A Novel Fingerprint Identification Fuzzy System Using a Center-Distance Weighted Local Binary Pattern

Despite the rapid development of the automated fingerprint identification system, some challenging fields need more enhancement. Typically, fingerprints suffer from distortion, partial cuts, and noise, making the identification process uncertain and more probabilistic. In this research, we aim to improve the fingerprint identification system using the well-known image-based local binary pattern method based on assigned weights determined by the distance from the center area of the processed fingerprint image. The proposed Fuzzy Fingerprint Identification System (FFIS) starts with image enhancement using Fourier domain analysis, then the image is cut into 200 × 200 around the core-area, after that, the feature vectors of the local binary images are extracted and matched based on the distance of their histograms. Finally, a fuzzy approach is used to retrieve the suitable linguistic form to help clear the uncertainty of the results. The proposed Fuzzy Fingerprint Identification System (FFIS) showed its efficiency through extensive experiments conducted on the FVC2002 database.