The Effect of Image Compression in the Classification of Multimodal Biometrics

A computer-based application is recognition of Biometrics. It determines the authentication for different persons. In this paper the biometrics images are taken for feature extraction and compare it with the registered CASIA Face Image Database, CASIA Fingerprint Image Database, CASIA Fingervein Image Database and CASIA Iris Image Database. In this paper we applied different techniques for biometrics recognition and they were as follows (Principal Component Analysis (PCA) for face recognition - Gray Level Co-Occurrence Matrix (GLCM) Feathers will extraction for fingervein recognition - Minutiae Features will extraction for fingerprint recognition - Gabor filter Features will ex-traction and Hamming Distance for Iris recognition). We then applied the image compression technology using Discrete Cosine Transform (DCT) technology to all databases available for each biometric and then applied the above techniques for biometrics recognition. This was done to determine the effect of image compression on recognition and accuracy rates. The purpose of this paper is to improve the recognition rate of biometrics in the system. The major goal of this study is to make easy implementation and to enhance the efficiency, better recognition rate and high speed by applying Support Vector Machine (SVM).