monochrome processing circuit 406 converts the image signal into a monochrome image signal and compresses the information amount.

FIG. 9 shows the configuration of the monochrome processing circuit 406. The dot-sequential RGB image signal from the linear density conversion circuit 403 is latched by a latch 801,
Input to 802,803. The clock control 804 controls the transfer clock TCLK (3 pixel clock GCLK) from the image processing control unit 107.
The clock of the double cycle is phase-shifted to generate a three-phase clock, and the clock of each phase is supplied to the latches 801, 802, 803. As a result, for example, the R signal
The signal is latched by the latch 802, and the G signal is latched by the latch 803. The R signal and the B signal output from the latches 801 and 802 are added by an adder 805 and further halved by an arithmetic unit 806. The output of the arithmetic unit 806 and the G signal output from the latch 803 are added by the adder 807 and further halved by the arithmetic unit 808. The output of the arithmetic unit 808 is represented by 0.5G + 0.25R + 0.25B, and becomes a monochrome image signal. The reason why the weight of the G signal is increased is that the visibility of human eyes is stronger than green. Since the signals handled by the monochrome processing circuit 506 are digital signals, the operation units 805 and 807 which perform the operation of 1/2
Can actually be realized by a bit shift operation, and does not require a special circuit. Note that the monochrome processing circuit 504 in FIG. 5 may have the same configuration as that in FIG.

The signature area 202 thus converted into a monochrome image signal
Are stored in the buffer memory 411 and input to the density detection circuit 407. The density detection circuit 407 detects the density of the signature image on the signature area 202. This density detection is a process of making the maximum density of the signature image printed on the certificate 200 almost constant while preserving the gradation, irrespective of the density of the signature image on the signature area 202, and making the density easy to read. It is necessary when performing.

As shown in FIG. 10, for example, the density detection circuit 407 includes a latch 901, a timing generation circuit 902, RAMs (random access memories) 903 and 904, counters 905 and 906, and buffers 907 and 90.
8 increment counter 909, buffers 910 and 91
2 and an address control circuit 911. FIG. 11 shows a time chart of signals of respective parts in FIG.

The monochrome image signal from the monochrome processing circuit 406 is
The data is latched in the latch 901 by the transfer clock TCLK from the image processing control unit 107. The latched monochrome image signal is provided to RAMs 903 and 904 as address data RAM-ADR, and data written at an address designated by RAM-ADR is read from RAMs 903 and 904. The read data RAM-DAT from the RAMs 903 and 904 is
2 are preset in the counters 905 and 906 by a preset clock PRCLK generated in synchronization with the transfer clock TCLK. After this preset, the contents of the counters 905 and 906 are incremented by one with an increment clock INCLK. The output CNT-Q of the counters 905 and 906 is
Entered into 7,908. Thereafter, the enable pulse BUF-EN is supplied from the timing generation circuit 902 to the buffers 907 and 908, and the write pulse RAM-WT is supplied to the RAM 903 and 904, so that the contents of the counters 905 and 906 are changed to the RAM 903 and 904.
904 is written to the above address.

Hereinafter, the same operation is repeated. Since the addresses of the RAMs 903 and 904 are given by the monochrome image signal, they correspond to the level (luminance value) of the monochrome image signal. Also RA
The data stored in each address of M903 and 904 indicates a value obtained by accumulating the number of appearances of the luminance value corresponding to the address by the increment counter 909. The level of the monochrome image signal corresponds to the density of the signature area 202. Therefore, in the RAMs 903 and 904, the signature area as shown in FIG.
A density histogram of the 202 image is created. When the generation of the density histogram is completed, the address data is supplied from the address control circuit 911 to the RAMs 903 and 904 via the buffer 912 under the control of the image processing control unit 107, so that the contents of the RAMs 903 and 904 are read. The data is transferred to the image processing control unit 107 via the buffer 910. After this, RAM903,
904 is cleared.

The image processing control unit 107 executes software density determination processing of the image of the signature area 202 from the density histogram obtained by the density detection circuit 407. In the density determination process, for example, a peak P of the appearance frequency in the density histogram of FIG. 12 is detected, and an eight-level density determination is set by dividing between the density level at this peak P and the zero level. Use levels. The density of the monochrome image signal in the signature area 202 is determined based on these density determination levels, and the determination result is converted into a 2-bit digital code by the main control unit 10.
Output to 1.

