JPH05266247A - Picture data processing system - Google Patents

Abstract

PURPOSE:To carry out a processing by separating slips, etc., recorded with the prescribed items in the routine format into a form section and a recorded section. CONSTITUTION:When the picture data on a document consisting of the routine form and the recorded section to be written in the column is inputted from an input section 2, a form deciding section 3 makes an analysis based on the information on the routine form registered in a form information section 4 as for the input picture data and decides the type of the routine form. Based on the type of the decided routine form, a form separation section 5 separates the form section. The picture data corresponding to the recording section separating the form section are imparted to a data coding section 6 to be compressed and united with the form information to generate coding data.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image data processing system for processing a slip or the like in which a predetermined form is filled with predetermined items.

[0002]

2. Description of the Related Art Conventionally, a computer is often used for processing slips and the like in which predetermined items are entered in a standard form.

By the way, an operator manually inputs the entry contents of a slip while watching the slip, but recently, the slip is read by a scanner and the entry contents are read from the read image. A method of inputting a computer by judging the letters and symbols of "" is also considered.

[0004]

However, in the case of adopting such a method, since the image data of the form portion of the slip and the image data of the actual entry portion are fetched together and processed together, Since the processing becomes complicated and malfunctions are likely to occur, and even when the read image data is encoded, the image data of the form part and the fill part is all encoded, so the amount of data used increases and the memory Not only a large capacity is required for the means, but also in the case of transmitting to another place, the transmission efficiency may be lowered due to the large amount of data.

The present invention has been made in view of the above circumstances, and can simplify processing and prevent malfunction.
In addition, it is an object of the present invention to provide an image data processing system that can minimize the amount of data used and can improve the transmission efficiency in the case of transmission.

[0006]

According to the present invention, there are provided an input means for inputting image data of a document consisting of a fixed form and an entry section filled in the entry field, a form information means for pre-registering information on the fixed form, and an input. Form determination means for analyzing at least the type of the fixed form by analyzing the image data input from the means based on the information registered in the form information means, and the image according to the type of the fixed form determined by the form determination means Form separation means for separating the form part from the data, image data corresponding to the entry part in which the form part is separated by the form separation means are compressed, and are integrated with the form information determined by the form determination means, and encoded data And data encoding means for generating

Further, the present invention is a form generating means for generating a fixed form from the form information included in the encoded data, and a decoding for decoding the image data corresponding to the entry portion included in the encoded data to restore the original image. And a synthesizing unit for synthesizing and outputting the fixed form generated by the form generating unit and the original image restored by the decoding unit.

[0008]

As a result, according to the present invention, the processing of a voucher or the like consisting of the standard form and the entry section filled in the entry fields is separated into the standard form part and the entry part of the voucher, and the form part has the form information. In addition, when the image data corresponding to the entry part where the form part is separated is compressed and encoded respectively, and these can be integrated and output, when processing the fixed form part and the entry part together The processing complexity of can be avoided. Further, since the image data is encoded only in the portion corresponding to the entry portion, the amount of data used can be minimized.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 shows a schematic structure of a data encoding unit. In the figure, reference numeral 1 is a document in which necessary items are filled in a standard form, for example, a slip, and this slip 1 is taken into the image input unit 2 as image data by a scanner or the like. In this case, the image input unit 2 may take in the image data of the slip sent from the image communication device such as a facsimile.

The image data taken in by the image input section 2 is sent to the form determination section 3. Form judgment part 3
Analyzes the input image data based on the information about the form registered in the form information section 4, that is, the form of the blank form, and determines the type of the form, the position on the image,
The size etc. are judged.

The determination result of the form determination unit 3 is sent to the form separation unit 5. The form separation unit 5 separates the form portion from the input image based on the result determined by the form determination unit 3, and outputs the image data of only the input data.

