JPH07120393B2 - Character recognition / graphic processing device - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a character recognition / graphic processing apparatus, and more particularly to graphic processing and character recognition that require fine line density such as seal matching.
The present invention relates to a character recognition / graphics processing apparatus suitable for simultaneous scanning of a single sheet by scanning once.

[Conventional technology]

When performing character recognition and graphic processing on one form, it is a big problem how to read the image data of the form and at what binarization level.

That is, for character recognition, considering the processing speed and the recognition rate, a fine line density is not necessary (for example, about 8 dots / mm is enough), and the binarization level is low, that is, It is good to capture at a level that does not erase the characteristics of the characters.

On the other hand, in the graphic processing, the finer the line density is, the better the binarization level is to some extent to capture fine lines clearly.

As described above, conventionally, the following method has been used for the two contradictory conditions.

(1) A method of allocating image data to character data and image data according to area (Japanese Patent Laid-Open No. 60-581) (2) Multiple levels of binarization can be set, and any level can be set according to the target form (3) Although there is only one line density for one form, the line density can be changed according to the form (JP-A-59-81773). ) [Problems to be Solved by the Invention] However, as described above, when character recognition and graphic processing are performed on one sheet, considering the processing speed and the recognition rate, character recognition is not so much. Fine line density is not necessary, and it is better to capture at a low binarization level, that is, at a level that does not erase the characteristics of characters. In graphic processing, finer line density is better and 2
The digitization level should be high to some extent to capture fine lines clearly.

On the other hand, it is conceivable that image data having different linear densities and binarization levels can be obtained by one scanning (reading), but in each of the conventional methods described above,
In either case, there is a problem that it is not possible to obtain image data having different linear densities and binarization levels by scanning once for one sheet.

The present invention has been made in view of the above circumstances, and an object of the present invention is to solve the above-mentioned problems in the conventional character recognition / graphic processing apparatus and to meet the respective purposes of character recognition / graphic processing. The object of the present invention is to provide a character recognition / graphic processing apparatus capable of performing accurate character recognition / graphic processing by reading the image data of the entire surface such as a form to be processed in one scan.

[Means for solving problems]

The above object of the present invention is to provide an image sensor having a constant scanning line density, an analog amplifier circuit for amplifying pixel information read by the image sensor, and a preset for performing graphic processing, which is connected to the analog amplifier circuit. A second binarization circuit having a different binarization level, and a second binarization circuit connected to the analog amplifier circuit and having a level different from the binarization level for performing character recognition processing, The second 2
A linear density conversion circuit connected to the binarization circuit for converting the linear density of the image sensor into another linear density;
And a first DMA circuit for transferring the graphic processing image data of the binarizing circuit to the image memory, and for transferring the character recognition image data of the linear density converted by the linear density conversion circuit to the image memory. And a second DMA circuit.

[Operation] In the present invention, the image input means is a first image input means having a first linear density and a binarization level necessary for character recognition to input an image at a relatively low level,
And a second image input unit having a second linear density and a binarization level for image input at a high level necessary for graphic processing, and both image input units are operated at the same time, The image information necessary for character recognition and the image information necessary for graphic processing are obtained by one scanning.

〔Example〕

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

In this embodiment, a 16-dot / mm sensor is used, character recognition processing is performed at a linear density of 8 dots / mm, and seal matching processing is performed at 1
An example of a character recognition / graphics processing device that performs a linear density of 6 dots / mm will be described.

FIG. 1 is a block diagram of a character recognition / graphics processing apparatus showing an embodiment of the present invention. In the figure, 11 is a central processing unit (hereinafter referred to as "CPU") for controlling the system, 12 is a memory for storing programs and data of the CPU 11, 1
Reference numeral 3 denotes an image memory for storing binarized image data for seal matching, and 14 denotes an image memory for storing binarized image data for character recognition. Further, 15 is a scanner capable of simultaneously reading two types of image data, as will be described later, 16
Is a file device that stores a reference registered imprint pattern, 17 is a CRT display that displays the character recognition result, the seal matching result, the imprint pattern, etc., 18 is a keyboard for the operator to input commands, and 19 is a bus Shows the line.

