Informatica

This study uses the r-theta transformation technique to map a fingerprint image to the straight-line signals. Subsequently, the “vector magnitude invariant transform” technique is applied to them to generate an invariant magnitude for person identification. This technique can solve the image rotation problem. Various vertical magnitude strips are generated to deal with the image-shifting problem. The algorithm developed in this study can precisely classify the fingerprint images.

This paper presents a method of minutiae based fingerprint matching that is robust to deformations and does not do fingerprint alignment. It concentrates on comparing rotation and translation invariant local structures defined by minutiae point and its neighboring minutiae points. Then the collection of most probable correspondences of matched minutiae is found. Finally, the local structures of higher order are validated. All three steps are completely rotation and translation invariant, robust to nonlinear deformations and do not use any fingerprint alignment. Experimental results on publicly available as well as internal databases show an improved performance of the proposed method in comparison with the traditional minutiae based algorithms that perform fingerprint registration.