The image data composed of only the entry data separated by the form separation unit 5 is sent to the encoding unit 6. The encoding unit 6 compresses the data length of the image of the entry data. Then, the data compressed by the encoding unit 6 is integrated with the information such as the type, position, and size of the form detected by the form determination unit 3, and is generated as one encoded data. The encoded data generated in this way is sent to the data transmission unit 7, transmitted to the outside, or stored in the data storage unit 8. Next, details of the configuration shown in FIG. 1 will be described.

FIG. 2 illustrates the form information section 4 used in the form determination by the form determination section 3. In this case, in the form information section 4, as information about the forms of a plurality of types of slips to be used, the form image 20 representing the image of the blank slip, the form image 20 is blurred, and the resolution of the form is increased. An image 21, a compressed form image 22 with a further reduced resolution of the compressed form image 21 and a compressed compressed image 2 with a reduced resolution of the compressed form 22 image 2
Stores 3 respectively. In this case, in FIG. 2, the size of the image is shown assuming that the size of one pixel is constant, so that it is shown that the image with a small overall size has a low resolution. As described above, the form information section 4 stores compressed image series having different resolutions calculated from the blank form image for each form type.

The above-described blurring process can be specifically performed by the following means. For example, the pixels of the input image are sampled every other pixel in the vertical and horizontal directions, the values of the pixels around the pixel are weighted and added, and this is divided by the total of all weights to obtain the blurred pixel value of the sample pixel. To do. As the weight, for example, it is effective to use the value of the weight table shown in FIG. Here, the pixel with the value 3 shown in FIG. 3 is the sample pixel, and the pixels with the values 2 and 1 are the surrounding pixels. The values 1, 2, and 3 of these pixels are weights for addition. In this case, since the sample points are taken every other pixel, the resolution drops to half and the image size becomes 1/4.
The blurring process is performed when the slip form is registered, and the form information of the form information section 4 is created in advance. Note that this processing is called filter processing, and other methods can be applied.

Then, the form determination unit 3 determines the type of form based on the form information registered in the form information unit 4 as described above. In this case, a compressed image is obtained by performing the same processing as the above-described blurring processing on the image of the input slip in which the necessary items are entered. Then, the smallest compressed image form 23 is extracted from the plurality of form information stored in the form information section 4, and these are compared.

The comparison in this case is performed based on the similarity or distance between the two images. For example, regarding the similarity, each image is regarded as a vector consisting of a sequence of pixel values, and the angle in the vector space of both images is calculated from the inner product calculation. Is used, and as the distance, distance measurement that simply calculates the sum of squares of the differences for each pixel value is used. Then, a plurality of registration slips for which the likely similarity and distance are obtained from the result of this comparison are left for the next process.

In the next process, the same matching process as described above is performed on the second smallest compressed form image 22 among the remaining registered slips, and the number of candidate slips that can be candidates for the input slip is reduced. In the same manner, the processing is repeated for the compressed form image 21, and finally the type of the input slip is determined by comparing the original images. In this case, if the type of the input slip is determined before the comparison of the original images, these processes can be ended midway.

In such a matching process, better results can be expected if the matching is performed in consideration of the positional deviation of the image, the fluctuation of the size, the rotation and the like. For example, by varying the compressed image by conversion processing such as parallel movement, enlargement / reduction, rotation, etc. in a possible range and performing matching, it is possible to perform the matching processing by excluding these variable elements. In this method, not only those effects but also the variation component in the lower compression image can be roughly determined, so that the variation component in the upper matching can be narrowed down, so that it is more stable, accurate and faster. Various matching processes are possible. In particular, regarding the rotation, if 90 degrees rotation or 180 degrees rotation is included in the variation range, it is possible to deal with a slip that has the wrong direction when the slip is set in the scanner. In addition to the operation of converting the input image into a function having no fluctuation component, the fluctuation component thus obtained stores the fluctuation state together with the form information.
It can also be used for processing of transferring an input image in a changed state when reproducing the restored image.

As for the matching processing, other than the black-and-white image, the matching processing using color information on the slip is also effective, and the slip determination processing based on the result of recognizing a specific character or symbol on the slip. Is also effective. A method in which the matching processing is not performed for the entire image but limited to a part thereof is also advantageous in terms of processing speed. Of course,
It goes without saying that the method based on these combinations is effective. Another example of matching is the following method.

