印刷体汉字识别有着很高的实用价值,主要工作包括图像预处理、字符的定位分割、图像的特征提取和图像识别,并以Visual Studio 2005作为开发环境实现一个完整的基于最短欧氏距离模式匹配的印刷体汉字识别系统.解决了在字符定位分割时出现的"过分离"问题.经测试,识别率可以达到97.6%,基本符合实际应用要求.
The recognition of the printed Chinese characters is of practical value.It mainly concerns the image pretreatment,the location and segmentation of the characters,the extraction of the feature and the image recognition,and developes an entity template-matching system based on the nearest Euclid distance,and resolves the over-split problem.Through the tests,the rate of the recognition can reach 97.6%,and it basically meets the needs of practical application.