JP4238711B2 - Mark sheet method - Google Patents

Description

The present invention relates to a mark sheet used for various tests, questionnaires, questionnaires, forms having mark selection fields, and the like.

Conventionally, a mark sheet is widely known as a method of acquiring data as data by recognizing the mark position by OMR. This method has the disadvantages that a dedicated reader is required and that the mark entry frame and other parts are printed in dropout color, so that the description content of the mark sheet is difficult to visually recognize.

Recently, with the spread of faxes and image scanners, the scanned mark image is read by an optical scanner such as an image scanner or fax machine, and the original mark sheet image is compared with the marked image by image processing. Thus, many optical image reading methods for recognizing marks have been proposed. These proposals eliminate the need for a dedicated reading device, enable a monochrome color mark sheet, and solve the main drawbacks of the optical image reading method. However, many of the known proposals for these optical image reading methods follow that marking is performed by painting a portion surrounded by a mark frame such as ◯, 0, “”, □, etc. with a writing instrument.

In this case, the mark object itself has no meaning as a character, and the mark objects are densely arranged in a line, so that the writer can easily mark the wrong place. The mark work by filling is a burden on the respondent psychologically because there is a large possibility of making an error in the recognition of the mark if the filling is incomplete as well as the labor required for the mark work.
Several proposals for improving these drawbacks have been disclosed, such as Patent Document 1 and Patent Document 2, but it is difficult to say that these defects have been sufficiently improved.

Although not directly related to the present invention, proposals of mark sheets using table frames or ruled lines are disclosed in Patent Document 3, Patent Document 4, and Patent Document 5.
Japanese Laid-Open Patent Publication No. 2002-7980 Japanese Patent Laid-Open No. 5-19436 Japanese Patent Laid-Open No. 2002-120413 Japanese Patent Laid-Open No. 6-215206 JP-A-5-276322

In many cases of the conventional OMR method and the known image recognition method, for example, a method is used in which the inside of a frame, such as ◯, 0, “”, □, etc., is filled and marked. For this reason, the respondent must mark the inside of the frame to be marked almost completely with a writing instrument. For this reason, it takes a long time to work with the mark. In addition, in the case of a mark sheet for visually observing a mark that has been widely used in the past, a method of entering a circle, x, or check mark in the blank of the answer field and answer words such as “Yes” and “No” Although it is commonly used as both marking means for marking by enclosing with an ellipse or the like, these conventional convenient means cannot be used in the known system. In addition, this type of frame that only indicates the position and has no meaning is continuously arranged in a line, so that the respondent selects the appropriate frame to be marked to ensure a limited narrow range. The painting work requires a great deal of concentration on the respondents and is a psychological burden on the respondents. These factors contribute to the unpopularity of respondents to automatic reading mark sheets.
An object of the present invention is to propose a reading method that reduces the burden of mark work on a mark sheet respondent.

The basic means for solving the problem is described in claim 1, and a table formed by a plurality of columns formed by a plurality of horizontal ruled lines and a plurality of vertical ruled lines and a description in the columns. Specify the mark table, specify multiple fields that contain only characters and blanks in the mark table from the mark table column, specify any character or blank in each mark column as the mark target location, Read result table data obtained by recognizing a marked table obtained by marking a figure for marking on a mark target portion with a writing tool and recognizing it with an OCR having both a character recognition function and a ruled line recognition function. Process the data and compare the character string in each mark field of the mark table with the corresponding character string in the reading result table to recognize the presence or absence of a mark for each mark field. A mark featuring It is to propose a over door system.

The basic means is based on the following principle. As in the conventional method, the mark sheet images before and after the mark are directly compared, it is determined whether or not the mark sheet is marked by an image change between them, and the position of the mark is not identified by the position of the changed image. Each of the images on the mark sheet is once converted into character type data including ruled line information, and the character string before and after the mark in the mark column is compared to determine whether or not the mark sheet is marked. At this time, overwriting on the character by OCR is recognized as a change of character type, and the entry of the mark figure in the blank is recognized as addition of the character to the blank. Note that recognizing the entry of a mark figure in a blank as an addition of a character to the blank can be referred to as a misrecognition as a character recognition function of the OCR, but the misrecognition is reversely used. The marked part is identified by the position in the table of the mark column to which the mark target part belongs.

