In this paper, we have presented a promising approach of Harris
corner detector for palmprint feature extraction. The corner
information extracted using Harris corner detector is compared
with other feature vector in the database using Hamming
distance similarity measurement method using sliding window
method. It is observed in the results that the sliding window
method takes less time in matching the feature vectors as
compared to match by correlation method. This shows that
sliding window method does not depend on the number of
corners detected. It can be concluded that feature matching by
sliding window method has come out to be best method with
high accuracy and less comparison (matching) time.
Experimental results clearly show that Harris corner detector
methodology has the ability to discriminate similar palmprints
with recognition rate of 97.5%. The high accuracy and less
comparison time make this system rapid, genuine and reliable
authentication system.