JPH0528258A - Form input device with character/graphic separation device - Google Patents

Abstract

PURPOSE:To easily input a form without permitting an operator to erase a character part for the form where a character and a graphic are mixed by providing a form processing means which executes a copy processing with a separated graphic and executes conversion into form information. CONSTITUTION:A picture input part digital-inputs a document picture where the character and the graphic are mixed, and an image memory 2 stores inputted digital data. A character/graphic separation part 3 executes a processing obtaining a picture consisting only of the character and a picture consisting only of the graphic from the original picture where the graphic and the character are mixed in a ruled line, which is stored in the image memory 2. A form picture memory 4 stores the graphic picture separated by the character/graphic separation part 3. A form processing part 5 encodes the form for inputting it to a computer for the graphic picture stored in the form picture memory 4, executes the copy processing of vectorized data and executes a processing for conversion into the document of the computer.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention separates characters and figures from a document image in which characters and figures (form part) are mixed,
The present invention relates to a form input device for inputting only a graphic part in plain text.

[0002]

2. Description of the Related Art When a form such as a slip is input to a computer (such as a word processor), an operator generally inputs it through an input device such as a keyboard or a mouse. In recent years, in order to eliminate the hassle of operating keyboards, etc., a document printed on paper or carefully handwritten is captured as an image, and the line portion of the form is vectorized, and form input is performed. Research is being carried out (for example, IEICE Technical Report PRU87-
35).

[0003]

However, since these are processes for the image of only the line part, the input from a general form document in which characters and figures (form part) are mixed is once performed by the operator. Was necessary to erase the character part. Also, as the target of form input,
Since it is only straight lines, it is only a printed matter or a nicely handwritten document, and a roughly handwritten document (form)
It was impossible to input. The present invention has been made to solve these problems of the prior art. That is, it is an object of the present invention to provide a form input device capable of easily inputting a form even if the form has a mixture of characters and graphics without the operator having to erase the character portion. It is a thing.

[0004]

In order to achieve the above object, the present invention provides image input means (1 in FIG. 1) for inputting a document image as digital data, and an image memory (FIG. 1) for storing the input image. 1), 2), a character / figure separating means (3 in FIG. 1) for separating a figure and a character in an image, and a form processing means (5 in FIG. 1) for performing a clearing process on the separated figure and converting it into form information. ) And are provided.

[0005]

The image inputting means (1) reads an image (for example, FIG. 2) including a figure and a character to be used as form information and stores it in the image memory (2). The stored image (for example, FIG. 2) is separated into a character (for example, FIG. 3) and a graphic (for example, FIG. 4) by the character / graphic separation means (3).
Then, the separated figures are subjected to a clear writing process by the form processing means (5) and converted into form information. According to the present invention, by performing character / figure separation processing before form processing, even for a form in which characters that cannot be handled by existing form input are mixed, only the form portion is extracted and a computer or word processor is extracted. Can be input to the application device such as. Further, according to the present invention, since the fair copy processing is performed on the extracted form portion, it is possible to input even a roughly handwritten form.

[0006]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS As shown in FIG. 1, a form input device according to one embodiment of the present invention comprises an image input section 1 for digitally inputting a form document, an image memory 2 for storing the input image, and characters only. Character / figure separating unit 3 for separating the image of FIG. 3 and an image of only a figure, a form image memory 4 for storing an image of only a figure (form) separated by the character / figure separating unit 3, and the figure (form) The image processing unit 5 includes a form processing unit 5 that vectorizes only the image and converts it into a document format.

More specifically, the form processing section 5 is a vectorization section 51 for vectorizing a line segment on an image containing only a figure (form), and a clear copy for applying a shaping process to the vectorized image. It is composed of a processing unit 52 and an application processing unit 53 for converting to a document format for input to a computer (word processor or the like).

The operation of this embodiment having the above arrangement will be described below. The image input unit 1 digitally inputs a document image in which characters and figures (forms) are mixed.
FIG. 2 shows an example of a document image (original image). The image memory 2 stores the digital data input by the image input unit 1. The character / graphic separation unit 3 stores an image of only characters as shown in FIG. 3 from the original image in which the graphic (form) of the ruled line and the character as illustrated in FIG. Processing for obtaining an image of only a figure (form) as shown in FIG. Various character / graphic separation methods have been proposed, and any one of them may be selected and used. The selection of the character / graphic separation method depends on the form of the input document. For example, if the target is a form document that is composed only of vertical / horizontal lines, then, for example, Iwashiro 3
A Study on Introduction of Production System in Figure Separation Processing "Technical Report of the Institute of Electronics and Communication Engineers PRL83-63,
When the method described in p.67-74 can be used and the ruled line is constituted by a dotted line or the like, for example, Japanese Patent Application Laid-Open No. 2-15
The method of 9690 can be used. The form image memory 4 stores the graphic (form) images separated by the character / graphic separation unit 3. The graphic image in the original image of FIG. 2 is as shown in FIG.

