JP3771371B2 - Mark recognition method and apparatus, and recording medium - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a mark recognition technique.
[0002]
[Prior art]
Conventionally, in the field of mark recognition, a special sheet called an OMR sheet is often used, and a check mark is often entered with a pencil or the like in a preset mark entry frame. In such an OMR sheet mark recognition apparatus, generally, when an OMR sheet is optically read and a line segment of a certain length and thickness is detected in the mark entry frame, a mark is placed in the mark entry frame. Judge that it is filled in.
[0003]
In addition to the normal mark entry frame, a reference mark entry field is provided at the top of the OMR sheet so that the writer can enter a plurality of marks in the reference mark entry field. The average thickness a and deviation b of the line segments entered in the entry field are detected, and when the detected line segment thickness c in the mark entry frame is c ≧ ab, the mark is entered. A mark recognition method for determining that a mark is written in a frame has also been proposed (Japanese Patent Laid-Open No. 6-333066).
[0004]
[Problems to be solved by the invention]
For example, when taking a questionnaire for evaluation of products, services, etc., if mark recognition can be used, the questionnaire results can be aggregated efficiently. However, when a questionnaire form is created using an OMR sheet, there is a problem that it takes a considerable amount of money and time, and it is easy for respondents to become a psychological barrier to fill in a limited check box.
[0005]
In order to avoid such a problem, it is desirable to be able to use a questionnaire form in which a selection item is printed on plain paper and the selection item is surrounded by a circle mark or a check mark is added on the selection item. However, in such a questionnaire form, the written mark overlaps with the printed character of the selection item, and there is a large variation in the mark entry position, shape, size, and the like. With the conventional mark recognition technology based on the premise that a mark is entered in the entry frame, accurate mark recognition cannot be expected.
[0006]
Accordingly, an object of the present invention is to provide a mark recognition method and apparatus suitable for mark recognition on a document such as a questionnaire sheet in which only selection items are printed on plain paper.
[0007]
[Means for Solving the Problems]
In order to achieve the above object, in the mark recognition method of the present invention, an image of a document without a mark and an image of a document with a mark are input, and an entire difference image is obtained from the two input images. Generated and the check area of the document is specified. The check area is an area in which a mark may be entered, such as a range surrounding a selection item on a plain paper questionnaire, and is not a limited area such as a mark entry frame of an OMR sheet. . Then, for each specified check area, the number of black pixels and the black pixel ratio on the generated difference image are measured, and the presence or absence of a mark in each check area is determined based on the measured number of black pixels and the black pixel ratio. Determined. For example, when the number of black pixels is equal to or greater than a first predetermined number, or when the number of black pixels is equal to or greater than a second predetermined number and the black pixel ratio is equal to or greater than a third predetermined ratio, it is determined that there is a mark. Is done. In addition, in order to deal with the duplicate mark entry when the mark entry is made alternatively, a logic check is performed on the result of the mark presence / absence determination.
[0008]
Further, the mark recognition apparatus of the present invention generates means for inputting a document image without marks, means for inputting a document image with marks, and generates an entire difference image from the two input images. Means for designating a check area of the document, means for measuring the number of black pixels and the black pixel ratio on the generated difference image for each designated check area, and the number of black pixels and black pixels measured A means for determining the presence / absence of a mark in each check area based on the ratio, and a means for performing a logical check on the determination result of the presence / absence of the mark are provided.
Is the same image.
[0009]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, a mark recognition system using a computer will be described as an embodiment of the mark recognition method and apparatus of the present invention. FIG. 1 shows a flow of mark recognition processing of this mark recognition system, FIG. 2 shows a functional block configuration, and FIG. 3 shows a hardware configuration. The mark recognition apparatus of the present invention can also be incorporated in other information processing equipment such as a character recognition apparatus and a copying machine.
[0010]
The hardware configuration shown in FIG. 3 is a general computer. In FIG. 3, 50 is a CPU, 51 is a work area of the CPU 50, RAM used as a storage area for programs and data, 52 is a hard disk device, 53 is a display device, 54 is a pointing device such as a mouse, and an input device such as a keyboard. , 55 is an image scanner, and 56 is a drive for reading and writing portable recording media such as floppy disks and CD-ROMs. These elements are connected to each other via a system bus 58.