First, the ruled line portion is extracted from the input image, and this is matched with the information regarding the ruled line registered in the form information section 4. By determining the correspondence relationship between each ruled line in the image input by this matching and each ruled line registered in the form information section 4, the one having a good corresponding relationship, that is, the thickness, length, position, etc. of the ruled line are good. Matching forms are selected. At this time, the expansion / contraction of the input image and the overall positional deviation information can be obtained from the positional deviation information of each ruled line. In addition, it is possible to extract information on a non-linear expansion / contraction state. These pieces of information are added to the encoded data and output.

In the above embodiment, an example of ruled lines has been described. However, this can be used for figures, photographs, illustrations, logos, bar codes,
It can be anything such as a character string. This is because, for example, a technique for extracting a part that looks like a character is already known, and it is easy to extract the part for other parts as well, and matching of these parts is possible as in the case of ruled lines. Because there is. For example, regarding a character string, matching can be performed by comparing the positions of the smallest rectangles that surround the character string. Next, FIG. 4 illustrates the form separation unit 6 that separates the slip form portion from the input image based on the form determination result of the form determination unit 3.

In this case, (a) of the figure shows an input image of a voucher, in which numerical values are entered in a character frame printed on the voucher. From this state, as shown in FIG. 3B, the character box portion 31 of the slip generated from the form information which is the determination result of the form determination unit 3 is also used as the form position information and the variation component obtained from the form determination unit 3. It is superposed on the input image 32 of the slip based on the.

As a result, as shown in FIG. 6C, the character frame portion 33 of the form portion is erased from the input image 32 of the slip as the processing result in FIG. That is, here, the character frame portion 33 is replaced with the white pixel. Next, as shown in FIG.
The black pixels adjacent to the character frame portion 33 are changed to white pixels. Further, as shown in (e) of the figure, a process of removing a small isolated black region is performed for the black pixel portion remaining in the process of (d) of the figure. In this case, the small isolated black region 34 is determined to be a connected region of small black pixels, but these regions 3
4 can be estimated to have been connected to the disconnection part by the character frame of the large area 35 at the center from the respective positional relationships,
Such small areas 34 are not deleted here, and all other small areas are deleted.

The possibility that the small region 34 is connected to the large region 35 is that the small region 34 exists on the cut side of the cut portion of the large region 35 between the cut portions of the large region 35 and the small region 34. , It is judged by comparing the lengths of both cutting parts and having a length that does not substantially contradict each other, and that the distance along the cutting line between both cutting parts is a size that does not substantially contradict. It

Then, as shown in (f) of the same figure, with respect to the cut portion of the remaining black region, a portion corresponding to the cut line,
That is, with respect to the portion adjacent to the cutting line having the same length as the cutting line, the black portion deleted by the processing described in FIG.

In this way, only the filled-in portion is extracted from the slip in which the form portion and the filled-in portion are mixed. Then, the extracted image shown in FIG. 6F is sent to the data encoding unit 6. In FIG. 6G, the character frame generated from the form information and the filled-in portion extracted by the above-described processing are displayed in an overlapping manner, and the display after decoding is as described above.

In addition to the above-mentioned processing, for example, the portion to be filled in is analyzed, and if the disconnecting portions can be connected to each other, these are connected to virtually reproduce the portion erased by the form information. Is also possible. The written portion can be analyzed by, for example, stroke analysis.

When the input image data is represented by color information or grayscale information, the specific portion can be removed from the form according to these information. That is, it is a method in which an input image is separated by its color and a specific color portion is processed as a storage portion. According to this, it is possible to perform different processings for a plurality of colors of the saved portion, for example, a black portion is taken out as a character portion, data is compressed, a red portion is taken out as a seal stamp portion, and a seal stamp collation is performed. ..