On the other hand, in the conventional optical image system, the marked portion is identified by the geometric position on the mark sheet.
For the sake of simplicity, let us assume a mark sheet printed on white paper with black ink. In a conventionally known method, a narrow range whose position is generally determined by a frame or the like is generally filled and marked, and a change in blackness or image density in the narrow range is detected by a photoelectric tube in the OMR method and by image processing in an optical image reading method. By recognizing the markings. On the other hand, the basic means of the present invention detects a marking by causing a figure to be entered in a mark column and marking it, and detecting a change in a character in the mark column. Therefore, the range of the image range used for detection is greatly different. In the conventional method, the object to be detected is a limited range that is less than or equal to the outline of a normal character, but in the case of the present invention, even if it is the smallest, the range of the size of the character and its surroundings, sometimes the size of the character Use a wide range of several hundred times the target. Such a fundamental difference from a conventionally known method is the basis for producing various features of the present invention.

In addition to the basic means of the present invention described above, the problem is solved by the following additional means (a) to (e) related to claim 2.
(B) Marking a mark target portion other than a character having a hollow shape on the outer peripheral frame by overwriting a graphic that does not contain a coating surface as a mark graphic, or (b) a blank positioned in front of the first character in the mark field or By specifying the space after the end character as the mark target location, (c) By specifying multiple characters or blanks in one mark field as multiple mark target locations and entering the mark figure in a lump (D) a character string that can be easily inferred to have the meaning of the order is designated as the mark target location, and (e) a character string that can be easily inferred to have the meaning of correctness or the like. It is to specify the mark target part.

Here, the role of each means of the present invention and the relationship between them will be described. The basic means has a direct effect of improving the visibility by dividing the mark part with a ruled line, and also makes it possible to recognize the marking without misrecognition using the additional means (a) to (c). Make the means available. As described above, the additional means (a) to (c) are features that can be used as a result of the basic means, but are also means for further enhancing the effects of the present invention, and therefore, they are referred to as additional means in this specification. The additional means (d) and (e) are lower means for enhancing the effects of the present invention by utilizing the additional means (a) to (c).

The feature of the additional means (a) is that it is possible to mark by simply filling in a figure that does not contain a painted surface, which brings about many advantages. The first advantage is that the time required for the marking operation can be improved as compared with the conventional method of filling the inside of the hollow frame-shaped character. In addition, the conventional marking that fills in a narrow range that is delimited by parentheses and the like has required the respondent to be burdened with dexterity and attention, but this burden is greatly reduced.
The greatest advantage produced by (a) is that the mark method that is commonly used in the conventional mark sheet of the visual method can be adopted as it is. That is, when a character is designated as a mark target place, a conventional mark means method can be used in which an answer word such as “Yes” or “No” is set as a mark target part and enclosed by an ellipse. On the other hand, by designating a blank as a mark target portion, a conventionally used ◯ mark, X mark, or check mark can be entered as a mark figure. In known mechanical reading methods, including proposals using image recognition, marks could not be recognized without errors using such conventional mark means.

