Chinese character framework inflection point detection methods and systems based on eight-neighborhood direction coding

Abstract

The invention discloses two Chinese character framework inflection point detection methods and systems based on eight-neighborhood direction coding. The first measurement method comprises a preprocessing step, a stroke segment segmentation step, a stroke segment coding step and an inflection point detection step. The second measurement method comprises a preprocessing step, a stroke segment segmentation step, a stroke segment coding step, an inflection point judgment step and an inflection point detection step. According to the methods, unique rules of direction coding chains of different strokes of Chinese characters are fully considered, research and summarization are carried out on the basis of eight-neighborhood direction coding, main direction coding, legal direction coding and illegal direction coding of different types of stroke segments are given, a stroke segment corresponding to a direction code chain containing three or more different direction codes is determined to be a stroke segment containing inflection points, and handwriting pixel points corresponding to illegal direction codes in the direction code chain are determined to be inflection points.

Description

technical field [0001] The invention belongs to the technical field of information processing of Chinese character writing, and more specifically relates to a method and system for detecting an inflection point of a Chinese character skeleton based on eight-neighborhood direction coding. Background technique [0002] There are many kinds of strokes of Chinese characters, and the writing of different people is also quite different. These factors affect the accuracy of machine handwritten Chinese character recognition and evaluation. However, we can use information such as the number and location of inflection points to confirm the stroke type and stroke structure of the current stroke, effectively improving the accuracy of handwritten Chinese character recognition and evaluation. [0003] The existing inflection point detection methods include detection methods based on edge pixel curvature and multi-neighborhood changes of pixel points. These methods are easily disturbed by ...

AI Technical Summary

Problems solved by technology

[0012] Aiming at the defects of the prior art, the object of the present invention is to provide a Chinese character skeleton inflection point detection method and system based on eight-neighborhood direction coding, aiming to solve the problem of using the unique law of the direction coding chain of different strokes of Chinese characters to detect the inflection point of handwritten Chinese characters question

Images