Title :
Hand-written numeral recognition based on Zernike moment
Author :
Tong, Xiao-jun ; Zeng, Shan ; Zhou, Kang ; Zhao, Kai ; Jiang, Qin
Author_Institution :
Dept. of Math. & Phys., Wuhan Polytech. Univ., Wuhan
Abstract :
The Zernike moment of the handwritten numerals is screened by the random variable correlation coefficient in statistics. We make use of the fuzzy C-mean cluster induction of the hand-written numerals based on the screened Zernike moment. The example testified this method effective, and provides the theory basis for further establishment of the hand-written numeral recognition standard library. In this process, we have also discovered two reasons which cause the fuzzy C-mean cluster to be un efficient.
Keywords :
fuzzy set theory; handwritten character recognition; pattern clustering; Zernike moment; fuzzy C-mean cluster induction; handwritten numeral recognition; random variable correlation coefficient; Application software; Clustering algorithms; Handwriting recognition; Libraries; Mathematics; Optimization methods; Pattern analysis; Pattern recognition; Random variables; Wavelet analysis; FCM cluster; Zernike moment; handwritten numeral recognition;
Conference_Titel :
Wavelet Analysis and Pattern Recognition, 2008. ICWAPR '08. International Conference on
Conference_Location :
Hong Kong
Print_ISBN :
978-1-4244-2238-8
Electronic_ISBN :
978-1-4244-2239-5
DOI :
10.1109/ICWAPR.2008.4635806