In the additional means (A), there are advantages and disadvantages depending on whether the mark target portion is a character or a blank. First, an advantage of designating a blank as a mark target location is that a simple graphic can be used as a mark graphic than when a character is designated as a mark target location. This is because, in the case of characters, it is necessary to use a figure for which the character to be marked is recognized as another character by OCR, but in the case of a blank, if it is recognized as some character by OCR This is because it is sufficient. Therefore, in the case of a blank, even the minus sign can sufficiently fulfill the role of the mark figure. On the other hand, when a character is designated as a mark target location, it may be necessary to designate a graphic larger than the case of a blank as a mark graphic in order for the character that is the mark target location to be recognized as another character by OCR. It is disadvantageous. However, it may be advantageous to specify characters. The first advantage is that the recognition accuracy at the time of mark recognition can be improved as compared with a case where a blank field having no character in the field is a mark field and a blank is a mark target portion. Let's assume a case where a blank is used as a mark field and a minus sign is used as a mark figure. In this case, there is no problem because the mark is correctly recognized for the marked target field. However, it may be mistakenly recognized that a mark target field that is not marked is marked. This is because dirt due to an eraser flaw may be recognized by the OCR as having a comma-like character in the mark target field, which may result in a false recognition that an unmarked mark field has been marked. is there. On the other hand, when a character is designated as a mark location, this type of error can be corrected because it can be detected by the logic that the character in the mark column does not change to indicate that the column is not marked. On the other hand, if the blank field is set to the mark field, the only way to compensate for the above-mentioned misrecognition is that if the character in the mark field is a comma or other character, recognition of the mark is temporarily suspended and the marked table is visually checked. There is no choice but to reexamine and correct. The second advantage of specifying a character as a mark target is better than specifying a blank, because it makes it easier for respondents to recognize that a mark is a target than when targeting a blank. is there. Further, this can be emphasized by the additional means (d) (e).

Next, the advantage of the adding means (c) is that in the adding means (b), in addition to the advantage of having a blank as a mark target, characters such as numbers corresponding to the symbols of choices are used in the characters or character strings in the mark field. Alternatively, the meaning of the mark field can be more clearly informed to the respondent by using a phrase that is meaningful as an answer such as “Yes”.
The advantage of the additional means (d) is that it has the advantage of making the character a mark target position in (a) in addition to the advantage of using the blank as the mark target position in the additional means (b), and also the marking. Since the number of logics that can be used for the determination is larger than in the case of (A), it can be corrected more accurately than in the case of (A). Furthermore, since there are multiple characters, it is possible to use multiple characters such as “Yes” and “No” for the additional means (e) (f), so the additional means (d) (e) is more effective. Can be.

Any additional means (d) (e) can be used to mark other than the intended mark part by the respondent by matching the expression of the mark part with the option code of the answer or giving the answer correct or incorrect meaning. It is a means for reducing.

Reduction of mark work burden on respondents The reduction of the mark work burden, which is an object of the present invention, has been realized as follows.
Visibility is greatly improved by separating each mark column with a ruled line. In addition, it is possible to match characters such as numbers, alphabets, etc. that can be easily inferred with the mark as a mark figure or a mark location with the answer option code. As a result, it is possible to make a mark figure or a mark target portion that is semantically related, and further, the visibility is improved psychologically as a result of not requiring the mark position as strictly as in the conventional method. In addition, a method of entering a circle mark, a cross mark, or a check mark in a blank space commonly used in a visual reading type mark sheet and a method of marking an answer word such as “Yes” or “No” by enclosing it with an ellipse, etc. Both are now available. In addition, it is possible to mark by simply entering simple figures (including symbols) or simple figures such as straight lines, including short lines, “○”, “×”, etc., and surrounded by a frame. Compared with the conventional method that required to fill the entire blank area, the mark work time was significantly reduced. In this way, the burden on the mark sheet writer can be greatly reduced by shortening the mark time and improving the visibility including psychological effects.

Improving the convenience of the mark sheet user The mark sheet is determined only by recognizing the character type, so the mark sheet may be printed in a single color. The equipment necessary for practical use of this mark sheet method can be implemented only with a general-purpose scanner, image recognition method software that can recognize general-purpose table ruled lines, and general-purpose spreadsheet software. , All data processing related to mark sheet application, from mark sheet design to mark recognition processing, statistical processing of mark recognition results, and graph creation, is easy operation and consistent without advanced technical knowledge. It was possible to carry out.

Prior to the description of the embodiments, terms used in the claims of the present invention will be described. The mark table referred to in claim 1 is a main component of a mark sheet such as a test answer sheet and a questionnaire answer sheet, and is arranged on the mark sheet clearly separated from other parts.
The other parts include arbitrary characters, figures, photographs, tables other than the mark table, and the like.
The mark sheet is a thing in which characters, figures, and the like are drawn on a surface of a medium such as paper or film by using a technique such as ink. In general, a mark sheet is a printed matter in which dark characters or figures are drawn on the surface of a white, light or colorless medium, but on the contrary, white, light or colorless characters or figures are drawn on the surface of a dark medium. The printed matter can also be used as the mark sheet of the present invention.

