Finger vein recognition method based on LBP and Gabor direction Weber local descriptor

The invention relates to the technical field of finger vein image recognition, in particular to a finger vein recognition method based on an LBP and a Gabor direction Weber local descriptor. The method comprises the following steps: 1, inputting a finger vein image I; 2, combining the obtained LBP graph with a direction feature graph obtained by a Gabor filter, constructing a two-dimensional distribution histogram, and outputting an LGWLD feature vector H. According to the method, a uniform mode LBP is constructed to replace a WLD differential excitation operator, so the uniform mode LBP can fully reflect pixel gray scale change information of a local area, and noise interference is reduced; the Gabor filter is used for replacing the WLD gradient operator, the filter can effectively describe the line features of the finger vein image due to consideration of the influence of multiple neighborhood pixels, and experimental results on FV-TJ and FV-USM finger vein databases show that the method provided by the inv