A new approach to the verification of chinese signatures with variant orientations and scales using relaxation and state-space search methods

摘要: Abstract A signature verification system usually consists of a learning stage and testing stage. Always an assumption is made by most previous studies that the variations in orientation scale signatures done signer two stages should be minimized, otherwise, errors might occur. In this study, we propose new method to solve problem. By treating each as point pattern consisting set critical points, relaxation proposed reference (15) extended first, next best matching pairs are found using state-space search like ∗ algorithm. According results, affine transformation matrix computed finally. applying signatures, variation problem mentioned above thus solved. Experimental results given show superiority method.