By the form processing unit 5, for the figure (form) image stored in the form image memory 4,
The code for inputting to a computer (word processor, etc.) is coded (vectorized), the vectorized one is written (shaped), and converted into a computer document. Specifically, first, the vectorization unit 5
According to 1, vectorization processing is performed by obtaining both end points of the line segment for the form image. The vectorization processing includes the communication technique PRL83-8 and the communication technique P.
RL85-24, Practical Technique PRU86-89, JP-A-2
It is possible to use the method described in the literature such as JP-A-210586. In addition, Japanese Patent Application Laid-Open No. 2-
It is preferable to use the method described in Japanese Patent No. 105265 (the title of the invention, “Image data vector conversion device”), because it is possible to perform a clear copy process together with vectorization.

FIG. 5 shows an example of the configuration of vectorization using the "image data vector conversion device". The vectorization unit 51 can convert the image data into vector data by a simple process mainly including scanning. As shown in FIG. 5, the binary image stored in the form image memory 4 is orthogonalized. A predetermined process is performed by scanning in a direction (here, the X-axis direction and the Y-axis direction). The scanning and processing in the X-axis direction are performed by the X-axis direction scanning unit 511, the continuous black pixel counting unit 512, the black pixel centroid extracting unit 513, and the centroid connecting unit 514.
The scanning and processing in the Y-axis direction are performed by the Y-axis direction scanning unit 515, the continuous black pixel counting unit 516, and the black pixel centroid extracting unit 5.
17, the center of gravity connecting portion 518. The result of each process is
It is shaped by the vector shaping unit 52. The processing in the Y-axis direction and the processing in the X-axis direction are substantially the same except that the scanning directions are different. Here, the Y-axis direction will be described as an example. Y
The scanning of the axial scanning unit 515 is not performed in pixel units, but is performed by skipping some pixels. The inter-scan line width S, which is the interlace width, can be set to any width.

The continuous black pixel counting unit 516 counts how many black pixels are continuous while scanning. Based on the counting result, the black pixel centroid extraction unit 517 extracts the centroids of consecutive black pixels. The center of gravity connecting section 518 connects the black pixel center of gravity extracted by the black pixel center of gravity extracting section 517 to form a vector. If a constant distance is set in advance and the distance between the black pixel centroids is smaller than the constant distance, the two are connected to form a vector. However, if the distance is larger than the certain distance, they are not connected. FIG. 6 is a diagram for explaining a situation in which the extracted centroids are connected to form a vector.
The distance between the black pixel center of gravity A and the black pixel center of gravity C is larger than the fixed distance, and thus is not connected. Therefore,
The vector V3 represented by the alternate long and short dash line is not formed (however, the vectors V1 and V2 may be shaped into the vector V3 at the stage of shaping operation of the clean copy processing unit 52 described later). Similarly, the black pixel centroids D and E are connected to form the vector V4, but the black pixel centroids E and F are not connected to each other and the vector V5 is not formed.