[0011]
A mark recognition processing program for executing the mark recognition processing procedure including steps S1 to S9 shown in FIG. 1 (or for realizing the functions of the functional blocks 20 to 26 shown in FIG. 2) on the hardware of the computer. 60 is placed in the RAM 60 and executed by the CPU 50. For example, the mark recognition processing program 60 is read from the portable recording medium 57 by the drive 56 and directly loaded into the RAM 51, or temporarily stored in the hard disk device 52, and loaded into the RAM 51 from the hard disk device 52 at the time of execution. Specific areas 62 to 68 of the RAM 51 are allocated as the various memories 30 to 36 in FIG. A control program such as an operating system is also loaded from the hard disk device 52 into the RAM 51.
[0012]
Hereinafter, the processing content and configuration of the mark recognition system will be described with reference to FIG. 3 as necessary and mainly with reference to FIGS.
[0013]
In step S <b> 1, the binary image input unit 20 inputs a document master image (a binary image of a document in which marks are not written) that is a target of mark recognition, and stores it in the master image memory 30. Specifically, for example, an unmarked document is read by the image scanner 55 and the read image is stored in the master image memory 30, or the master image read in advance and stored in the hard disk device 52 is mastered. Read into the image memory 30. In FIG. 4, 100 is an example of a master image.
[0014]
In step S <b> 2, the area designating unit 21 inputs information for designating the check area of the document, and stores the check area information in the area information memory 32. Here, the check area is an area in which a mark may be written, in other words, an area for checking the presence or absence of a mark. For example, in the case of a questionnaire sheet in which selection items are circled, individual selection items Is a region that surrounds a part of a selection item. In the master image 100 of FIG. 4, for example, an area indicated by a broken line is designated as a check area. The check area can be specified by a method in which, for example, a master image is displayed on the screen of the display device 53 and the corner position of the check area is indicated on the screen by operating the mouse of the input device 54 or the like. Also, when handling a document of the same type as a document that has already been processed and the information of the check area is stored in the hard disk device 52 as a file, for example, the type of the document is designated and the check area The check area can be designated by reading the information into the area information memory 32.
[0015]
In the next step S <b> 3, the binary image input unit 20 inputs a data image of the same type of document (a binary image of a document in which a mark has been entered) and stores it in the data image memory 31. Specifically, for example, a document on which a mark is written is read by the image scanner 55, and the read image is stored in the data image memory 31, or a data image read in advance and stored in the hard disk device 52 is stored. Read into the data image memory 31. In FIG. 4, 101 is an example of a data image. The data image 101 is the same image as the master image 100 except that the mark M is written in the selection item “2.
[0016]
In the next step S <b> 4, the difference image generation unit 22 generates an entire difference image from the master image on the master image memory 30 and the data image on the data image memory 31 and stores it in the difference image memory 33. To do. More specifically, for example, the difference image between the master image and the data image is corrected, and the difference image is generated by taking the exclusive OR of the two images. In FIG. 4, reference numeral 102 denotes a difference image between the master image 100 and the data image 101, and it can be seen that only the mark M written on the document is extracted. In addition, although the area | region shown with the broken line in the difference image 102 represents the check areas c1-c5, it does not actually exist in a difference image. In addition, there is a possibility that black pixels (noise) due to an error in correcting image misalignment or a dirty document are included in the difference image.
[0017]
In the next step S5, the black pixel number counting unit 23 refers to the difference image and the area information, counts the number of black pixels on the difference image for each check area, and determines the obtained number of black pixels in the black pixel number memory 34. To store. In the next step S6, the black pixel ratio calculation unit 24 refers to the area information and the number of black pixels, calculates the area (total number of pixels) of each check area based on the area information, and calculates the black area with the area of each check area. The black pixel ratio of each check area is obtained by dividing the number of pixels, and the obtained black pixel ratio is stored in the black pixel ratio memory 35.
[0018]
In the next step S7, the mark determination unit 25 refers to the black pixel ratio and the number of black pixels, determines the presence / absence of a mark for each check area, and stores the result in the mark determination result memory 36. The determination rule is, for example, marked when “the black pixel ratio is equal to or greater than the predetermined value Th1” or “the number of black pixels is equal to or greater than the predetermined value Th2 and the black pixel ratio is equal to or greater than the predetermined value Th3”. It is determined that there is. The predetermined value depends on the size of the check area, the type of writing instrument used for mark entry (related to the thickness of the mark entry line, etc.) and the like. For example, Th1 = 3%, Th2 = 10, Th3 = 1% is chosen. According to the mark determination method using the number of black pixels and the black pixel ratio, the length, thickness, and shape of the written mark, the intersection between the mark and the printed character, the noise included in the difference image, and the like It is possible to determine the presence or absence of a mark that is not easily affected. In this way, in the difference image 102 of FIG. 4, it is determined that only the check area c2 is marked.
