Multimodal biometric system for Personal identification using Palmprint and finger knuckle print Fusion techniques

Abstract

Biometric based identification systems are becoming very popular newlinein recent years since security is the main concern in day to day affairs It is an newlineart of automatically identifying persons based on their physiological and or newlinebehavioural characteristics When only a single trait is used in a biometric newlinesystem it suffers from problems like noisy data intra and inter class newlinevariations non universality user acceptance and spoof attacks A multimodal newlinebiometric system provides solution to these problems by combining newlineinformation from multiple sources In multibiometric system information newlinefusion need not necessarily combine different biometric traits but also from newlinemultiple sensors multiple instances multiple algorithms and multiple sample newlinesystems newlineIn this research work two biometric traits palmprint and finger newlineknuckle print which has high user acceptance reliability and stability are newlineconsidered The images in the data base are first preprocessed to extract a newlineregion of interest ROI which can be used for feature extraction and newlinematching This step serves to improve the quality of the images and to extract newlineonly useful information and also reduce the influence of variations resulting newlinefrom the image data acquisition process newline newline