A similar method can be used even when the input image is expressed in shades. Even in these cases, of course, the degree of displacement or expansion / contraction of the unnecessary portion may be obtained from the registered data.

The entry portion may be extracted not for the entire area of the slip but for each entry column. Here, if the extraction area corresponding to the entry field is called a block, this block is defined in advance when the form is registered, and may not be related to the entry field. This is defined in the form of an area surrounded by ruled lines, for example. Of course, the entire area may be defined as one block, but defining each entry field has the following advantages.

First, the efficiency of data compression and the quality of the output image of the reproduced image can be improved in that noise existing in areas other than the block area can be canceled. Also,
Since an appropriate block can be selectively taken out and a voucher can be reproduced when a reproduced image is created, it is possible to create a completed voucher in which various modifications are made. For example, an application example thereof is to virtually have a copy voucher creating function by generating and outputting two types of vouchers that include a stamp image and vouchers that do not include the stamp image when creating a reproduction image. Of course, it is possible to reconfigure the slip as a form different from the form at the time of input by changing the position of the block when reproducing, and one integrated slip such as a list can be created from the information of multiple slips. It is also possible to create, to generate a plurality of slips from one slip, and to generate a plurality of slips from a plurality of slips. These functions are the same not only for the form generation but also for the display.

Further, since the processing method and the compression method can be changed for each block, for example, it becomes possible to distinguish the block in which the form part is not erased from the block in which the form portion is not erased. This function is indispensable because the form cannot be erased in the shaded entry fields. In such a block in which the form is not erased, the form of the area is not generated when the original image is reproduced.

The point that data management can be performed in block units improves the flexibility of the system. For example, this function is very effective in a system in which keyboard input and slip input are mixed for the same type of slip. What is keyboard input? The slip form image is displayed on the display, and the operator moves the cursor on the screen to the entry field of the form image and uses the keyboard or a function such as menu selection that appears on the screen to enter the entry field. It is a method of inputting items, which is the opposite of the automatic input of entry slips. Although the mixture of these causes an important problem in system design, the introduction of the block can easily solve this problem. In other words, the character data and the block in the entry field input by the keyboard may be treated as data of two different expression formats in the same category. Both can be treated as code data by recognizing the characters and symbols in the block, but there is no problem in displaying the block as it is even if the expression format of both is different.

On the other hand, in communication, only necessary blocks may be transmitted, or the blocks to be transmitted may be changed for each destination. Further, also in storage, it is possible to collectively store the same block on different slips or different slip types. At this time, the keyboard input data and the block can be mixed and stored. If the optimum data compression method is used for each block, the compression efficiency will be further improved.

To summarize the above, the above-mentioned processing is basically constituted by erasing the form portion and the small area by the form information. What is important here is that noise is reduced as an object to be erased even in an adjacent area or a small area of the form portion, whereby the data compression efficiency for the image of the next filled portion is significantly improved. For example, the vertical line, which is the unerased portion of the form part in the image shown in FIG. 4C, causes a significant reduction in compression efficiency in the run length encoding process, but in FIG. 4F this is erased. is doing. In the present invention, a small area which is considered to be connected to a larger area is restored as a black area together with a portion erased as an adjacent portion of the foam. As a result, it is possible to prevent a phenomenon such as a blank portion of the cut portion in the reproduced image and improve the image quality. Further, even if the form part of the input image is slightly deformed and changed, one of the basic advantages of the present invention is that such a part can be erased to some extent in the reproduced image, that is, it can be displayed in high quality, that is, beautifully. Is.

Then, the image of only the filled-in portion, in which the slip form portion is separated by the form separating portion 6, is sent to the data encoding portion 6 and is subjected to data compression by, for example, the run length encoding method or the like. Form judgment part 3
By combining with the information such as the type of form obtained in step 1, one encoded data which is the final output form is created and output. FIG. 5 shows a format example of encoded data. In this case, the data is the form data section 51.
And the filled-in image data section 52.

