Indian Journal of Science and Technology

Abstract

Objectives: The essential requirement for a successful hashing method involves two distinct stages projection and quantization. In general, the projection stage is given much importance than the quantization stage. This stage has been concentrated in this paper which has equal importance as projection stage. Methods/Analysis: The using of Manhattan Distance method has been proposed in this paper instead of the widely used Hamming Distance, since it destroys the neighbourhood structure while measuring the similarity between points in the hashcode space. Findings: The problem of destroying the neighbourhood structure that existed in Hamming Distance is overcome by Manhattan hashing. Novelty /Improvement: The outperformance of Manhattan distance compared with Hamming distance has been shown and also, this paper has made an attempt to implement them in our Biocryptosystem to show its efficiency
Keywords: Biocryptosystem, Biometrics, Hamming Distance, Manhattan Hashing, Similarity Measure Functions