JPH07152856A - Optical character reader - Google Patents

Abstract

PURPOSE:To deal with an optical reader in the state of coexisting plural existing business forms and documents which are not designed for the optical character reader. CONSTITUTION:This optical character reader is provided with a business form registering part 16 which registers a partial image that can prescribed the type of a form as an identification (ID) image used for identification of the form type out of the images detected from the form to be processed and also registers the information accordant with the form type and used for reading the characters in response to the ID image, a business form identifying part 22 which identifies the form type used for reading characters based on the ID image registered by the part 16, a character reading part 26 which reads the character according to the identifying result of the part 22 and based on the business form information accordant with the form type registered by the part 16, and a data processing part 30.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an optical character reader for reading existing forms and documents which are not designed for optical character readers.

[0002]

2. Description of the Related Art Conventionally, the following method has been used in an optical character reading apparatus to handle a plurality of types of forms (or documents). First, there is a system in which the types of forms are manually sorted in advance and the types are read by an optical character reader.

Secondly, a character (ID number) is printed (or handwritten) at a fixed position on the form, the character is first recognized, and the type of the form is discriminated according to the recognition result and read. It is a method to do. Generally, a so-called ID processing function is known in which an ID number for identifying a form is printed at a fixed position on the form and the difference is automatically determined by reading the ID number. in this case,
Only forms for optical character readers designed to record ID numbers in fixed positions are processed.

[0004]

As described above, in the conventional optical character reader, in order to handle a plurality of types of forms,
It was necessary to sort by person and process each type, or to record the ID number at a fixed position on the form.

By the way, in general, existing forms which are not designed for optical character readers are widely used. There are many types of these existing forms.
Then, there is a demand for performing character reading processing on these existing forms by an optical character reading device.

However, as described above, in the first method, since there are many types of existing forms, the work load for classifying them becomes very large. Also, the second
In this method, since the ID number cannot always be recorded at the fixed position, the optical character reader cannot be used.

The present invention has been made in view of the above points, and an optical character reading device capable of handling a plurality of existing forms and documents which are not designed for an optical character reading device in a mixed manner. The purpose is to provide.

[0008]

SUMMARY OF THE INVENTION The present invention relates to an optical character reading device for reading characters recorded on a sheet, which is capable of prescribing the type of the sheet from an image detected from the sheet to be processed. Registering a specific image as an identification image for identifying the type of paper, and registering information corresponding to the type of paper used for character reading in association with the identification image, and the registration means. Based on the registered identification image, identification means for identifying the type of paper to be subjected to character reading and information corresponding to the type of paper registered by the registration means according to the identification result by the identification means are provided. And a reading unit for reading characters by using the reading unit.

[0009]

According to this structure, if there is a partial image of which the type can be specified on a sheet such as a form, if the partial image (ID image) is registered, this image is also displayed. And the paper type is automatically identified. Then, it becomes possible to read characters using various information referred to for reading characters according to the type of paper. Therefore, even a form that is not designed for an optical character reading device can be processed, and even if a plurality of types are mixed, the form can be read without performing a work such as classification.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing the configuration of an optical character reading device according to an embodiment of the present invention. As shown in FIG. 1, the optical character reading device includes an image reading unit 10, a display unit 12, an input unit 14, a form registration unit 16, an ID image identification dictionary 18, a form information table 20, a form identification unit 22, and form-specific information. The search unit 24, the character reading unit 26, the input form information storage unit 28, and the data processing unit 30 are configured.

The image reading unit 10 optically scans a form (input form) to be processed by an image scanner or the like, and detects image data. The display unit 12 is an interface with the user, and the image reading unit 1
The image detected by 0 is displayed, and the operation message to the user is displayed.

The input unit 14 has a mouse, a keyboard and the like, and inputs various instructions from the user. The form registration unit 16 includes a partial image (hereinafter referred to as an ID image. Details will be described later) that can define an existing form to be processed, and various information necessary for character reading (hereinafter, form information). (Referred to as) is associated and registered.

The ID image identification dictionary 18 includes a form registration unit 16
The ID image of the form registered by is stored as identification information (dictionary) for identifying the type of the form to be processed.

The form information table 20 includes a form registration unit 16
Is for storing the form information corresponding to each ID image. The form identification unit 22 is the image reading unit 1.
With respect to the image data detected by 0, the type of the form to be processed is identified based on the ID image registered in the ID image identification dictionary 18.

The form-specific information retrieval unit 24 is a form identification unit 22.
The corresponding form information is searched from the form information table 20 according to the type of the form identified by.
The character reading unit 26 reads characters from the image data detected by the image reading unit 10, using the form information retrieved from the form information table 20 by the form-specific information retrieval unit 24.

The input form information storage unit 28 is for storing the form information referred to by the character reading unit 26 and the data processing unit 30 retrieved from the form information table 20 by the form-specific information retrieval unit 24. . The data processing unit 30 performs various processes on the character reading result by the character reading unit 26.

Next, the operation of this embodiment will be described.
First, before starting the actual character reading process, the form registration unit 16 registers an ID image and form information for a form to be processed. The registration processing by the form registration unit 16 will be described with reference to the flowchart shown in FIG.

