Research On Acquisition And Quality Assessment For Hand Vein Image

With the rapid development of information technology, and the increase of information exchange, it has put forward higher requirements of the protection of personal information security, especially in personal identification technology. Human vein recognition is a new non-contact biometric identification technology, with the uniqueness, living, internal characteristics and other characteristics of the human vein, it has eliminated the possibility of fraud or plagiarism. So human vein recognition has a broad development prospect in security and authentication field.In this paper, with the research object of hand vein, with the target of acquiring a high quality image, aim at the low image quality, affected by the light of the surrounding environment in vein recognition, we make a in-depth research in design of image acquisition devicese and image quality assessment method, and establish a set of infrared acquisition, infrared light source control and image quality assessment of vein image acquisition system. the hand vein image by collected can meet the needs of follow-up treatment.In the image acquisition, due to influence of the thickness of the hand, obvious levels of vein, the epidermis and pore characteristics, we have designed to the light module on the image acquisition, made the illuminance shot on hand equidistribution. According to PWM control technology, we make the PT4115 regulator constant current source to drive the group of near-infrared LED to output different duty cycles in a certain period of time with the use of PWM waveform generated by SCM, which acquires hand vein images under different light intensities.According to the statistical characteristics and structure characteristics of the image, this paper proposes to take the vein image feature as quality assessment indexes. The selected features describe the amount of information contained in the image from different angles, and fuse the different parameters to evaluate image quality together. It has selected the duty cycle range initially, for preparing to output the best vein image next step.This paper combine image quality assessment method with the human visual system, make the use of Contourlet transformation, propose hand vein image quality assessment algorithm based on statistical models with Contourlet transformation. This paper make Contoulet decomposition of vein image under multi-scale and different directions, acquire high and low frequency characteristics. According to Contourlet coefficients this paper has established statistical model, analysed and extracted the maximum image information of the different scales and directions of the vein images under different duty cycles. We take it Contourlet reconstruction, to extract the best quality image at the full extent.