The “ruled line” referred to in Claim 1 is generally also referred to as “table frame line”.
The “description” in claim 1 means characters, figures, photographs, etc. in the table column.
The “character” in claim 1 includes a symbol and includes a character used not only in Japanese but also in a foreign language. However, the characters do not include white space and control characters.
In order to achieve the effects of the present invention, it is preferable to use “characters” that are correctly recognized when read by OCR. In general, for Japanese, the typeface typefaces such as Mincho and Gothic are more suitable than line and cursive.
In claim 1, “description” is composed of arbitrary characters, a graphic including a photograph, and a blank.
The “Mark Table” in claim 1 may include characters in the columns and columns including figures and photographs in addition to blanks. However, as defined in claim 1, characters and characters are included in the columns. Only fields that contain only white space can be specified as mark fields.
As the “mark target portion” in claim 1, a character or a blank in the mark column is designated.

In the present invention, as specified in the claims, one or more mark target portions are designated in each mark column. In addition, you may designate the mark object location which differs separately for each mark column.
In the present invention, the range of the mark target portion is a space between the adjacent character and the range occupied by the virtual rectangle having the smallest area that can surround the figure of the character when the mark target portion is a character. If the mark field is horizontal writing in the range of the sum with the range, the range including the blank space between the character and the upper and lower horizontal ruled lines, or in vertical writing, the blank space between the character and the left and right vertical ruled lines It is set as the range which added the range which added the range. On the other hand, the range when the mark target portion is blank is a quadrangular range defined by a plurality of ruled lines and character boundaries surrounding the blank. Here, the boundary line of the character is assumed to be an outer peripheral line of a virtual quadrangle having the smallest area that can surround the figure of the character.
However, this range does not limit the entry of entries outside this range. Actually, in the method of the present invention, in many cases, the mark recognition is correctly performed even if the entry exceeds this range.

The “mark graphic” referred to in claim 1 includes a graphic having a character shape in addition to a normal graphic drawn with dots, lines, curves, and the like.
Writing instruments used for “writing” in claim 1 include stamps in addition to ordinary writing instruments such as pencils and ballpoint pens.
The “written by writing” in claim 1 is performed by adding a mark figure to the blank if the “marked location” is blank, and when the “marked location” is a character. Do this by overwriting on the character.

The “OCR having both a character recognition function and a ruled line recognition function” in claim 1 is a method in which a marked table consisting of a plurality of columns including a character string is optically captured as an image, and the relative processing of each column is performed by image processing. This means a device that reads and outputs as a table by converting the characters into the approximate characters and recognizing the characters in each column as characters while maintaining the position.
In general, an OCR having both a character recognition function and a ruled line recognition function has a performance of automatically recognizing handwritten characters and figures as similar characters while ignoring misidentification. Must use this kind of OCR. Although not a general OCR, a type of OCR that recognizes and outputs only characters and ruled lines after removing graphics from an image is not suitable for the present invention.
Further, the OCR of the present invention includes a system constituted by a scanner and commercially available character recognition software having a ruled line recognition function.
Some OCRs having a character recognition function and a ruled line recognition function have a function of recognizing handwritten characters in addition to printed characters. In the present invention, this type can also be used. When this type of device is used, there is an advantage that a handwritten portion such as a respondent's name can be simultaneously read and output as data along with the mark reading.

The data of the reading result table in claim 1 is data that can reproduce and display the reading result table if it is input to an appropriate apparatus. Normally, this table data consists of data obtained by adding ruled line data to character data corresponding to the character strings in each column in the table, and the data is a ruled line character code representing a ruled line between the character codes of each character in the table, This is character type data in a format in which a control code representing a line feed is inserted.

