Handwritten Character Recognition Using Improved Thinning Algorithm and Fuzzy EX-OR Operator.

• Main Authors: Chen ,Chih-Wen, 陳智文
• Other Authors: Chen, Liang-Chia
• Format: Others
• Language: zh-TW
• Published: 1996
• Online Access: http://ndltd.ncl.edu.tw/handle/04856118623126453330

碩士 === 中原大學 === 電子工程學系 === 84 === In this paper,an improved thinning algorithm and a fuzzy EX-OR operator are presented for handwritten character system. The thinning algorithm is improved the iterative thinning algorithm and parallel thinning algorithm. By using some 9x9 windows to preserve necessary and essential structures. It is good with respect to both connectivity and noise immunity. Futher it is about two times faster than the iterative method. The results show that the improved algorithm overcomes some of the disadvantages found in traditional thinning algorithm. We propose that a new operator classifier by using fuzzy logic theory can be applied efficienty to complex classification problems. The operator classifier hassimple calculation and can reduce running times to obtain better calssify result. It has good result to process the similar characters. The experiment show that the proposed two methods are effective. We present results from the U.S. Postal Services database. The recognition rate is about 99.23% base on our handwritten system.