Character segmentation method and device

Abstract

The invention provides a character segmentation method and device. According to the method and device, multiple continuous touching characters can be segmented into complete single characters, the condition that a complete character is segmented into halves is avoided, and the accuracy rate of character recognition is increased to a certain extent. The character segmentation method comprises the steps that a to-be-segmented character image is subjected to vertical projection, the middle points of blank spaces are looked up from the projected image to serve as pre-segmentation points, and thena pre-segmentation point set is obtained, wherein the blank spaces are points of which vertical projection values are less than a set value; the average width of single characters is calculated according to the total character number and the total character width; the pre-segmentation point set is traversed, the space between every two pre-segmentation points is calculated, and an actual segmentation point set is determined by combining the average width of the single characters; and the actual segmentation point set is traversed, pixel points between the adjacent actual segmentation points are determined, and then a single character image after segmentation of the to-be-segmented image is obtained.

Description

technical field [0001] The present invention relates to the field of computer technology, in particular to a character segmentation method and device. Background technique [0002] OCR text recognition technology has extensive and in-depth applications in many fields today, such as scientific research, reading, information retrieval, etc. Using OCR text recognition technology, texts and reports can be processed in batches, improving processing efficiency and reducing labor costs. In the OCR text recognition process, it can be roughly divided into: image input---binarization processing---denoising and cutting---character intelligent recognition---result output, among which denoising and cutting is an important link, namely Remove useless noise in the picture, cut the target text in the picture that needs to be recognized in units of characters, each character forms a separate picture, and provides usable materials for the subsequent intelligent character recognition link, th...

AI Technical Summary

Problems solved by technology

[0007] In the process of implementing the present invention, the inventors found that there are at least the following problems in the prior art: In actual business scenarios, only relying on the traditional projection method to solve the denoising and cutting of pictures, the accuracy rate is very low in many cases, because the business The variety and complexity of pictures in the scene

like Figure 4 , after the characters are binarized, some characters in the picture are glued together and cannot be completely separated, so the projected image cannot be divided according to the 0-value point of SumY

Images