JP2000020639A - Optical character reader - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce operator's correction work by recognizing the images of a document by using format control information for the front and rear sides of the document and printing out the recognized result on the document read. SOLUTION: This optical character reader is provided with a scanner 2 for reading out images from the front and rear sides of a document and outputs from the scanner 2 are respectively stored in a front image storing part 11 and a rear image storing part 12. A field information extraction part 50 finds out the reading field information of the front and rear sides of the document by document analysis and a front/rear decision control part 60 extracts and prepares front format control information and rear format control information from format control information 40. A recognition part 20 reads out the front and rear sides of the document by using the front and rear format control information. A display device 80 displays the read results and a printing part 70 prints out identification characters indicating, whether or not characters to be corrected are included in the document on the document.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a system for recognizing characters on both sides of a form and generating format control information necessary for obtaining code data that can be processed by a computer. Front) Create format control information for both sides of a form without changing the system that creates format control information only for the side, and without making the operator aware of the front and back of the form and the type of size. The present invention relates to an optical character reader that prints the results of reading the front and back of a form on a form while simultaneously reading both sides.

[0002]

2. Description of the Related Art In a conventional optical character reader, accurate character cutout information is indispensable for reading characters written on various slips and forms at a high recognition rate, and the information is format controlled. Information (hereinafter referred to as FCI). FCI for reading both sides of a form simultaneously
Is table format control information (hereinafter FF)
CI) and back format control information for the back side (hereinafter BFCI). The conventional optical character reading device realizes reading of both sides of a form while switching the two format control information for the front side and the back side prepared in advance for each side of the form. As an example of the above, there is a technique disclosed in Japanese Patent Application Laid-Open No. 7-182443.

[0003]

However, the above-mentioned conventional optical character reader has the following problems to be solved. A conventional optical character reader can create a system that can create only the FFCI of only the surface of a form.
It was necessary to change so that FCI and BFCI could be created. In addition, the operator is aware of the front and back of the form and
FCI and BFCI must be created. In addition, when reading forms with different sizes or forms with the front and back of the forms mixed together, the type of form and the symbol for distinguishing between the front and back of the form must be specified in advance. It is necessary to print on a form, or to create one FFCI and BFCI for one type of form, and there is a problem that work efficiency is poor because a lot of labor is required. In addition, there is no means to print the reading result such as whether there is unread or misreading on the back of the form, and it is necessary to check which form and which form has the character to be corrected on each form. However, there is a problem that work efficiency is poor because a lot of labor is required for confirming the correction of the form.

SUMMARY OF THE INVENTION An object of the present invention is to solve the above-mentioned conventional problems, without changing a conventional system that can create only FFCIs, and changing the type of form size and the symbols for distinguishing front and back. Even if the front and back FCIs of multiple types of forms are created as the front FCI of one form without printing on the form in advance, the input can be made without the operator being aware of the front / back and size of the form. An object of the present invention is to provide an optical character reading device that automatically creates FFCI and BFCI for each side of a completed form and can read both sides of the form.

Another object of the present invention is to provide a form that requires correction of read characters, a form that does not require the form, and an identification character having a meaning to identify whether the form to be corrected is a front side or a back side. An object of the present invention is to provide an optical character reading device that can print on a form.

[0006]

In order to achieve this object, an optical character reader according to the present invention comprises an upper surface sensor for scanning the upper surface (front surface) of a document and a lower surface for scanning the lower surface (back surface) of the document. A scanner with a sensor for
And an image buffer for storing the front and back form images obtained by the sensor for the lower surface, a memory for storing the front and back images of the form from the image buffer, and extracting information on both-side read fields of the form from the form image. Means, a memory for storing field information of the front and back surfaces of the extracted form, format control information necessary for reading characters on the form, and format control using the field information of the front and back surfaces of the form. Means for extracting and creating the format control information of the front side of the form and the format control information of the back side of the form from among the information, a memory for storing the format control information of the front side and the back side of the created form, Means for recognizing and processing the image of the form stored in the memory using the format control information on the back side, And means for printing on took form, and means for displaying the read contents, and wherein the scanning both sides of a document at the same time.

Further, the optical character reading apparatus according to the present invention converts format control information of the front and back sides of a form from the extracted field information of the front and back sides of the form without using format control information defined by the operator. It is characterized by having means for creating.

Further, the optical character reading apparatus according to the present invention provides a format control information for a front side of a form once prepared and a format control information for a back side so as to accurately read even forms mixed upside down. Without providing a means for switching the format control information, each time the form is read one by one, the format control information for the upper surface and the format control information for the surface facing the upper surface for each side of the inserted form are It is characterized by having means for creating.

Further, the optical character reading apparatus according to the present invention includes means for reading both sides of one form in advance and creating two format control information for the front and back sides of the form from the format control information. It is characterized by having.

