Based On The Mel Frequency Cepstrum Coefficient Palmprint Recognition Research

In recent years, with the development of scientific technology and the electronic products, the high-tech electronic products gradually increase in people’s daily lives and production activities. On many occasions, it is necessary to certify and recognize the identity of the people. Traditional ways of certification and confidentiality can not meet the requirements of the people any more in assurance certification security. Thus, the biometric technology of with the help of individual biological characteristics to recognize identity is quietly emerging. Biometric identification technology is difficult to copy and counterfeit, so it can be a very effective solution to safety certification.The biometric identification technology is considered to be the best security authentication technology in the modern society and has been applied to access control, e-commerce, financial, military, and so on. Due to the high recognition rate of biometric technology and broad market prospects, it has aroused wide concern and become a hot research topic in the area of safety certification.Palmprint recognition is an emerging biometric authentication technology in recent years. Compared with the iris, fingerprint, face and other biometric technology, it has many advantages:such as, ease of collection, the stability of main characteristics, obviousness, easy acceptance to people, convenience and low cost. Palmprint recognition technology has become an economic and practical recognition technology.Main contents as follows:1. It discusses the disadvantages of the traditional identification technology, and introduces the development history, research status and assessment criteria of the biometric identification technology2. It discusses the palmprint image methods of acquisition and the key technology of palmprint image preprocessing and features extraction.3. Based on many classifical feature extraction algorithms, this paper proposes a new method of palmprint feature extraction. It takes the Mel Frequency Cepstral Coefficients as features of palmprint images4. It introduces in detail the principle of BP neural network, and constructs BP neural network classifier to recognize above mentioned palmprint features. It obtains satisfactory experiment results.