Novo algoritmo de segmentação e realce de imagens de impressões digitais

• Ano de defesa: 2018
• Autor(a) principal: Ramos, Ramon Celestino
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal da Paraíba; Brasil; Informática; Programa de Pós-Graduação em Informática; UFPB
• Palavras-chave em Português: Biometria; Impressão digital; Realce; Segmentação; Extração de características; Biometrics; Fingerprint; Enhancement; Segmentation; Feature extraction; CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: https://repositorio.ufpb.br/jspui/handle/123456789/13344

Biometrics provides a reliable authentication mechanism using physical or behavioral traits to identify users based on their natural characteristics. Fingerprint recognition is one of the most used biometrics approach, since its high accuracy and low cost make the system more affordable and acquire satisfactory results. However, fingerprint recognition is still an open problem, since false acceptance and false rejection errors are still found in matching algorithms. This research aims to create new methods in order to facilitate the feature extraction process, improving image quality through the use of inovative segmentation and enhancement techniques, seeking to reduce error rates and achieve competitive results among state-of-the-art algorithms. The main contributions of this work were the creation of the following methods: segmentation of the region of interest of fingerprint images, reaching error rates lower than the best segmentation algorithms in the world in 10 of 12 databases evaluated, obtaining an average gain of 5.6% over the best current segmentation algorithm; creation of the complete feature extraction method, based on state of art works, making corrections and innovations at key points to obtain better results. The feature extraction method was evaluated through the submission of the algorithm to the Fingerprint Verification Competition (FVC), obtaining promising results, being classified as the second best algorithm between research groups and the only algorithm of Brazilian origin. In addition, when the comparison with the base enhancement algorithm of this work was done, the method developed in this research obtained gains of 21% in accuracy.