Optimizing the recognition rates of unconstrained handwritten numerals using biorthogonal spline wavelets
Suzete Correia, J.M. de Carvalho · 2002
In this paper an approach for off-line recognition of unconstrained handwritten numerals is presented. This approach uses the Cohen-Daubechies-Feauveau (CDF) family of biorthogonal spline wavelets as a feature extractor for absorbing local variations in handwritten characters and a multilayer cluster neural network as classifier. Experiments with the bases CDF 2/2, CDF 2/4, CDF 3/3 and CDF 3/7 were performed using the handwritten numeral database of Concordia University of Canada. The results show that CDF biorthogonal wavelets yield a performance improvement of 2.4% in numeral recognition, compared to the results obtained with the Haar wavelets.