Segmentation of fingerprint images using the gradient vector field

In this paper, the properties of the fingerprint image gradient are discussed and their application to segment the fingerprint image is presented. The algorithm proposed here uses the magnitude of gradient and the variance of gradient vectorpsilas direction. The method takes advantage of the two prominent properties of fingerprint, i.e., the clarity of ridge to valley and the slow change of fingerprint texture orientation. This technique can not only segment the fingerprint images but also measure the quality of sub-block or overall fingerprint segmented. This algorithm is tested for a large number of fingerprint images with the variation of quality, and is found to be effective and robust. The results of segmentation of fingerprint images using our method proposed here are presented.