FIG. 2 is a block diagram showing a detailed structure of the scanner 15. In the figure, 41 is an image sensor having a linear density of 16 dots / mm, 42 is an analog amplifier circuit, 43 is a first binarization circuit for graphic processing, and 44 is a second binarization circuit for character recognition. , 45 is a linear density conversion circuit for converting the linear density from 16 dots / mm to 8 dots / mm, 46 is a DMA circuit for transferring graphic processing image data to the image memory 13, and 47 is for character recognition A DMA circuit for transferring image data to the image memory 14 is shown.

Here, the binarizing circuits 43 and 44, for example, when the output range of the analog amplifier circuit 42 is 0 to 1v, the binarizing level of the binarizing circuit 43 for graphic processing is 0.7v, and the character recognition is performed. An arbitrary binarization level is set according to the purpose so that the binarization level of the binarization circuit 44 for use in the system is set to 0.4v. Also,
The line density conversion circuit 45 has, for example, a buffer for two scans, transfers data to the DMA circuit 47 every other dot in the main scan direction, and transfers only one scan out of two scans in the sub scan direction to the DMA circuit 47. By transferring the data to the circuit 47, the linear density is converted from 16 dots / mm to 8 dots / mm.

Here, 8 dots / mm and 16 dots / mm as shown in FIG.
An operation example in the case of having two mm buffers (13 and 14) will be described.

FIG. 4 is a processing flowchart of the CPU 11 showing an outline of the operation of the apparatus of this embodiment configured as described above.

The operation of this embodiment will be described below with reference to FIGS.

This processing is described by a program stored in the memory 12 and executed by the CPU 11.

The CPU 11 uses, for example, I
By inputting D number processing start, the form 51
Is instructed to read (process 21), and after reading, image data of the entire surface of the form of 16 dots / mm in the designated area of the image memory 13 and 8 dots in the designated area of the image memory 14 are read.
The image data of the entire surface of the / mm form is stored as shown in FIG.

The character recognition process is performed by the image memory 14 in the pre-process 22.
Cut out only one character from. At that time, noise removal and normalization processing may be performed at the same time. In the extracted character pattern, the feature amount is extracted in the feature extraction processing 23, and the dictionary 25
The contents read from are sequentially collated in the collation process 24, and the final result is given by the decision process 26.

These processes are performed for the recognition target characters of one form, and the recognition result is displayed on the display 17 (process 33).

On the other hand, in the seal recognition process, the pre-processing 27
The collation imprint pattern is cut out from 13 and the center position is detected (processing
28). Further, the registered imprint pattern is read from the file 16 (process 29), stored in a specific area of the image memory 13, and the center position is detected (process 30).

In the rotation angle alignment process 31, the rotation angles of the collation imprint pattern in which the center position is detected and the registered imprint pattern are matched, and both imprints are superimposed in the overlay process 32 and displayed on the display 17 (process 33).

In the center position detection processing 28, 30, for example, the center of the circumscribed rectangle of the imprint frame may be calculated, or the center of gravity may be calculated. That is, it is a process of obtaining a reference of a position that does not depend on the inclination. The rotation angle matching process 31 is based on the correlation of the peripheral density distributions about the center, or is a method of simply rotating one of the imprints to find the point having the maximum similarity.

As shown in Fig. 3, 8 dots / mm and 16 dots / mm
With two buffers, the processing efficiency can be improved by the processing described below.

First, using the image data of the buffer 14 of 8 dots / mm,
The area around the form (the shaded area) is detected, and the address on the form memory is detected. Here, the ID on the form is recognized, and the character recognition is performed according to the read ID information and the parameters given in advance, according to the 8 dot / mm buffer 1 described above.
Characters are cut out from each field of 4 and processed.