Referring back to FIG. 4, the image signal of the photographic area 201 converted by the linear density conversion circuit 404 is stored in the buffer memory 411 as an RGB image signal, and is input to the color determination circuit 408. The color determination circuit 408 determines the color of the photographic image pasted on the photographic area 201, that is, determines whether the photograph is a color photograph or a monochrome photograph. This color determination is necessary when performing processing for further improving the image quality of photo printing performed by the printing unit 109. In color printing, four-color printing in which BK (black) is added to three colors of Y (yellow), M (magenta), and C (cyan) as ink colors is generally used. In the present embodiment, as the recording method of the printing unit 109, particularly, the electronic printing method which is an essential condition for achieving automation and online printing as a recording method of a photographic image and a signature image, high image quality can be obtained in a photo-like manner. The thermal sublimation recording method is adopted. In the thermal sublimation recording method, since it is difficult to obtain a pure black dye, a method corresponding to achromatic printing by superimposing YMC inks has been conventionally used. When such superposition is used, it is difficult to achieve both color reproducibility and achromatic color reproducibility because ideal ink characteristics cannot be obtained and the characteristics of the image receiving paper. Therefore, the reproducibility of a photographic image is further improved by using parameters optimized for each of a color conversion process that emphasizes reproduction of a full-color image and a color conversion process that emphasizes processing of a monochrome photograph. The color conversion is a process of converting an RGB image signal into a YMC ink amount signal, which is one of the processing functions of the output image processing unit 107 in the case of the present embodiment. The color determination signal is used for parameter switching information in this case.

As shown in FIG. 13, the color determination circuit 408
21,322, ROM 323, shift registers 324,326 and AND circuits 325,327. The color difference calculation circuits 321 and 322 respectively output RGB color difference signals. The color determination circuit 408 determines a predetermined threshold value in a color space represented by the color difference signals RG and GB, and performs a color determination on a pixel (dot) basis using the threshold value. The ROM 323 outputs 1-bit data of "H" when the color difference signals RG and GB provided as address inputs are equal to or larger than the above threshold value, and "L" otherwise. “H” indicates that the corresponding pixel is color. The shift register 324 has, for example, a 12-bit configuration, and serially shifts the output of the ROM 323 in accordance with the pixel clock GCLK from the image processing control unit 107. The shift register 324 and the AND circuit 325 check whether or not the ROM 323 determines that the pixel is a color signal for 12 consecutive pixels. The shift register 326 also has a 12-bit configuration, and the AND circuit 325
Output from the image processing control unit 107 to the main scanning synchronization signal HSYNC.
Shift. With the shift register 326 and the AND circuit 327, the shift register 324 and the AND circuit 325
It is checked whether or not the line is determined to be a color signal continuously, and a color determination signal is output. 12 dots x 12 lines
Since the color determination signal corresponds to approximately 1 mm ² on the application form 100, the color determination signal indicates a determination result that the photograph in the photograph region 201 is a color photograph when the color signal portion is 1 mm ² or more in total. The color determination signal from the color determination circuit 408 is transferred to the output image processing unit 108 via the image processing control unit 107. Output image processing unit 108
If the color determination signal transferred to the printer is a determination result of a color photograph, the printing unit 109 performs color printing.

The image signal of the handwritten character area 203 converted by the line density conversion circuit 405 is output to the dropout color processing circuit 409.
In addition, in the quantization circuit 410, preprocessing of the recognition processing in the character recognition unit 105 is performed. Image reading unit 104 is color image sensor 3
By using 05, the dropout color processing circuit 409 utilizes the high degree of freedom in selecting the ink color to be used for the application form 100 by selecting each plane of the RGB image signal. The system can be adapted to multiple applications with different out colors. As the dropout color, for example, two types of blue type and orange type
Type is used. The blue system is optimized to drop out when the B signal is selected, and the orange system is optimized to drop out when the R signal is selected. The image signal of the handwritten character area processed by the drop-out color processing circuit 409 is quantized by the quantization circuit 409 to, for example, a quaternary level in accordance with the specification of the character recognition unit 105, and then the buffer memory 41
Stored in one. Thus, in the buffer memory 411,
Processing results of the image signals of the signature area 202, the photograph area 201, and the handwritten character area 203 are stored.

