Handwritten Signature Verification Technique based on Extract Features

In this paper, we describe a new method to identify the handwriting signature. There are many methods for signature verification, but most of them require segmentation or connected component analysis. A robust Verification technique for identifying and verifying human-signature images is presented. The Verification process utilizes the determinant value that produces the features for the handwritten signature. Image’s determinants values are computed, by divided image into blocks then find centric point to calculate distance. The least square criterion is then utilized to determine the similarity between the existed (in Database file) signatures with a new signature’s images. A threshold value is designed to differentiate between existed and non-existed stored signature’s images.