Signature recognition by segmentation and regular line detection

This paper presents a method of identifying handwritten signature by a good algorithm capable of identifying a signature with high accuracy. Here first a signature image is segmented and then data is extracted from individual blocks. After data is collected then accuracy is measured from both numbers of segments along with regular line matching coefficient. A mathematical formula is used to perform the accuracy measurement. A test signature is compared with a standard signature that is in the database. As the level of accuracy is determined from both number of segments and similarity of lines in these segments, reliability of this method is very high. It has the capability to identify even a skilled forgery.