%0 Conference Proceedings
%4 sid.inpe.br/sibgrapi/2015/07.22.13.26
%2 sid.inpe.br/sibgrapi/2015/07.22.13.26.14
%T Análise da acurácia de métodos de descrição 2D para o reconhecimento facial 3D
%D 2015
%A Queiroz, Jonathan Coutinho Luz de,
%A Pamplona Segundo, Maurício,
%@affiliation Universidade Federal da Bahia
%@affiliation Universidade Federal da Bahia
%E Vieira, Thales Miranda de Almeida,
%E Mello, Vinicius Moreira,
%B Conference on Graphics, Patterns and Images, 28 (SIBGRAPI)
%C Salvador, BA, Brazil
%8 26-29 Aug. 2015
%I Sociedade Brasileira de Computação
%J Porto Alegre
%S Proceedings
%K reconhecimento facial, biometria, visão computacional.
%X With the growing need for strict access control of computational systems, face recognition has shown to be viable and effective in authentication systems. In particular, face recognition based on 3D images by converting them to 2D images has been studied due to the reduced influence of factors such as illumination and pose. This approach consists in the application of projection techniques to obtain 2D images, which allows the application of traditional 2D description methods for recognition. The focus of this work is the analysis of the accuracy achieved by these methods in the 3D face recognition. To this end, we compared the obtained results for UGS, PCA, HOG, LBP and LDA descriptors. The main contribution is the observation that robust methods such as HOG, LBP and PCA yield lower performance than UGS, which is much more simple, when used for 3D face recognition. This demonstrates the necessity of developing description methods specifically for this task.
%@language pt
%3 main-ready.pdf