In the present invention, the data of the reading result table is processed to determine the presence or absence of a mark in each mark column. For this purpose, a mark is identified and its position is compared by comparing the character string in each mark column of the virtual table with the character string in each mark column of the corresponding mark table based on the data of the reading result table obtained by OCR. May be identified. Also, the data of the reading result table obtained by OCR is output to spreadsheet software, etc., converted into the reading result table in the software, and then processed in the same manner for the existing table in the software, A determination may be made. In general, it is easier to design a data processing program by outputting data and then processing the data once it has existed, rather than performing data processing on a virtual table. The latter also has an advantage that it can be corrected at this stage because the actually displayed table can be visually confirmed.

The “character in the outer peripheral frame hollow shape” as used in claim 2 is a character having a shape that is formed only by an outer frame that goes around the outer periphery, such as a circle, a square, a triangle, etc. . An example of the outer peripheral frame hollow character is shown in FIG.
The term “painted surface” as used in claim 2 means a figure in which the inside of a closed curve drawn with a line is painted with the same color as the line. The figure that does not contain a painted surface refers to a figure drawn with dots and lines that does not include any painted surface. For reference, FIG. 3 shows an example of a figure that does not contain a painted surface and an example of a figure that contains a painted surface.
The “figure that does not contain a painted surface” as used in claim 2 means a graphic that is formed by only dots and lines and that does not include a painted surface in any part of the graphic. FIG. 3 shows an example of a figure that does not contain a paint surface and an example of a figure that contains a paint surface.

The “character string that can be easily inferred to have the meaning of the order” as used in claim 2 is a single character or a plurality of characters representing numbers such as numbers, circled numbers, and Chinese numerals, and alphabets A, B, C, which means a character string composed of one or more characters such that the characters have the meaning of the order according to the conventions such as C, A, I, U, E.
In claim 2 (e), “a character string that can be easily inferred to have a meaning such as right or wrong” means a meaning required by the respondent as an answer, for example, neither positive nor negative. Means a character string consisting of one or more characters such as “Positive”, “Yes”, “No”, “I don't know”, “Yes”, “No”, “YES”, “NO”, etc. .

FIG. 1 shows an explanatory diagram in which a mark is attached to the mark sheet of Example 2 as an example of the mark sheet of the present invention. With reference to this figure, terms and features in the mode of the present invention will be described. This mark sheet includes a table composed of 27 columns formed by eight horizontal ruled lines (reference numeral 1, etc.) and five vertical ruled lines (reference numeral 2, etc.). This table is a mark table. The mark sheet includes parts other than the mark table made up of characters (symbol 10 or the like) and figures (symbol 11). The column of the mark table can be divided into a column having characters and a column having no characters. In this example, the column denoted by reference numeral 5 is a column containing only blanks, and the others are all columns containing characters. In the present invention, any of a field containing only characters, a field containing white space, and a field containing characters and white space can be designated as a mark field. In this example, a field containing only white space is not designated as a mark field. In this example, twelve fields including the characters “Yes” or “No” in the direction indicated by the arrow 3 are mark fields, and the other 15 fields are fields other than the mark field.
According to the description indicated by reference numeral 11, the mark target portions are the characters “Yes” and “No”
It is specified that the mark figure is two parallel lines.

In all of the following examples including this example, ten simulated respondents were made to answer one mark sheet each.
In this embodiment, the basic means and the adding means (a) are characters to be marked, and (d) is further used. Ten questions were given answers to each of the four questioned questions with the symbols a, i, c, and d, and one of the choices was selected and answered to 10 students. The marked mark sheet of student number 1 which is one examinee of the answer sheet is shown in FIG. The table included in FIG. 1 is a marked table referred to in the claims. In addition, a table obtained by removing two parallel lines written in the mark target column of the answer column from the marked table is a mark table, but it is omitted for the sake of space saving. According to the description in the mark sheet, two parallel lines are designated as a mark figure, and characters “a”, “i”, “c”, and “e” in the answer column are designated as mark target portions. As is clear from this figure, according to the present embodiment, according to the adding means (d), characters that can be easily inferred in the order of a, i, c, and e are used as the mark target portions, and the representation of the mark target portion is used as an option code It was possible to eliminate anxiety about incorrect marking of unintended parts by respondents. Since a, i, c, and d are characters of the outer peripheral hollow frame shape at the same time, the additional means (a) is also used. In addition, two additional parallel lines, which are figures that do not contain a painted surface, are used by the adding means (a). The burden of marking work is light because it only overwrites parallel lines.

