Efficient person identification by Finger-Knuckle-Print based on Discrete Cosine Transform Network

Abstract

One of the current trends in human identification is the development of new emerging methods. Due to increased security concerns and the development of counterfeiting techniques. This development depends on the unique parts of the human body that can be identified and used as a means of identifying a person. Including fingerprints, iris, lips, etc. Most of the systems and methods are slow or require expensive technical equipment, for this, we suggest a new approach for personal authentication using Finger-Knuckle Print through with a novel texture descriptor, Discrete Cosine Transform Network (DCTNet) and support vector machine (SVM) classifier. Fingerknuckle- print is one of the emerging biometric traits.Recently it has been found FKP is highly rich in textures and can be used to uniquely identify a person. The study also takes the unimodal and multi-modal biometric systems results along with their methods of information fusion in score level, which does not require special equipment and can be used in systems where fast detection is needed. Our methods significantly out performs stateof the art methods.