An optical character reading apparatus according to the present invention is provided with a form for determining whether a read character needs to be corrected and a form for which the read character is not required, and whether the surface to be corrected in the form to be corrected is a front side or a back side. Means for printing an identification character having the meaning described in (1).

Further, the optical character reading apparatus according to the present invention can accurately read both sides of a form even when forms having different sizes and forms in which the front and back sides of the forms are mixed are mixed. Then, without printing in advance the types of forms with different sizes and the symbols for distinguishing the front and back of the forms on the form, the form is converted from the format control information for each side of the input forms with different sizes. It is characterized by comprising means for extracting and creating the format control information for the front side and the format control information for the back side every time each page is read.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, an optical character reader according to an embodiment of the present invention will be described in detail with reference to FIG. FIG. 1 is a configuration diagram of an optical character reading device according to an embodiment of the present invention. A scanner 2 for reading the front and back images of a form is provided, and its output is stored in a front image storage unit 11 and a back image storage unit 12 of the recognizing device 1 via an image buffer 5, and then field information is extracted. The form analysis is performed by the section 50 to obtain information on the read fields on the front and back sides of the form, and the field information on the front and back sides of the form is stored in the front field information storage section 14 and the back field information storage section 15. Using the field information generated as a result of the analysis, the front and back determination control unit 60 extracts and creates the front format control information and the back format control information from the format control information 40, and uses the front and back format control information. The recognition unit 20 reads the front and back of the form. The reading result is displayed on the display device 80, and the printing unit 70 prints on the form an identification character having a meaning indicating whether or not there is a character that needs to be corrected in the form.

In the scanner 2, the upper surface sensor 3 scans the upper surface (front surface) of the form to acquire the form data. Further, the lower surface sensor 4 scans a surface (back surface) opposite to the surface of the form scanned by the upper surface sensor 3 to acquire the form data. This makes it possible to read both the front and back sides of the form. The acquired form data is stored in the front image storage unit 11 and the back image storage unit 12 via the image buffer 5.

The format control information 40 includes field information (position, size) for all fields to be read in the form as shown in FIG. In FIG. 2, X and Y are position coordinates having the origin at the upper left end of the form, L and H represent the width and height of the field to be read, and W represents the width of the character frame of the read character.

The field information extracting unit 50 performs form analysis to obtain a field frame from the form image, and extracts information (position, size, etc.) of the field to be read in the form as shown in FIG. The extracted field information on the front side of the form (hereinafter, referred to as FFI) is stored in the front field information storage unit 14, and the field information on the back side of the form (hereinafter, referred to as BFI) is stored in the back field information storage unit 15.

The front / back determination control unit 60 converts the FFI to the FFI
FFCI and BFCI are extracted and created according to the flowcharts shown in FIGS. next,
The processing of the front / back determination control unit 60 will be described with reference to the flowcharts of FIGS.

First, a field is extracted from the FFI,
The field is set to field A (S1). Next, a field is extracted from the FCI, and the field is set as a field B (S2). Field B is F
If the field is the first field extracted from the CI, the field B is set to the field C (S3, S4). Otherwise, go to the next step. If field B extracted from FCI is more similar to field A than field C, field B is newly set in field C (S5, S6). If not, proceed to the next step. It is determined whether the field B is the last field of the FCI (S7). If it is the last field,
The field C is stored in the FFCI (S8). Otherwise, retrieve the next field B from the FCI and repeat the operation. Next, it is determined whether the field A is the last field of the FFI (S9). If it is the last field, go to step A, otherwise, FFI
Then, the next field A is taken out from the above, and the above operation is repeated. Up to the step A, the FCI field for each field on the front surface of the form is obtained, and the FFCI is created.

A field is extracted from the BFI, and the field is set as a field D (S10). next,
The field is extracted from the FCI, and the field is set to the field E (S11). Field E is FC
If it is the first field extracted from I, field E is set to field F (S12, S13).
Otherwise, go to the next step. If the field E extracted from the FCI is more similar to the field D than the field F, the field E is newly set in the field F (S14, S15). If not, proceed to the next step. It is determined whether the field E is the last field of the FCI (S16). If it is the last field, the field F is stored in the BFCI (S1).
7). Otherwise, fetch the next field E from the FCI and repeat the operation.

Next, it is determined whether the field D is the last field of the BFI (S18). If it is the last field, the process is terminated; otherwise, the next field D is extracted from the BFI and the above operation is repeated. Until the processing of the last field of the BFI is completed in step S18, the field of the FCI for each field on the back side of the form is obtained, and the BFCI is created.

By performing the above processing, it is possible to create FFCI and BFCI for each form.