The clear copy processing unit 52 forms a roughly handwritten form by combining or contacting the vectors, deleting or removing erroneous vectors, and correcting the inclination. FIG. 7 shows various cases of vector shaping performed by the fair copy processing unit 52. V6 to V16 are vectors. (A) Deletion of short vector When the length of the vector is less than or equal to a certain threshold, it is determined that this is an error vector formed due to an error in the extraction of the center of gravity, and the vector is deleted. (B) Vector combination If the distance between the end points of two vectors is smaller than a predetermined distance and the angle formed by the two vectors is smaller than a predetermined angle, the process of "vector combination" is performed. This is a process in which a vector is represented by a single vector that connects end points that are separated from each other. Figure 7
A specific example of vector combination is shown in (a). The distance H between the closer end points of the vectors V6 and V7 is smaller than the predetermined distance HT, and the angle θ formed by one vector is the predetermined angle θ.
If it is smaller than T, the two vectors V6 are shaped into one vector, which is a vector V8 connecting between the end points which are separated from each other. (C) Vector contact In the process of "vector contact", when the distance from one endpoint of one vector to the other vector is smaller than a predetermined distance, the endpoint is contacted with the other vector. It is a process of shaping. Specific examples of vector contact are shown in (b) and (c) of FIG. 7. Figure 7
(B) indicates the case where the vectors V9 and V10 intersect, and (a1, b1) to (a4, b4) are the coordinates of the end points of each vector. As shown in the left diagram in (b) of FIG. 7, only K is protruded from the intersection of the short portions of the vector V10. If this K is smaller than the predetermined distance, as shown on the right side of FIG. 7B, the part of the protruding K is cut into a vector V11 and shaped so as to be in contact with the vector V9. To be done. The left diagram of (c) of FIG.
Although it is separated from the vectors V9 and V10, the distance, that is, the vector V10 is extended by the distance M to form a vector V11, which is brought into contact with the vector V9. The coordinates (c1, c2) of the contact point are obtained in the same manner as in the case of (b) in FIG. FIG. 7D illustrates a state in which the vectors are combined and the vectors are combined. The left figure shows the state before vector connection and vector contact, and the right figure shows the state after that. The combination of the vectors V12 and V13 is brought into contact with the vector V14. The resulting vectors are vectors V15 and V16. By the shaping process in the clear copy processing unit 52 as described above,
Erroneous vectors are deleted and gaps between vectors are corrected.

FIG. 8 is a schematic view of the above-described state of the clean copy of the entire figure. That is,
By performing linear approximation on the line graphic image as shown in FIG. 8A, the result as shown in FIG. 8B can be obtained. By subjecting this result to a shaping process to correct the deviation of the end points and the intersections, the result as shown in FIG. 8C can be obtained. When the print document or the form handwritten neatly is input, the clear copy processing unit 52 may detect the fact and omit the execution of this function.

The application processing unit 53 converts the vectorized form into a document of a computer or an input format of each application device (application). As described above, the document form drawn on the paper can be input as the form document that can be handled by the computer.

In the above-described embodiment, only the figure (form) image from the original image is the object of form input, but an image containing only characters is left and character recognition processing is performed on the image to perform character recognition. It becomes possible to fill in forms including. In addition, by adding this character recognition input process, the meaning of each row and each column of the form can be understood, and application to a later application can be facilitated. For example, by knowing the destination column in the form, it becomes possible to obtain statistics for each destination.

[0016]

According to the present invention, by performing the character / graphic processing before the form processing, even if the form has a mixture of characters which cannot be dealt with by the existing form input, the form portion is automatically generated. It can be extracted and input to an application device such as a computer or word processor. Further, according to the present invention, since the fair copy processing is performed on the extracted form portion, it is possible to input even a roughly handwritten form. Therefore, according to the present invention, there is an advantage that the form of the document can be extremely easily input to the application device.

[Brief description of drawings]

FIG. 1 is a block diagram of an embodiment of the present invention.

FIG. 2 shows an example of an image of an input document having a form in which characters and figures are mixed.

FIG. 3 is a diagram showing an example of a separated character image when the image of FIG. 2 is separated by a character / graphic separation process.

FIG. 4 is a diagram showing an example of a figure (form) image separated from the separated original image by the character / graphic separation processing when the image of FIG. 2 is separated by the character / graphic separation processing.

FIG. 5 is a diagram showing a configuration example of a vectorization unit and a clean copy processing unit.

FIG. 6 is a diagram illustrating a situation in which extracted centroids are connected to form a vector.

7A to 7D are diagrams showing various cases of vector shaping performed by the fair copy processing unit.

8A to 8C are diagrams showing a state of a clean copy.

[Explanation of symbols]

1 ... Image input section, 2 ... Image memory, 3 ... Character /
Figure separation unit, 4 ... Form image memory, 5 ... Form processing unit, 51 ... Vectorization unit, 52 ... Clean copy processing unit,
53 ... Applied processing unit.

Claims (1)

Claim: What is claimed is: 1. An image input means for inputting a document image as digital data, an image memory for storing the input image, and a character / graphic separation means for separating a figure and a character in the image. A form input device with a character / graphic separation device, comprising: a clear copy processing unit for performing a clear copy process on a separated graphic; and a form processing unit for performing a clear copy process on the separated graphic and converting it into form information.