For example, the image data of the form to be registered is detected by the image reading unit 10 and the form image is displayed on the display unit 12. The form registration unit 16
The user is instructed by the input device 14 (mouse or the like) of the area of the displayed form image to be the ID image.

The form registration unit 16 cuts out the ID image data from the image data based on the position data indicating the range of the ID image area input from the input unit 14 (step S1). Specifically, as the ID image, a portion unique to the form type such as a picture, mark, logo, character, or graphic provided on the existing form is selected. For example, when the form image as shown in FIG. 3 is displayed, the portions shown in A and B in the drawing are designated as ID images. The form registration unit 16
Position data indicating the range of the input area and the size of the ID image are fetched (step S2).

Next, the form registration unit 16 instructs the area (reading position) to be the target of character reading in the form image. That is, when the character reading process is performed, the position where the character image to be processed is to be cut out from the form image is designated. The form registration unit 16 inputs the position data indicating the reading position input from the input unit 14 and registers it in the form information table 20 (step S3).

Here, according to the format of the form,
A plurality of reading positions can be designated for one form image. For example, the reading position is designated for each item (address, name, telephone number, etc.) provided on the form. Next, the form registration unit 16 determines the processing definition for each reading position (each item) (character recognition information such as the character type, font, and character spacing referred to in the character recognition processing, and data such as logical check). (Including information for processing) is input and registered corresponding to the reading position of each item (step S4).

The form registration unit 16 normalizes the size of the ID image obtained in step S1 and further creates a dictionary used for recognition by, for example, the composite similarity method, and forms registered in the form information data 20. It is registered in the ID image identification dictionary 18 in association with the information (step S5).
S6).

As described above, the registration process by the dictionary registration unit 16 is performed for each form type to be processed, and the ID image (identification dictionary for recognizing the form) and the form information are registered respectively.

Next, the case of executing the character reading process will be described. First, the image reading unit 10 detects form image data by optically scanning the form to be processed. At this time, the form is previously ID
If the image is registered, a plurality of form types may be mixed.

The form identifying unit 22 uses the dictionary for ID image recognition registered in the ID image identifying dictionary 18 for the form image detected by the image reading unit 10.
It is identified whether the form currently being processed includes an ID image and which form type the ID image indicates.

The form-specific information retrieval unit 24 is a form identification unit 22.
The form information corresponding to the form type is searched and retrieved from the form information table 20 and stored in the input form information storage unit 28 in accordance with the identification result by the.

Next, the character reading unit 26 uses the information for character recognition of the form information stored in the input form information storage unit 28, and the information is input for each reading position corresponding to each item of the form. The character image is cut out from the form image and the character recorded in the form is recognized by referring to the character recognition dictionary.

The data processing unit 30 uses the information for data processing of the form information stored in the input form information storage unit 28 for the character recognition result by the character reading unit 26,
Various processes are performed to generate the final character reading result.

In this way, by registering a partial image capable of defining the type of form as an ID image according to the type of form to be processed, it is not designed for an optical character reading device. Even if it is, it is possible to read characters. Further, if the ID image is registered, the character reading process can be realized without problems even if a plurality of form types are mixed.

In the above embodiments, the composite similarity method is used for recognizing the ID image, but other recognition methods can be used. In this case, the form registration unit 16 creates an ID image identification dictionary according to the recognition method and registers it in the ID image identification dictionary 18.

Further, the information included in the form information is not limited to the one described above, and any information can be registered in association with the ID image for identifying the form type. Further, although the form has been described as an example in the above-described embodiment, the form can be applied to a form other than a form such as a document as long as it includes a partial image that can be an ID image.

[0032]

As described above, according to the present invention, a partial image capable of defining the type of paper is registered as an identification image for identifying the type of paper, and characters are read in association with the identification image. Since information corresponding to the type of paper used for recording is registered, it is possible to handle a plurality of existing forms and documents that are not designed for an optical character reader in a mixed manner.

[Brief description of drawings]

FIG. 1 is a block diagram showing the configuration of an optical character reading device according to an embodiment of the present invention.

FIG. 2 is a flowchart for explaining a processing procedure in the form registration unit 16 in the present embodiment.

FIG. 3 is a diagram showing an example of an ID image used in this embodiment.

[Explanation of symbols]

10 ... Image reading unit, 12 ... Display unit, 14 ... Input unit,
16 ... Form registration unit, 18 ... ID image identification dictionary, 20 ... Form information table, 22 ... Form identification unit, 24 ... Form-specific information retrieval unit, 26 ... Character reading unit, 28 ... Input form information storage unit, 30 ... Data Processing unit.

Claims (1)

[Claims] 1. An optical character reading device for reading characters recorded on a sheet, a partial image for which the type of the sheet can be defined is selected from the images detected from the sheet to be processed. Based on the identification image registered by the registration means, which is registered as an identification image for identifying, and which is used for reading characters in association with the identification image, and which registers information according to the type of paper. hand,
Identification means for identifying the type of paper to be read, and reading means for reading characters using information corresponding to the type of paper registered by the registration means according to the identification result by the identification means. An optical character reading device comprising: