On Analysis of Multi-dimensional Features for Signature Verification
J. Mahmud, Chowdhury Mofizur Rahman · 2006
This paper aims to verify offline signatures using improved feature analysis and artificial neural network. Feature analyzer can reduce the large domain of feature space and extract invariable information. We incorporated different features from multi-dimensional feature analysis perspective. For verification from extracted features, we used neural network classifier. Instead of using feed forward neural network, multiple feed forward neural networks are used which are trained in the form of ensemble. Using such ensemble makes the system more general than a regular single neural network based system. Use of resilient back propagation for each neural network training, provides faster recognition. Using cross validation techniques, we performed significant amount of testing. Experimental evaluation of the signature verifier is reported