Further, the image is converted into the address of the buffer 13 of 16 dots / mm by calculation using the address of the form position on the memory obtained by the buffer 14 of 8 dots / mm described above,
Similarly, according to the ID information, the imprint image is cut out according to the parameters given in advance, and the collation process is performed.

As described above, the processing such as the position detection of the form can be shortened by considering the processing time and using the buffer having the coarse linear density.

In the above-mentioned embodiment, the seal stamp collation is mentioned as an example of the graphic processing, but the present invention should not be limited to this, and in general, graphic processing and image processing which require a different linear density from character recognition. It can be applied to

In the above embodiment, 16 dots / mm and 8 dots / mm
The case where image data of two surfaces having different linear densities from mm is taken in by one reading is shown, but it goes without saying that this linear density can be an arbitrary value depending on the purpose.

The density conversion method is not particularly limited, and it goes without saying that the image sensors may be individually provided.

Furthermore, by using a sensor with different dropout colors or a multicolor sensor, and by loading the image data into each image memory, even if it is on the same field, printing or filling can be done according to the dropout color. It is also possible to recognize different characters or images by changing the color of the characters.

Similarly, depending on the purpose of use, for example, image data with different gradations, such as one-level multi-valued image data of 16 levels and the other binary data using a dither pattern,
It is also possible to take in each image memory.

It is also possible to take in one as normal dot image data for character recognition or graphic processing and the other as image data compressed for data transfer and storage into the image memory.

〔The invention's effect〕

As described above, according to the present invention, the image input means has the first image input having the first linear density and the binarization level for performing image input at a relatively low level necessary for character recognition. Means and a second image input means having a second linear density and a binarization level for high-level image input necessary for graphic processing, and both of these image input means operate simultaneously. Since the image information necessary for character recognition and the image information necessary for graphic processing are obtained by one scan, a character recognition / graphic processing apparatus capable of performing accurate character recognition / graphic processing is provided. It has a remarkable effect that it can be realized. In particular, it is effective when character recognition and character image reading are performed on characters on the same field, such as displaying an image of an unread character or an erroneously read character.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram of a character recognition / graphic processing apparatus showing an embodiment of the present invention, FIG. 2 is a block diagram showing a detailed configuration of the scanner 15, and FIG. 3 is a line. FIG. 4 is a diagram for explaining an operation example when two buffers having different densities are provided, and FIG. 4 is a processing flowchart in the apparatus of the embodiment. 11: CPU, 12: Main memory, 13: Graphic processing image memory,
14: Image memory for character recognition, 15: Scanner, 17: CRT display, 18: Keyboard, 41: Image sensor, 43: Binary circuit for graphic processing, 44: Binary circuit for character recognition, 45: Linear density Conversion circuit, 46, 47: DMA circuit, 51: Form.

 ─────────────────────────────────────────────────── ─── Continuation of front page (56) Reference JP-A-58-38063 (JP, A) JP-A-60-65662 (JP, A) JP-A-59-47672 (JP, A) JP-A-59- 156069 (JP, A) JP 54-76007 (JP, A) Actual development 55-58768 (JP, U)

Claims (1)

[Claims] 1. An image sensor having a constant scanning line density, an analog amplifier circuit for amplifying pixel information read by the image sensor, and connected to the analog amplifier circuit,
A first binarization circuit having a predetermined binarization level for performing graphic processing, and connected to the analog amplification circuit,
A second binarization circuit having a level different from the binarization level for performing character recognition processing, and a linear density different from the linear density of the image sensor connected to the second binarization circuit. To a linear density conversion circuit, a first DMA circuit for transferring the graphic processing image data of the first binarization circuit to an image memory, and a linear density character converted by the linear density conversion circuit. A character recognition / graphics processing apparatus comprising: a second DMA circuit for transferring recognition image data to an image memory.