Research And Achievement On Algorithm For Human Palm-Dorsa Vein Recognition

The hand vein recognition is a new biological feature recognition, which is developing in recent years. We can carry on the body's identification through the hand vein. Compared with fingerprint recognition, it has many advantages, such as non-contact.This paper based on last extensive technologies in this field, the dissertation is trying to make studies on vein recognition algorithms. Main works are as follows:1.vein image processing: We design and validate the algorithms of removing vein image background, image smoothing; Through experiments, we consider image smoothing as necessary step of vein image processing; Mading a comparison of segmentation effect between particle equality measurement algorithms and the method of threshold image; We analysis the effect of template size on threshold image algorithms; The design of thinning algorithms use the composed mode of middle approaching and classical algorithms; Base on various noise of thinning image, design removing noise method respectively.2.feature extraction and matching: introduce the matching method horizontal moving of points and calculating distance. Aiming at the character of vein image; proposed a feature extraction and matching method based on line angle reference vector.