[0019]
The next step S8 is a step necessary only when a document in which a mark is alternatively written is targeted or a document including a check area in which a mark is alternatively written is targeted. is there. For example, the example of the document shown in FIG. 4 is an example in which a mark is written in only one of the five check areas, so this step S8 is necessary. In this step S8, the logic check unit 26 refers to the mark determination result, and when it is determined that two or more of the plurality of check areas where the mark should be alternatively written are determined to have a mark, Judged as an error. If it is determined that there is a logical error, for example, the number of black pixels in two or more check areas determined to have a mark are compared with each other, and the mark determination result is set so that only the check area with the largest number of black pixels is marked. A process of correcting or invalidating all mark determination results of the form or mark determination results related to a specific check area related to a logic error is performed. I marked a choice with a pencil, but erased it with an eraser and re-marked another choice, but the mark is also valid if the previously entered mark is not erased enough Can be recognized as In such a case, there is a high possibility that the number of black pixels of a mark that has been erased inadequately is less than the number of black pixels of a mark that has been added later, so the correction processing of the mark recognition result as described above may be effective. Many.
[0020]
In the case of continuously performing mark recognition of the same type of form, the processes of step S1 to step S8 are repeatedly executed, the mark recognition process ends when there is no form to be processed next, and the mark recognition result for each form Is obtained in the mark determination result memory 36.
[0021]
【The invention's effect】
As is clear from the above description, according to the present invention, various formats such as a questionnaire sheet for printing a selection item on plain paper and having a mark surrounding the selection item or a check mark on the selection item are printed. It is possible to recognize the mark on the document.
[Brief description of the drawings]
FIG. 1 is a flowchart illustrating a processing flow according to an embodiment of the present invention.
FIG. 2 is a block diagram showing a functional block configuration of an embodiment of the present invention.
FIG. 3 is a block diagram showing a hardware configuration of an embodiment of the present invention.
FIG. 4 is a diagram illustrating an example of a master image, a data image, and a difference image.
[Explanation of symbols]
20 Binary image input unit 21 Region designation unit 22 Difference image generation unit 23 Black pixel number counting unit 24 Black pixel ratio calculation unit 25 Mark determination unit 26 Logic check unit 30 Master image memory 31 Data image memory 32 Region information memory 33 Difference image Memory 34 Number of black pixels memory 35 Black pixel ratio memory 36 Mark determination result memory 50 CPU
51 RAM
52 Hard Disk Device 53 Display Device 54 Input Device 55 Image Scanner 56 Drive 57 Portable Recording Medium 58 System Bus 60 Mark Recognition Processing Program 100 Master Image 101 Data Image 102 Difference Image M Marks C1 to C5 Checked Areas

Claims (6)

A step of marking inputs a document image of the blank, the steps of: inputting an image written mark is entered, and generating the entire difference image between two images the input, written approval region a step of designating the to each check each region where the specified, based on the step and the measured number of black pixels and the black pixel ratio to measure the number of black pixels and the black pixel ratio on the generated difference image And determining the presence or absence of a mark in each check area. 2. The mark recognition method according to claim 1, wherein in the step of determining the presence / absence of a mark, when the number of black pixels is equal to or greater than a first predetermined value, or the number of black pixels is equal to or greater than a second predetermined value and a black pixel ratio. A mark recognition method characterized by determining that a mark is present when is equal to or greater than a third predetermined value.   3. The mark recognition method according to claim 1, further comprising a step of performing a logical check on a determination result of the presence / absence of a mark. Means mark inputs a document image of the blank, means for inputting an image written mark is entered, and means for generating the entire difference image between two images the input, written approval region and means for designating, in each check each region where the specified, means for measuring the number of black pixels and the black pixel ratio on the generated difference image, based on the measured number of black pixels and the black pixel ratio And a means for determining the presence or absence of a mark in each check area.   5. The mark recognition apparatus according to claim 4, further comprising means for performing a logical check on a determination result of the presence / absence of a mark.   A computer-readable recording medium on which a program for causing a computer to execute the processing of each step of the mark recognition method according to claim 1, 2 or 3 is recorded.

Images