Off-line arabic signature recognition and verification

Off-line signature recognition and verification is an important part of many business processes. It can be used in many applications such as cheques, certificates, contracts and historical documents. In this paper, a system of two separate phases for signature recognition and verification is developed. A recognition technique is developed based on a multistage classifier and a combination of global and local features. New algorithms for signature verification based on fuzzy concepts are also described and tested. It is concluded from the experimental results that each of the proposed techniques performs well on different counts.