Similarity Computation Based on Feature Extraction for Off-line Chinese Signature Verification

In this paper we present a new method for off-line Chinese signature verification. The approach is based on feature extraction of every segment segmented from the signature image. After preprocessed and segmented, the signature image is segmented to some segments. Every segment is represented by a set of seven features. By using these segments and its features, the degree of similarity between the questioned sample and the n genuine signature samples is computed. Only a little number (3 to 5) of genuine samples for each writer is needed by using a simplified version of question document expert´s classifier. Experimental results show our approach is promising to distinguish random and simple forgeries from genuine signatures effectively.