The examinee marked and answered according to the instructions on the mark sheet. The appearance of the marked table of the student number 1 is shown in FIG. The image of each answer sheet that has been answered is read with a scanner, converted to tabular data containing character and ruled line information with commercially available character recognition software that has both a character recognition function and a ruled line recognition function, and the data is commercially available. Output to spreadsheet software. FIG. 6 shows a reading result table displayed on the display as a table in the spreadsheet software. The number at the left end of the table is the row number of the cell of the spreadsheet software, and the alphabet above the table is the column code of the cell. Here, each column (that is, each cell) of the reading result table obtained by reading with the OCR shown in FIG. 6 is in a position that is completely similar to the position of the corresponding column of the mark table. Correspondence between the column position (that is, the cell position obtained from the row number (1 to 5) and column symbol (A to K) of the cell of the reading result table) and each mark column of the mark table can be easily performed. In FIG. 6, it can be seen that the character at the marked portion has changed to various characters different from the original character. Next, the character string in each mark field of the original mark table is converted into the character string in each mark field of the original mark table by a data processing function using VBA, script, etc. provided in the calculation software. It is determined that the character strings are different from each other as a result of the comparison, and the processing result is displayed as 1 in a place marked in another table in the calculation software and 0 as a place not to be displayed. Is shown in FIG. After checking the duplication of marks in two or more options from the data shown in FIG. 7 and the absence of a mark, the option codes marked for each question (for example, the first question is “A”) are shown in FIG. The answer option list for the question is entered in the row of student number 1 by the data processing function of the software. At this time, although omitted for simplification of explanation, in the same answer sheet, a separate table for student numbers is provided for student numbers, and each student is marked, and the student is recognized in the same way. The data was acquired and automatically entered in each student number cell in FIG. Repeat the same procedure for the 10 marked tables for all candidates. Further, FIG. 8 shows the correct answer for each question.
Furthermore, the answer options for each question in the table of FIG. 8 are compared with the correct answer, and each student is asked for correctness for each question. The score, correct answer rate, rank and correct answer rate by question and average correct answer for each student The rate was calculated using the same calculation function. Using the software's graph creation function, it was possible to create and display graphs such as a score distribution table with simple operations.

This embodiment is an example using basic means and additional means (a) (c) and (e).
(In this embodiment, as in the first embodiment, as in the first embodiment, a combination of a scanner and a character recognition software having a commercially available character recognition function and ruled line recognition function is used as an OCR, and the output of the OCR is used as a character. The data was processed by inputting the data into commercially available character recognition software and then processing the data into a table, that is, reading the image of each answered answer sheet with a scanner and using a commercially available character recognition function and ruled line recognition function. The data is converted to data containing information on characters and ruled lines with the character recognition software, and the data is output to a commercial spreadsheet software and read in the spreadsheet software, which is the result of reading with the OCR. The result table was subjected to processing for mark determination and position identification using the data processing function attached to the spreadsheet software. The marked mark sheet shows an image.The marked table is a table portion in this image.Also, since it is obvious, the mark table is omitted from the mark table, and “Yes” is displayed in the answer column of this image. 9 is obtained by removing the parallel drawn lines drawn above “No.” FIG. 9 shows a read result table obtained by reading the marked table of the marked mark sheet by OCR. You can see that “Yes” and “No” in the marked part have been changed to other characters, and in this example, two or more marks can be marked at the same time by overwriting “Yes” or “No” with two lines. Since the location is specified, means (d) is used.
Furthermore, by using “Yes” or “No” having a meaning such as right or wrong at the mark target location, and using the additional means (e), the visibility of the mark target location to the respondent is enhanced.
Also in this example, as described in the basic means, the presence or absence of marking can be determined by comparing the entire character string in the corresponding reading result table column with the corresponding character string in the mark table. However, as another logic, it is possible to determine whether or not each mark field is marked by the fact that the first character is a different character type from the corresponding first character in the mark table, and it corresponds to the mark field as another logic It can also be determined by the fact that the end character in the column of the read result table to be used is a character type different from the end character of the corresponding mark table. If no error occurs in OCR character recognition, these three kinds of logic all give the same mark recognition result. However, if there is an error in OCR character recognition and there is a difference between these results, it can be corrected by re-checking the marked sheet visually. In addition, an error determination that a mark field that is not actually marked is marked can be similarly corrected. In this example, the results of the three types of logic were the same for all ten mark sheets.

