A Chinese-character stroke extraction based on contour information

• Main Authors: Wu, Bohom, 吳柏宏
• Other Authors: Lee, Chung Nan
• Format: Others
• Language: zh-TW
• Published: 1996
• Online Access: http://ndltd.ncl.edu.tw/handle/45411719320395398769

碩士 === 國立中山大學 === 資訊工程研究所 === 84 === In this thesis a new stroke extraction algorithm is proposed. The algorithm is based on the contour information of characters. In the proposed algorithm, the contour is first extracted, then the dominant corner points are detected. Third, the character structure is extracted by the contour information and a modified cross section graph (CSSG). The purpose of the CSSG is to find the end and junction regions of characters. Finally, the Bezier curve is used to evaluate the continuity of strokes.Experimental results show that the proposed algorithm can extract the strokes correctly based on the human perception. The algorithm can avoid the distortion of thinning algorithms and gives a better results than the direction method. This research provides a solid basis for the structure matching of Chinese characters.