Fingerprint Recognition Techniques

Abstract

Convex hull is a popular structure in computational geometry. It’s a useful tool in constructing structures like Voronoi diagrams, and we use this structure for fingerprint recognition. This paper utilizes the Graham’s scan algorithm for computing convex hull in a specified space which is tested for its reliability, performance, uniqueness and stability. Here we analyze and compare the time, space complexities with the K-NN clustering for better performance. The Graham’s scan algorithm produces an accurate complexity of O(n log n) despite the size of the hull. The key concept is to pre process the input and uses the order to help decide which points to pick. This work also contributes to the pattern recognition and cluster orientation.