Extraction and selection of binarised statistical image features for fingerprint recognition

Abstract

Due to their simplicity and efficiency, histogram-based descriptors are very used in the task of fingerprint recognition. In this work, we use a novel histogram based descriptor called binarised statistical image features (BSIF). The experiments have conducted on the standard FVC2002 database. We have extracted the BSIF histograms from sub-images around the core point of the fingerprint image and concatenated them to construct the final features vector. The experiments have shown that an increasing number of extracted sub-images lead to an increasing recognition rate, but lead also to higher dimension histogram which decreased performance of the system regarding computing time and memory capacity. To avoid this problem we have used a feature selection method based on the mutual information called interaction capping (ICAP) which selects the relevant bins of the BSIF histogram. The results showed that using feature selection method could reduce the dimensionality leading to a less computational complexity.