The similarity determination in the above operation is performed by obtaining the absolute value of the difference between the pieces of information of the field information (position and size) shown in FIG. 2, and the fields having the smallest difference value are the most similar. Judge that there is.

The recognizing unit 20 uses the FFCI to store the image of the front side of the form stored in the front image storage unit 11, and the BFCI to store the back side image of the form stored in the back image storage unit 12. Character recognition is performed by a known technique, and both sides of a form can be read.

As a result of reading the form, the printing unit 70 prints a meaningful identification character for determining whether or not there is a character to be corrected in the form on the form. FIG. 5 shows an example of a printing method. FIG. 5A shows a case where a character to be corrected is included only on the surface of the form. In this case, an identification character "A" is printed on the surface of the form. (2) is a case where a character to be corrected is included only on the back side of the form. In this case, the identification character "B" is printed on the surface of the form. (3) is a case where the form includes characters to be corrected on both the front side and the back side.
In this case, an identification character "AB" is printed on the surface of the form. (4) is a case where the form does not include a character to be corrected. In this case, an identification character "O" is printed on the front surface of the form.

The printing position, direction, identification character, and printing surface are not limited to the example shown in FIG. The printing position can be at any position on the form, and the printing direction can be vertical, horizontal, or oblique. The identification characters are printable and identifiable characters.
Either side is possible.

Therefore, according to the present invention, it is possible to create a plurality of types of forms without changing the conventional system that can only create format control information for the front side of a form, and without the operator being aware of the type of front and back and size of the form. Even if all the fields of the form are defined as the fields on the surface of one form, it is possible to read both sides of a plurality of types of forms even if the fields for reading the form are freely defined.

Conventionally, in the case where forms of different sizes are mixedly read, one format control information has to be created for one type of form. Even when forms with different sizes are mixed, the format control information of each form can be combined and created as one format control information, so that the operator does not need to be aware of the front and back and size of the form, Creation work can be performed easily.

Further, according to the present invention, even in the case where a form in which the front and back sides are upside down is mixed, the form is not provided with a control means for switching between the format control information for the front side and the format control information for the back side of the form once created. Each time the document is read, the format control information for the upper surface of the form and the format control information of the surface facing the upper surface are extracted and created from the format control information 40 by the front / back determination control unit 60 in accordance with each side of the form. Therefore, both sides of the form can be read using the two format control information.

Further, according to the present invention, one form is read in advance, and the format control information 40 is used by the front / back determination control unit 60 to determine the format control for the top of the form according to each side of the read form. The information and the format control information of the surface facing the upper surface are extracted and created, and the two format control information can be used to continuously read both sides of a plurality of forms.

According to the present invention, the field information extraction unit 50 uses only the FFI and BFI obtained by analyzing the form from the form image without using the FCI defined by the operator necessary for reading the form. The front / back determination control unit 60 creates format control information for the front side of the form from FFI and format control information for the back side of the form from BFI, and uses the two format control information
It is possible to read both sides of a form.

Further, according to the present invention, the operator can easily determine the form to be corrected and the surface to be corrected in the form based on the identification characters printed on the form surface. Therefore, the display device 80 can select a screen to be immediately corrected from the identification characters printed on the form. This makes it possible to improve the work efficiency of the operator's correction confirmation.

Further, according to the present invention, even when forms of different sizes or forms with the front and back inverted are mixedly read,
The format control information 40 is determined by the front / back determination control unit 60 in accordance with each side of the input forms having different sizes without printing in advance the types of the forms having different sizes and the symbols for distinguishing between the front and back of the forms. Thus, the format control information for the upper surface of the form and the format control information for the surface facing the upper surface can be extracted and created, so that both sides of the form can be read using the two format control information.

The present invention is not limited to the form OCR of the present embodiment. The present invention can be implemented by a text reader type OCR or a form image input system.

[0033]

As described above, the optical character reading apparatus of the present invention utilizes field position information and size information obtained by form analysis of both sides of a form obtained by scanning with a scanner. The front and back judgment control unit extracts and creates the format control information for the front side and the back side of the form from the format control information freely defined by the operator without being aware of the front and back and size of the form. By using the two pieces of information, it is possible to read characters on both sides of a form even when forms having different sizes and forms having the opposite sides are mixed. Therefore, without changing the conventional system that can create only the format control information on the front side of the form, and without the operator being aware of the front and back and the type of size of the form, it is as if all Since the fields to read the form can be freely defined as if the fields are the fields on the surface of one form, even if forms of different sizes are mixed, one format control information can be created. Thus, the operation of defining the reading field can be performed smoothly, and the labor of the operator can be greatly reduced. Further, the identification characters printed on the form allow the operator to easily find the form to be corrected and the surface to be corrected in the form, thereby improving the work efficiency of the correction confirmation.