The output image processing unit 108 includes a buffer memory 421, a layout processing circuit 422, and a density correction circuit 42, as shown in FIG.
3. It comprises a color conversion circuit 424, a high-frequency emphasis circuit 425, a character layout processing circuit 426, and a bit expansion circuit 427. The output image processing unit 108 processes the image signal of the photo area 201 and the signature area 202 to be supplied to the printing unit 109 and the image signal of the handwritten character area 203 separately. That is,
The image signals of the photo area 201 and the signature area 202 output as bitmap data from the image processing control unit 107 via the buffer memory 421 are input from the buffer 421 to the layout processing circuit 422, The layout processing is performed according to the layout of the signature area.
The layout process is a process for setting a print position and a print size. The image signal processed by the layout processing circuit 422 is then subjected to density correction processing by a density correction circuit 423 based on an instruction from the image processing control unit 107. Note that the layout processing circuit 22 generates an area identification signal for distinguishing the photograph area 201 from the signature area 202 simultaneously with the layout processing.

The density correction circuit 423 is, for example, a ROM 500 as shown in FIG.
It is composed of The ROM 500 stores 16 types of density correction curves, and one correction curve is selected from the density determination signal and the area identification signal of the signature area 202 expressed by 3 bits. The layout-processed image signal from the layout processing circuit 422 is corrected according to the selected density correction curve and output. Specifically, the density determination signal, the area identification signal, and the image signal are stored in the ROM 500
, And an image signal corrected by the density correction curve is output from the ROM 500.

The image signal whose density has been corrected by the density correction circuit 423 is input to the color conversion circuit 424, where the conversion from the RGB image signal to the YMC image signal is performed as described above. For this color conversion process, for example, a function type process is used, and the optimal conversion is performed by switching parameters depending on whether the image is a color photograph, a monochrome photograph, or a signature. As a method of realizing the color conversion circuit 424, LU using ROM
Various known methods such as a T (look-up table) method, a non-linear conversion method, and a method using a masking equation can be used.

The YMC image signal from the color conversion circuit 424 is
According to 5, the MTF (modulation transfer f
unction) and high-frequency emphasis in consideration of the legibility of the printing result in the printing unit 109 is performed. The high-frequency emphasizing circuit 425 is supplied with the area identification signal generated by the layout processing circuit 422, so that the emphasis coefficient is set to the photographic area 201 and the signature area 202.
Can be switched to be optimal respectively. The image signal output from the high-frequency emphasizing circuit 325 is supplied to the printing unit 109 as photo / signature image data.

On the other hand, the handwritten character area 203 output as code data from the image processing control unit 107 via the buffer memory 421
Are input from the buffer 421 to the character layout processing circuit 426, and are subjected to layout processing according to the layout of the character area on the certificate 200. The image signal processed by the character layout processing circuit 426 is expanded into bit data corresponding to the print dot on the certificate 200 by a bit expansion circuit 427. This operation is called bitmapping. The bit data from the bit development circuit 427 is supplied to the printing unit 109 as character image data.

As shown in FIG. 16, a printing unit 109 in FIG. 1 prints a photograph / signature image and a character image in chronological order.
It has a signature image printing stage 621 and a character printing stage 622. The printing stage 621 receives the photo / signature image data from the output image processing unit 108 via the control unit 620 in the printing unit 109, and performs printing using the YMC ink by the thermal sublimation recording method as described above. The printing stage 622 receives the character image data from the output image processing unit 108 via the control unit 620, and performs printing by the energization transfer recording method. The reason why the current transfer recording method is used for printing character images is that the printing result using the ink ribbon used in this method is infrared OCR
(Optical character reading device) and because it is possible to use ink that is stable up to high temperatures, which is desirable for storage. Being able to check the characters printed on the certificate 200 by infrared OCR has many advantages when considering a security system. The printing unit 109 is provided with three sets of trays 623, 624, and 625 so that printing base papers having different printing styles and paper colors can be set. Tray 623,
Gate drivers 626, 627, and 628 are provided in 624 and 625, respectively, and these are controlled by the main control unit 101 via the print control unit 620. The main control unit 101 recognizes the work category determined by the application form 100 and the contents of the application form 100 (for example, the gender of the application form), and sends a signal specifying a printing base paper to the print control unit 620 based on the recognition.

The input image processing unit 106 shown in FIG. 4 has a buffer memory 411 in the output stage, and the output image processing unit 108 shown in FIG.
Has a buffer memory 421 in the input stage. Such a double-buffer configuration enables multitask processing of input / output operations, streamlines operator confirmation and correction work, and facilitates system construction such as online credit check.

The image input unit 111 in FIG. 1 is for inputting image information other than the image on the application form 100.
D (floppy disk device), ODD (optical disk device), MT (magnetic tape device), IC card reader, etc., a device that reads and inputs image information pre-recorded on a medium, or transmitted by a communication line A device for receiving and inputting incoming image information is conceivable. The image information handled by the image input unit 111 is a form in which face image information and signature image information are developed into a bitmap, a form in which character information is represented by a digital code, and a form in which these bitmaps and digital codes are mixed. Take.