This embodiment is another example in which the additional means (a), (c) and (e) are applied in addition to the basic means as in the second embodiment. Example 2 is the same as Example 2 except that the mark figure is an oval figure surrounding the outside of the mark part. FIG. 10 shows an image of a marked mark sheet entered by one respondent. FIG. 11 shows a reading result table obtained by reading the marked table of the marked mark sheet of FIG. 10 by OCR. The character strings in the corresponding mark fields were compared, and the marks were recognized as in Example 2. There was no error in mark recognition on all 10 mark sheets.

In this embodiment, the additional means (a), (b), (c) and (e) are applied in addition to the basic means. An image of a marked mark sheet completed by one respondent. This is the same as in the case of the mark sheet of Example 2 except that “Yes” and “No” spaces are added as mark locations.
FIG. 13 shows a reading result table obtained by reading the marked table of the marked mark sheet of FIG. 12 by OCR. The character strings in the corresponding mark fields were compared, and the marks were recognized as in Example 2. There was no error in mark recognition on all 10 mark sheets.
In the second embodiment, the mark figure is a parallel horizontal drawing line of two lines in the second embodiment. In this embodiment, one horizontal drawing line is designated. The reason for this is that in this embodiment, a single horizontal drawing line is sufficient because the mark is recognized without error by only a simple figure by the additional means (b). Next, for reference, the same experiment was repeated by changing the character recognition software to a commercially available character recognition software and ruled line recognition function made by another manufacturer. FIG. 14 shows a reading result table obtained by reading the mark sheet of the one respondent with the OCR. The marked characters in the mark column are completely different from those in the OCR, but the recognition of the marks was performed correctly in this case, and there was no error in the recognition of all the marks in the 10 mark sheets.

This embodiment is an application example in the case where the blank of the additional means (a) is used as the mark target place in addition to the basic means. FIG. 15 shows an image of a marked mark sheet entered by one respondent. Since the check mark, which is a graphic for a mark in this embodiment, is a graphic that does not contain a painted surface, an adding means (A) is used. The check mark that has been conventionally used as a positive answer can be used as a mark figure by the additional means (A). Further, as in the example of FIG. 17, a check mark can be arbitrarily entered in a wide mark field without any particular restriction on the position. This frees respondents from the burden of filling in a narrow area. This is simply to specify the range in which the function of the box surrounded by the ruled lines of the present invention accommodates characters. Like the mark frames such as ○, 0, “”, □ in the conventional method, This is derived from the fact that it completely differs from the function of specifying the filling range. This is a feature common not only to this embodiment but also to the system of the present invention generated by the basic means of the present invention. FIG. 16 shows a reading result table obtained by reading the marked table portion of the marked mark sheet of FIG. 15 by OCR. Whether all marks were marked or not was recognized by the presence or absence of characters in the fields. In order to prevent misrecognition that a non-marked column was marked, consideration was given to allow visual re-inspection when the mark column is a comma or a dot. There was no error in mark recognition on all 10 mark sheets. In fact, there was no opportunity for visual inspection.