[Brief description of the drawings]

FIG. 1 is a diagram showing an embodiment of an optical character reading apparatus for double-sided reading according to the present invention.

FIG. 2 is a diagram showing an example of field position and size information according to the present invention.

FIG. 3 is a part of a detailed flowchart of processing of a display determination control unit of the present invention.

FIG. 4 is a part of a detailed flowchart of processing of a display determination control unit according to the present invention;

FIG. 5 is a diagram illustrating an example of printing of identification characters on a form by a printing unit according to the present invention.

[Explanation of symbols]

1. Recognition device 2. Scanner 3.
... Sensor for upper surface, 4 ... Sensor for lower surface, 5 ... Image buffer, 10 ... Memory, 11 ... Table image storage unit,
12: Back image storage unit, 14: Front field information storage unit, 15: Back field information storage unit, 17: Front format control information storage unit, 18: Back format control information storage unit, 20: Recognition unit, 30: Control unit , 4
0: format control information, 50: field information extraction unit, 60: front / back determination control unit, 70: printing unit,
80 ... Display device.

Continued on the front page (72) Inventor Masatoshi Kurata 2880 Kozu, Odawara-shi, Kanagawa F-term in Hitachi, Ltd. Storage System Division (Reference) 5B029 BB02 CC13 CC26 DD10

Claims (10)

[Claims] An operator required to read characters on a form stores format control information without being aware of the front and back sides, a scanner for reading images on the front and back sides of the form, and an output from the scanner. Image buffer, a memory for storing front and back images of a form from the image buffer, a unit for extracting field information of a field from the image, and a memory for storing field information of the front and back sides of the extracted form Means for creating two format control information of the front and back sides of the form using the field information of the front and back sides of the form from the format control information, and a memory for storing the format control information of the front and back sides of the created form When,
Means for recognizing and processing the form image stored in the memory using the format control information on the front and back sides of the form, means for printing on the read form, and means for displaying the read content; An optical character reader for reading simultaneously. 2. A system according to claim 1, wherein both sides of the form are simultaneously read using format control information of the front and back sides of the form prepared using only the field information of the front and back sides of the form without using the format control information. An optical character reader as described. 3. A symbol for distinguishing between front and back sides of a form is not printed on the form in advance, and each side of the entered form is
Each time both sides of a form are read, the front-side format control information is read from the front-side format control information without a means for switching between the format control information for the front side and the format control information for the back side prepared in advance. Means for creating two types of format control information, one for the front and the other for the back side. 2. The optical character reader according to claim 1, wherein both sides of the form are read simultaneously. 4. A buffer for separately storing front and back images of a form, means for extracting field information of fields from the separately stored images, and a method for storing the extracted front and back field information of the form. A memory for separately storing, means for creating two format control information of the front and back of the form using the field information of the front and back of the form separately stored from the format control information, A memory for separately storing front and back format control information, and means for recognizing and processing the form images stored in the separate buffers using the separately stored front and back format control information of the form. 2. The optical character reading device according to claim 1, further comprising: reading both sides of the form simultaneously. 5. A means for preliminarily reading both sides of one form and preparing two format control information for the front and back sides of the form from the format control information, wherein the two format control information thus prepared are provided. 2. The optical character reading apparatus according to claim 1, wherein both sides of a plurality of types of forms are read at a time by using. 6. As a result of reading each form, a form is required to correct the read character, a form not requiring the correction, and a form to determine whether the surface to be corrected in the form to be corrected is a front surface or a back surface. 2. A device according to claim 1, further comprising means for printing an identification character.
An optical character reader as described. 7. The format control information is input to each side of a form having a different size without printing in advance a form for determining the type of the form having a different size or a symbol for distinguishing the front and back of the form. Means for extracting and creating the format control information for the front side and the format control information for the back side each time the form is read, and using the two format control information thus created. 2. The optical character reading apparatus according to claim 1, wherein even when forms having different sizes or forms having opposite sides are mixed, both sides of the form are simultaneously read accurately. 8. Even if the format control information of a plurality of types of forms is collectively created as one format control information, each time a plurality of types of forms are read out one by one from the one format control information. 2. The optical character reading device according to claim 1, further comprising means for creating format control information for the front side and the back side, and using the created format control information to simultaneously read both sides of the form. 9. A method for determining the types of forms having different sizes and symbols for distinguishing the front and back of the forms from the format control information for each side of the forms having different sizes without printing them in advance. An optical character reading apparatus comprising means for extracting and creating format control information for the front side and format control information for the back side each time a form is read one by one. 10. Even when the format control information of a plurality of types of forms is collectively created as one format control information, each time a plurality of types of forms are read out one by one from the one format control information. An optical character reader having means for creating format control information for the front side and the back side.