A Combined Feature Extraction Model Using SIFT and LBP for Offline Signature Verification System
Bhushan S. Thakare, Hemant R. Deshmukh · 2018
Increasing demand of security application in realtime scenario has grown rapidly during last decade. In this field, bio-metric applications are considered as most promising technique for user identity verification and identification. Other that bio-metric systems, signature based user verification systems are also used widely in various application such as banking systems use signature based transaction. This process of verifying signatures is called as offline signature verification. Several techniques have been introduced in this field but the performance of verification depends on the significant feature extraction technique. To deal with this issue, here we presented a combined approach for feature extraction where SIFT (Scale Invariant Feature Transform) and improved LBP (Local Binary Pattern) are combined together to obtain the robust feature model. furthermore, a classification study is also presented to measure the performance of proposed approach. Experimental study shows that proposed approach obtains promising performance for signature verification system.