Suvarna Joshi

Design of multimodal biometrics system based on feature level fusion

Multimodal system aims to fuse two or more biometrics traits of an individual to achieve improvement in FAR and FRR of biometrics system which in turn increases accuracy of system. In this paper we have proposed biometrics system based on biometrics traits face and signature. The performances of face and signature recognition can be enhanced using a proposed feature selection method to select an optimal subset of features. Signature is very important human characteristics which is required in all financial transaction for human identification. In case of financial transaction correct recognition is necessary otherwise it can lead to fraudulent activities. Face is most commonly acceptable and popular biometrics. Proposed algorithm fuses wavelet based features of face and signature. Wavelet based feature fusion method also gave very promising results. Hamming distance classifier is used to take decision whether person is genuine or imposter. Our experiments show that the proposed algorithm can achieve higher classification accuracy than offline signature and face based identification system. We have achieved false accept rate of 5.99% and 3% for multibiometrics system for ORL databases combined with Caltech and Ucoer real signature database resp.

Feature Extraction Using DWT with Application to Offline Signature Identification

Handwritten signature is most widely accepted biometrics for person identification. This paper proposes a novel algorithm for offline handwritten signature recognition. Target of this research is to present signature recognition based on coded wavelet coefficient. It works at global level for extraction of discriminate signature features using wavelet transform. Before extracting the features, preprocessing of a scanned handwritten signature image is necessary to isolate the signature part and to remove any unwanted background present. Wavelet transform has been used to extract features from preprocessed signature images. Wavelet coefficients are extracted from detail part of handwritten signature and further wavelet coefficients are coded. Wavelet coefficient coding results in image compression. This causes reduced feature vector size. Hamming distance has been used to find out distance between test signature pattern and training signature pattern. Experiments are carried on signature database for 56 users each of 24 genuine and 9 skilled forgery signatures. One more experiment is carried out on gathered database. Recognition success rate for genuine signatures is 95 %. FAR of proposed algorithm is about 0.22.