In this embodiment, in addition to the basic means, the adding means (A) and the adding means (B) are applied. FIG. 17 shows an image of a marked mark sheet entered by one respondent. Since the mark, which is the mark figure of this embodiment, is a figure that does not contain a painted surface, the adding means (A) is used. The circle that has been conventionally used as a positive answer by the basic means and the additional means (A) can be used as a mark figure, and the respondent can understand the meaning of the mark. Further, since the mark target portion is a space in front of the first character, an adding means (b) is also used. In the case of the addition means (b), there is an advantage that the characters in the column can be clearly shown to the respondent by obtaining meaningful characters or character strings associated with the answers as in this example. FIG. 18 shows an output result obtained by reading the marked table portion of the marked mark sheet of FIG. 17 by OCR. As shown in FIG. 18, the written circle is recognized as various characters by the OCR, but the presence or absence of marks in all the mark fields is correctly recognized by the basic means of the present invention. Mark recognition was performed on 10 marked mark sheets, and all marks could be recognized without error.

This embodiment is an example in which different means are designated and marked according to the mark column, and FIG. 19 shows an image of a marked mark sheet entered by one respondent. In this example, all of the marks except for the “question sheet” on the mark sheet are a mark table. In order to save space, the description of the following processing is omitted, but in this case as well, all 10 marked mark sheets could be recognized without error.

It is explanatory drawing of a term, Comprising: The code | symbol is attached | subjected to each part of the mark sheet of Example 2. FIG. Example of outer frame hollow characters Examples of figures that do not contain painted surfaces Respondent marked mark sheet for student number 1 in Example 1 Example of reading result table obtained by reading with OCR of student number 1 in Example 1 Mark recognition result of respondent of student number 1 in Example 1 Option answer results for each question of all respondents in Example 1 Example of marked mark sheet of Example 2 An example of a reading result table obtained by reading with the OCR of the second embodiment Example of marked mark sheet of Example 3 An example of a reading result table obtained by reading with the OCR of Example 3 Example of marked mark sheet of Example 4 An example of a reading result table obtained by reading with the OCR of Example 4 Another example of the reading result table obtained by reading with the OCR of the fourth embodiment Example of marked mark sheet of Example 5 An example of a reading result table obtained by reading with the OCR of Example 5 Example of marked mark sheet of Example 6 An example of a reading result table obtained by reading with the OCR of Example 6 Example of marked mark sheet of Example 7

Explanation of symbols

1 One of the horizontal ruled lines (this table has a total of 8 lines).
2 One of the vertical ruled lines (a total of 5 in this table).
3 Fields designated as mark fields (in this table, there are a total of 12 fields including “Yes” or “No”).
4 Fields other than the mark field that contain characters in the field (there are 8 in this table).
5 Fields other than the mark field that do not contain characters.
6 White space in front of the first character in the mark field (the first character “has six places in front of”).
7 White space behind the last character of the mark field (there are 6 places behind the last character “I”).
8 The first character in the mark column (four “yes” in “yes”, “yes” in front of each “no”, there are a total of twelve places).
9 Example of the last character in the mark field (4 “yes” in “yes” and 4 “e” in each front in “no”, there are 12 places in total)
10 Characters in an area other than the mark table in the mark sheet.
11 A figure of an area other than the mark table in the mark sheet.

Claims (2)

A table formed by a plurality of columns formed by a plurality of horizontal ruled lines and a plurality of vertical ruled lines and a description in the column is used as a mark table, and only characters and blanks are included in the mark table columns. By means of designating a plurality of fields including a mark field, designating any character or blank in each mark field as a mark target location, and writing a mark figure on the mark target location in writing The data of the read result table acquired by recognizing the marked table acquired by marking with the OCR having both the character recognition function and the ruled line recognition function is processed, and the character string in each mark column of the mark table And a mark sheet method for recognizing the presence or absence of a mark for each mark field based on whether or not the character strings in the corresponding fields of the reading result table are different from each other A figure that does not contain a painted surface is entered as a mark figure on the mark target area other than the outer frame character, and a space that is located in front of the first character in the mark field or a space that is located after the last character Can be easily analogized to specify the mark target location, to specify a plurality of consecutively aligned characters or blanks in one mark field as the mark target location, and to have the meaning of the order 2. The method according to claim 1, wherein at least one of designating a character string as a mark target location and designating a character string that can be easily inferred as having a meaning such as right or wrong as the mark target location. Mark sheet method.

Images