![]() | Only 14 pages are availabe for public view |
Abstract Accurate automatic person identification is critical in a variety of applications in our electronically interconnected society. Biometrics, which refers to identification based on physical or behavioral characteristics, is being increasingly adopted to provide positive identification with a high degree of confidence. Among all the biometric techniques, fingerprint-based authentication systems have received the most attention because of the long history of fingerprints and their extensive use in forensics. Fingerprints were one of the first forms of biometric authentication to be used for low enforcement and civilian applications. Contrary to popular belief and despite decades of research in fingerprints, a reliable fingerprint is still an open problem. In this thesis, we present some contribution to advance the state of the art methods in this field. First, quality enhancement of fingerprint images is important for a good performance of an Automatic Fingerprint Identification System (AFIS). So a hybrid fingerprint enhancement algorithm is presented. This algorithm is based on morphological enhancement in the additive wavelet transform domain and Wave Atom denoising. The performance of the proposed algorithm has been evaluated on a set of images. The results are compared with the results obtained using morphological enhancement only. Our proposed algorithm has given a better performance than using morphological enhancement only. The second contribution of the thesis addresses the problem of fingerprint identification. It presents a new fingerprint identification method based on Mel Frequency Cepstral Coefficients (MFCCs). This approach is based on treating the problem as a pattern recognition problem. Cepstral features are extracted from a group of fingerprint images, which are transformed first to 1-D signals by lexicographic ordering. MFCCs and polynomial shaping coefficients are extracted from these 1-D signals to form a database of features, which can be used to train a neural network. In the testing phase of this neural network, fingerprint identification is performed. The last contribution of the thesis presents a study for the sensitivity of the proposed cepstral approach for fingerprint recognition to synthetic pixels obtained through interpolation. The objective of this study is to allow fingerprint database compression through decimation. |