Off line signature recognition based on contourlet transform
Mina Fakhlai, Hamid Reza Pourreza · 2011
In this paper, we propose new offline Persian signature recognition based on the contourlet transform (CT). We utilize Support vector machine (SVM) as a tool to evaluate the performance of the proposed method. In proposed method, first signature image is normalized by size then image is enhanced to remove noise. After pre-processing, signature image is divided into four regions, contourlet coefficients are computed on each region. Next the histogram of orientation and direction of each region are computed so we have four histograms that are fed to a layer of SVM classifiers as feature vector. Our Persian dataset include 400 genuine images and 200 forgery images. Recognition rate is 98%.