Authors : S. Jeyanthi, N. Uma Maheswari and R. Venkatesh
Abstract: Each human being has distinctive fingerprints. Speed and accuracy are the prerequisites of a Biometric Authentication System. There are some cases several fingerprints can be merged or overlapped in nature they are in low quality also. These kind of overlapped fingerprints create a complicated problem to recognize the person because of separating images is very monotonous and time consuming task without the help from fingerprint experts. In low quality images feature extraction becomes a serious problem. The proposed research describes the Novel Automatic Fingerprint Recognition System (NAFRS) for overlapping images. This research deals with study of various image processing steps for overlapped images. A novel fingerprint separation algorithm is proposed. The proposed system includes preprocessing, feature extraction, fingerprint classification, recognition and retrieval of information that have been considered extensively. The challenge in the existing system is that in all fingerprint databases only a few fingerprint samples from each individual are available for training and testing. The proposed research is planned to improve the performance for overlapped image in terms of accuracy and speed with the increased number of samples for testing and training. This proposed research research tries to reduce false accept rate and false reject rate for latent images.
S. Jeyanthi, N. Uma Maheswari and R. Venkatesh, 2012. An Efficient Automatic Fingerprint Recognition System for Overlapped Images-Survey. International Journal of Soft Computing, 7: 235-241.