Character Segmentation System Based on C# Design and Implementation

Abstract At present, most of the OCR recognizing through individual character, thus the quality of character segmentation is the key point to affect the quality of OCR recognition system. This paper introduces the formula of projective method in analysis of preliminary segmentation for images. Moreover it applied analysis for connected spatial domain, the correct results shows that writing image well matched. After two analyses and segmentation, characters can be segmented correctly. In order to provide useful solutions to these two problems that characters keying must be performed rapidly and documents digitizing can be conserved for a long time. Therefore, we must place emphasis on the research and development of the character segmentation.