The form / data section 51 includes a total data length 53 indicating the total length of data, a data attribute 54 indicating information about a slip attribute such as an input site name and an input date, and a slip type determined by the above processing. Form type 5 showing expansion / contraction information and displacement information estimated from numbers and input images
5 and the block number 56 indicating the number of blocks stored in the filled-in image data section.

The input image data section 52 is divided into n block data sections 57. Each block data section 57 has a block length 571 indicating the length of the block data section, and a block slip within the slip. Position or number 572 that indicates the serial number of the block determined in the form definition registered in the form information section, and the position and size of the frame in the form removed in or around the block. Frame information 57 indicating information necessary for reproduction
3, an expression format 574 indicating an image expression format such as whether the image data is expressed in binary, a grayscale image, or a color image, and an encoding method 5 indicating an image compression method
75, and block encoded data 576 indicating compressed data. Here, when the block is blank, it is convenient for decoding if it can be expressed in the expression format 574 or the encoding method 575 without being encoded. Further, the frame information 573 may be determined at the time of separating the form, or may be registered in the form information section 4 in advance so as not to be related to the block.

For registration of form information, an image of an actual blank slip is input, predetermined registration data is created using the image data, and registered in the form information section 4. At this time, the operator inputs the form type, block definition, etc. using a keyboard, pointing device, etc., and registers them together with the data. It is also possible that the input image can be modified by the instruction of the operator.
According to this function, it is also possible to create form information from a completed slip. Alternatively, a form image may be automatically generated in the system directly from the form design data, and form information may be created using this. This form information can be transferred to an external device through a communication line, which facilitates unification of data in a distributed processing system in which a large number of systems are collected.

In addition to the accumulated encoded data, form information and a description of a method of the encoding method, for example, a detailed specification of data described in the encoded data such as a source code of an encoding program is stored. You may do it. By doing this, when the system happens to have no form information for a particular slip, or when it does not have a particular encoding program, detailed specifications can be referenced, so new data and methods can be updated. You can avoid the trouble that occurs when the specifications are not registered. This can be similarly applied through a communication line.
If the details of the received encoded data are unknown, it may be possible to make an inquiry to the destination.

Therefore, in this way, the processing for the slip 1 consisting of the fixed form and the entry portion filled in the entry fields should be separated into the fixed form portion and the filled portion of the slip. The form information is determined by the part 3, and the image data corresponding to the entry part is separated from the form part by the form separating part 5, and only the image data corresponding to this part is compressed and integrated with the form information to form one encoding. Since it is generated and output as data, it is possible to avoid the complexity of the conventional process of processing the fixed form part and the filling part together.
The processing can be simplified to prevent malfunction.

Further, since the image data is encoded only in the portion corresponding to the entry portion, the amount of data used can be minimized and the storage capacity of the storage means can be minimized. In addition, it is possible to improve the transmission efficiency due to the small amount of data when transmitting to another place. Next, FIG. 6 shows a schematic configuration of the decoding unit.

In this case, the coded data stored in the data storage unit 81 or sent through the data receiving unit 9 is given to the data decoding unit 10 and the form generating unit 11.

The form generation unit 11 searches the data registered in the form information unit 12 based on the contents of the form data unit 51 of the encoded data shown in FIG. 5, and generates a standard form. In addition, the data decoding unit 10 has the same structure as that of FIG.
The data in the block data part 57 of the input image data part 52 of the coded data shown in is decoded and restored to the original image.

The image of the slip entry portion generated by the data decoding unit 10 and the image of the form portion generated by the form generating unit 11 are sent to the form synthesizing unit 13.

The form synthesizing unit 13 superposes and synthesizes both images on the basis of the position information in the form information. As a result, the slip image input to the transmission unit shown in FIG. 1 is reproduced, and this is output from the output unit 14 of the printer or the display.

