Efficient finger vein image feature extraction method and system

The invention provides an efficient finger vein image feature extraction method and system, and the method comprises the following steps: 1, obtaining a pixel dispersion distribution diagram containing main line information through employing a pixel dispersion algorithm described by a local mean value and a standard deviation; step 2, performing multiple times of feature extraction on the obtained pixel dispersion distribution diagram by using a Kirsch filtering kernel by using a pixel dispersion extraction algorithm fused with main line information to obtain a pixel block combination; wherein the movement mode of the Kirsch is from left to right and from top to bottom; step 3, traversing all pixel blocks in the pixel block combination to obtain a global weight and a binary sequence; and step 4, calculating GLDP codes according to the obtained binary sequence and the global weight, and combining the obtained GLDP codes as pixel values of corresponding positions to finally obtain a feature image. According to