Research On Hand Vein Recognition Algorithm Based On Origin Static Moment Vector

With the advent of the information age.information security has become increasingly important.there are many occasions in which the personal authentication is required.There is no doubt that biometrics is one of the most important and effective solutions for this task.As a new biometric recognition,vein recognition has great advantages compared to most current biometric recognition technologies.However,the research of hand vein recognition technology is at the starting stage,so it is very necessary and significant to study the havd vein recognition.In this thesis,some key techniques of hand vein recognition have been investigated deeply, several novel and effective algorithms have been proposed such as image enhancement,feature extraction and feature matching etc.The main work in this thesis call be summarized as follows:(1) The issue on the contrast of gray-scale between hand vein texture and background is studied. Firstly,the region of interest is extracted on the collected image. a method is used to enhance image based on the combination of high-frequency emphasis filtering and histogram equalization. The experimental results demonstrate the target vein of image is clear after image enhancement. The fuzzy index is used to express the degree of improvement for image enhanced. Finally, some threshold segmentation algorithm and the thinning algorithm is investigated.(2) The feature extraction and feature matching of hand vein image is explored. A novel characteristic is extracted using hand vein image of thinning. Firstly, the region of interest is divided into multiple sub-image.Secondly the average static moment of origin is extracted for all sub-image.all average static moment of origin is combined into a feature vector.finally, the two-stage matching is used to match the hand vein pattern utilizing eigenvector and amount of junction.Experimental results show that this method is valid and feasible.