That is, in the receiving section of FIG. 6, the form data section 51 of the encoded data shown in FIG. 5 is interpreted according to the form information, and the original slip image is generated based on this. Further, the entered image data portion 52 is decrypted and superposed on the slip image. As for the printout, only the image of the entry portion may be printed on a sheet on which the slip image has already been printed, or the image in which both are overlaid may be printed on a blank sheet. For display display, if two kinds of display memories are prepared in the system, and the slip image is displayed on one side and the filled part image is displayed on the other side, it is not necessary to rewrite the slip image for displaying the same slip. Speed is improved.

The present invention is not limited to the above-mentioned embodiments, and can be carried out by appropriately modifying it within the scope of the invention. For example, in the above-described embodiment, the case where the transmission and reception units are one set has been described, but the present invention can also be applied to a network system configured by combining a plurality of sets. According to the present invention, since the slip image can be compressed into a data having a very small capacity, there is no problem in transmitting a large amount of slip image data through the communication line, and therefore the slip image input can be dispersed. It will be possible.

[0050]

As described above, according to the present invention, processing of a voucher or the like, which is composed of a standard form and an entry section filled in the entry fields, is performed by separating the standard form part and the entry part of the voucher. The part is used as form information, and the image data corresponding to the entry part where the form part is separated is compressed and encoded respectively, and these can be integrated and output. It is possible to avoid the complexity of the processing when the processing is performed, and thereby to simplify the processing and prevent malfunction. Also,
Since the image data is encoded only in the part corresponding to the entry part, the amount of data used can be minimized,
The storage capacity of the storage means can be reduced, and the transmission efficiency can be improved even when transmitting to another place.

[Brief description of drawings]

FIG. 1 is a diagram showing a schematic configuration of a data encoding unit according to an embodiment of the present invention.

FIG. 2 is a diagram for explaining a form information section of the embodiment shown in FIG.

FIG. 3 is a diagram for explaining blurring processing in a form information section of the embodiment shown in FIG.

FIG. 4 is a diagram for explaining a foam separating unit of the embodiment shown in FIG.

5 is a diagram showing a format of encoded data according to the embodiment shown in FIG.

FIG. 6 is a diagram showing a schematic configuration of a data decoding unit of the present invention.

[Explanation of symbols]

1 ... Input slip, 2 ... Image input section, 3 ... Form determination section,
4 ... Form information section, 5 ... Form separation section, 6 ... Data encoding section, 7 ... Data transmission section, 8, 81 ... Data storage section, 9 ... Data receiving section, 10 ... Data decoding section, 11 ...
Form generation unit, 12 ... Form information unit, 13 ... Form synthesis unit, 14 ... Output unit, 20 ... Form image, 21 ...
23 ... Form compressed image, 31 ... Character frame part, 32 ... Input image, 33 ... Character frame part, 34 ... Isolated black area, 35 ...
Large area, 51 ... Form data section, 52 ... Entry image data section.

Claims (3)

[Claims] 1. An input means for inputting image data of a document, which comprises a fixed form and an entry section filled in the entry field, a form information means for pre-registering information on the fixed form, and an input made by the input means. Form determination means for determining at least the type of the fixed form by analyzing the image data based on the information registered in the form information means, and a form portion from the image data according to the type of the fixed form determined by the form determination means And a form separating means for separating the image data corresponding to the entry portion in which the form part is separated by the form separating means, and is integrated with the form information judged by the form judging means to generate encoded data. Image data processing, comprising: Stem. 2. Form generating means for generating a fixed form from the form information included in the encoded data, and decoding means for decoding image data corresponding to the entry portion included in the encoded data to restore the original image. An image data processing system comprising: a fixed form generated by the form generating unit and a combining unit that combines and outputs the original image restored by the decoding unit. 3. An input means for inputting image data of a document consisting of a fixed form and an entry section filled in the entry field, a form information means for pre-registering information on the fixed form, and a form information means registered in the form information means. Form type specifying means for specifying the form type in advance based on the information obtained, form separating means for separating the form part from the image data according to the form type determined by the form type specifying means, and the form part by this form separating means An image characterized by comprising: data encoding means for compressing image data corresponding to the separated writing section and integrating with the form information determined by the form determination means to generate encoded data. Data processing system.