Wrist and Palm Vein pattern recognition using Gabor filter

Abstract

Wrist and palm vein pattern can be considering as a promising biometric technique for identification, through the study of the pattern of blood vessels that visible from the skin. This kind of recognition is very important for many reasons; vein exists inside of the human body makes it difficult to change pattern like shift the position of vein from part to another, unlike another method of techniques of recognition. In this paper work wrist and palm vein are studied for identification and verification, this work divided into three phases preprocessing, features extraction and recognition. in preprocessing phase apply resize and image “enhancement” using “CLAHE and 2-D Gaussian high pass filter”, the features of each image are extracted by using Gabor filters. LDA and PCA are used to minimize the dimension of the features set. For vein image, identification used Euclidean distance to measure the similarity. The average CRR of vein palm in proposed work is 94.49% and the average CRR of vein wrist is 92.33.