An approach to line feature representation and matching for palmprint recognition

A palmprint is a relative new biometric feature for personal authentication. Palm-lines, including the principal lines and wrinkles, are one of the most important features used in palmprint recognition. This paper proposes a novel approach of line feature representation and matching for palmprint recognition. To represent palm-lines, a vector, called line feature vector (LFV), is defined by using the magnitude and orientation of the gradient of the points on these lines. A LFV contains information about both the structure and thickness of the lines, thus its capability to distinguish between palmprints, including those with similar line structures, is strong. A correlation coefficient is employed to measure the similarity between LFVs of palmprints during the matching phase. 99.0% and 97.5% accurate rates are obtained in the one-to-one matching test and one-to-many matching test, respectively. The results show that LFV is robust to some extent in rotation and translation of the images. The accuracy, speed and storage of the proposed approach can meet the requirements of an online biometric recognition.