JPS62192887A - Feature quantity generating method in character recognizing device - Google Patents

Abstract

PURPOSE:To decrease a feature quantity and to realize the improvement of a recognizing ratio and the recognizing speed by unifying a part of the divided block of an input pattern, generating a histogram and designating the block to be unified. CONSTITUTION:An operator inputs unifying instruction data to a host computer 51 by using a keyboard 54. The host computer 51 writes the unifying instruction data into a memory 53. After an OCR processor 52 reads an original 55 and the direction coding and area dividing of an outline part are executed concerning the input pattern, the histogram is prepared. At the time of preparing the historgram, the OCR processor 52 reads the unifying instruction data of the memory 53, decides the block to be unified and unifies the histogram of plural instructed blocks. Thereafter, the OCR processor 52 compares the obtained histogram with the dictionary and determines a candidate character.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field] The present invention relates to a feature generation method in a character recognition device such as OCR.

[Prior art]

One of the character recognition methods in OCR, etc. is to divide an input pattern into multiple blocks by attaching a direction code to its outline, take a histogram for each divided block by its direction code, and use each histogram as a feature value. There is a method for character recognition. In this case, for example, if the input pattern is divided into 4×4 and eight direction codes are used, 4×4×8=128 dimensional feature amounts are extracted.

By the way, some of these feature quantities have high character discrimination ability, while others have low character discrimination ability. However, conventionally, distance calculations have been performed by treating all of these feature amounts in the same way, which has been one of the causes of a decrease in recognition rate and recognition speed.

〔the purpose〕

An object of the present invention is to improve recognition rate and recognition speed in a character recognition device that divides an input pattern into a plurality of blocks, takes a histogram for each direction code for each divided block, and performs character recognition using each histogram as a feature quantity. The aim is to improve the

〔composition〕

The present invention integrates some of the divided blocks to generate a histogram, and also allows the operator to specify the blocks to be integrated, thereby reducing the amount of features and thereby improving the recognition rate and speed. It is something that will be realized. below,
An embodiment of the present invention will be described with reference to the drawings.

First, region division of an input pattern will be explained with reference to FIG. First, a direction code is attached to the outline of the input pattern (step 21).

Next, the direction codes attached to the contours of this input pattern are counted and the total number is determined (step 22). Next, the division coordinates in the X direction and the Y direction are determined based on the total number of direction codes. For example, assuming that the area is divided into nXm, the input pattern is scanned in the X direction using the total number of direction codes, and the X coordinates at which the number of direction codes corresponds to each division point are determined (step 25). Similarly, the dividing point in the Y direction is 111.

m m27). Therefore, scan the input pattern in the Y direction,
The Y coordinate at which the direction code number corresponds to each division point is determined (step 28).

FIG. 1 is a flowchart for explaining feature amount generation according to the present invention.

First, a histogram of each direction code is created for each divided block (step 11).Next, the integrated instruction data from the host processor is read (step 12), and this integrated instruction data is referenced to block the corresponding block. (step 13). FIG. 3 shows an example of dividing an area into 4×4 areas, and FIG. 4 shows an example of blocks to be integrated and integration instruction data in this case.

FIG. 5 is a schematic block diagram of a hardware configuration that implements the method of the present invention. In FIG. 5, a host computer 51 and an OCR processor 52 share a memory 53. The operator uses the keyboard 54 to input integrated instruction data in the format shown in FIG. 4 into the host computer 51. This integrated instruction data is transferred to the host computer 5.
1 is written to the memory 53. On the other hand, OCR processor 52
reads the original document 55, attaches a direction code to the outline and divides the input pattern into regions according to the flowchart shown in FIG. 2, and then creates a histogram according to the flowchart shown in FIG. When creating this histogram, OCR
The processor 52 reads the integration instruction data in the memory 53, determines the blocks to be integrated, and integrates the histograms of the specified blocks. Thereafter, the OCR processor 52 compares each of the obtained histograms with a dictionary prepared in advance as a feature quantity of the input pattern, and determines candidate characters based on the distance.

Note that the blocks to be integrated can be determined by using methods such as multivariate analysis when creating the dictionary in advance and determining the discrimination ability.Also, the operator can change the blocks to be integrated for each character type or font. .

〔effect〕

According to the present invention, an input pattern is divided into a plurality of blocks by attaching a direction code to its outline, a histogram is obtained for each direction code for each divided block, and each histogram is used as a feature when performing character recognition. By merging some blocks to create a feature, the number of features can be reduced, and recognition speed can be expected to improve. Also,
Since the features to be integrated have low dictionary identification ability, it is expected that the recognition rate will improve and the dictionary capacity can be reduced.

[Brief explanation of drawings]

FIG. 1 is a flowchart for explaining the method of the present invention,
FIG. 2 is a flowchart for explaining region division of an input pattern, FIG. 3 is a diagram showing an example of region division, FIG. 4 is a diagram showing an example of integrated instruction data and integrated blocks, and FIG.
The figure is a block diagram of a hardware configuration for realizing the present invention. DESCRIPTION OF SYMBOLS 1...Host computer, 52...OCR processor, 53...Memory, 54...Keyboard. 55...Manuscript.

Claims (2)

[Claims] (1) In a character recognition device that divides an input pattern in which a direction code is attached to an outline into a plurality of blocks, takes a histogram for each direction code for each block, and performs character recognition using each histogram as a feature quantity, A method for generating feature amounts in a character recognition device, the method comprising generating a histogram by integrating parts of divided blocks. (2) A method for generating feature amounts in a character recognition device according to claim 1, wherein an operator arbitrarily specifies the blocks to be integrated.