A signal obtained by processing the image signal from the image reading unit 104 in the same manner as described above and a signal obtained by processing the image signal input by the image input unit 111 are sent to the printing unit 109, and can be printed on the certificate 200. . For example, the photograph area on the application form is omitted, and the image reading unit 104 reads only the signature area and the handwritten character area and outputs image signals of these areas. On the other hand, the applicant's face image is captured in advance by an electronic still camera or a video camera and stored as image data in any one of the above-described media, and this is stored in an image input unit.
The data is read out at 111 and input to the main control unit 101.

[Effects of the Invention] According to the present invention, for example, a color image of a face photograph, a black-and-white image such as a signature, an address, a name, and the like on an application created by an applicant applying for the issuance of a certificate such as a driver's license. , Date of birth, gender and other hand-written characters or characters such as printed characters are partially or wholly read by a single scanning scan, divided into regions, and image-processed for each region. By outputting the information to the printing unit, etc., all the information to be printed on the certificate can be processed simultaneously at a time without including a photographing process in the certificate issuance work. A certificate of record format can be issued. Similar effects can be obtained even when an image signal obtained by reproducing an image recorded on various recording media, for example, is input separately from the image progress obtained by reading the image on the application form. . Further, in the present invention, the applicant's face image on the certificate is an image signal obtained by reading a face photograph on the application form or an image signal of the applicant's face image input separately. Since the face image can be selected by the person himself, there is an advantage that mental dissatisfaction such as dislike of the expression can be avoided.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an overall configuration of an image reading processing apparatus according to an embodiment of the present invention, FIG. 2 is a diagram showing a layout of each area on an application form in FIG. 1, and FIG. 1 is a diagram showing the configuration of an image reading unit in FIG. 1, FIG. 4 is a block diagram showing the configuration of an input image processing unit in FIG. 1, FIG. 5 is a block diagram showing the configuration of a frame detection circuit in FIG. 6 is a block diagram showing the configuration of the mark detection circuit in FIG. 5, FIG. 7 is a block diagram showing one configuration of one linear density conversion circuit in FIG. 4, and FIG. 8 is a line in FIG. FIG. 9 is a diagram for explaining the principle of the density conversion circuit, FIG. 9 is a block diagram showing the configuration of the black and white processing circuit in FIG. 4, FIG. 10 is a block diagram showing the configuration of the density detection circuit in FIG. Fig. 11 is a timing chart of each part of the density detection circuit in Fig. 10. FIG. 12 is a diagram showing a density histogram obtained by the density detection circuit, FIG. 13 is a block diagram showing a configuration of the color determination circuit in FIG. 4, and FIG. 14 is a configuration of an output image processing section in FIG. FIG. 15 is a block diagram showing the configuration of the density correction circuit in FIG. 14, and FIG. 16 is a schematic diagram showing the configuration of the printing unit in FIG. 100 Application form 104 Image reading unit 105 Character recognition unit 106 Input image processing unit 108 Output image processing unit 109 Printing unit 111 Image input unit 200 Certificate

──────────────────────────────────────────────────続 き Continued on front page (58) Field surveyed (Int.Cl. ⁷ , DB name) H04N 1/38-1/393 B42D 15/10

Claims (3)

(57) [Claims] 1. An image reading means for reading an image in which each area of a color image, a black and white image, and a character image to be subjected to character recognition are mixed and outputting an image signal, and an image signal output from the image reading means for each area An image reading processing apparatus comprising: an area dividing unit that divides an image signal for each area; and an image processing unit that processes an image signal of each area divided by the area dividing unit so as to conform to a predetermined recording format. . 2. An image reading means for reading an image in which at least two types of image areas are mixed among a color image and a black and white image and a character image to be subjected to character recognition and outputting an image signal; Image input means for inputting an image signal of at least one type of character image of a character image to be recognized, area dividing means for dividing an image signal output from the image reading means for each area, the area An image reading processing apparatus comprising: an image processing unit that processes an image signal of each area divided by a division unit and an image signal input by the image input unit so as to conform to a predetermined recording format. . 3. The image reading means scans a document in a predetermined format selected from a plurality of formats to read an image on the document, and the area dividing means outputs the image from the image reading means in accordance with the format of the document. 3. The image reading apparatus according to claim 1, wherein the divided image signal is divided for each of the regions, and the image processing unit processes the image signal divided by the region dividing unit in accordance with a format of the document. Processing equipment.