Evaluation Of Image Quality And Liveness Detection For Hand Vein Recognition

The21st century is increasingly complex, mobility and frequent interaction of the information age. Information exchange has covered many fields, which also gave birth to the information security requirements. As a "stepping stone", the traditional authentication methods because of its structure is relatively simple, relatively easy to steal, is facing the risk of being imitated. On the other hand, the biometric authentication methods based on physiological biometric characteristics or behavioral traits formed, with its universality, uniqueness, stability, and difficult to copy, is getting a lot of attention. Many researchers are carrying out related work, and some results have been used in the areas of finance, government, and infrastructure. According to relevant survey, biometrics market is an emerging market with maintain a high growth rate. The biometric technology is and will play an important role in the security of personal information.The hand vein as a branch of biometrics is an emerging technology. It mainly bases on the absorption capability of blood to near-infrared light. This technology is currently developing very rapidly in Asia, and there are many commercial products have been launched. In particular, the dorsal hand vein recognition system used in2008Beijing Summer Olympics is a good application. Considering its enormous potential and future development, this paper carried out some related work.The main innovation in this paper as follows:1) The quality of data source determines the performance of the system in some extent. So, combining with problems encountered in the experimental and the future research works, this paper carried out quality assessment for dorsal hand vein image. According to the characteristics of dorsal hand vein image and related work of other biometric methods in this field, this article selects the contrast, effective area, clarity, geometry offset and information entropy as the indicators, and uses the weighted average method to calculate a quantitative score for evaluation the quality of dorsal hand vein image.2) Consider any known security system is faced with the risk of be forge, this paper also did some research works in liveness detection for dorsal hand vein. The author designed and obtained a database containing live and fake dorsal hand vein samples. According to the principle of vein image formed, this article carried out the identification work in both the time domain and frequency domain, and ultimately determine the method of regional block energy in the frequency domain as the resolution for liveness detection.