%0 Conference Proceedings
%4 sid.inpe.br/sibgrapi/2017/09.09.02.55
%2 sid.inpe.br/sibgrapi/2017/09.09.02.55.51
%T Multi-Frame super-resolução a partir de uma única imagem por estereoscopia
%D 2017
%A Nacif, Felippe Leitão,
%A Fernandes, Leandro Augusto Frata,
%@affiliation Universidade Federal Fluminense
%@affiliation Universidade Federal Fluminense
%E Torchelsen, Rafael Piccin,
%E Nascimento, Erickson Rangel do,
%E Panozzo, Daniele,
%E Liu, Zicheng,
%E Farias, Mylène,
%E Viera, Thales,
%E Sacht, Leonardo,
%E Ferreira, Nivan,
%E Comba, João Luiz Dihl,
%E Hirata, Nina,
%E Schiavon Porto, Marcelo,
%E Vital, Creto,
%E Pagot, Christian Azambuja,
%E Petronetto, Fabiano,
%E Clua, Esteban,
%E Cardeal, Flávio,
%B Conference on Graphics, Patterns and Images, 30 (SIBGRAPI)
%C Niterói, RJ, Brazil
%8 17-20 Oct. 2017
%I Sociedade Brasileira de Computação
%J Porto Alegre
%S Proceedings
%K super-resolução, lentes estereoscópicas, iterative back projection.
%X Super-resolution (SR) is a set of techniques that estimates a high-resolution (HR) image from one or more low-resolution (LR) images. Single-frame SR usually relies on learn-based methods to enhance a LR image. Multi-frame SR, on the other hand, may not assume previous knowledge of the scene and use information extracted from multiple LR images taken from slightly different viewpoints to estimate a HR view. We present a multi-frame SR approach that uses a stereoscopic camera lens attachment to capture two overlapped views of the scene at the same time with the same image sensor. We adapt the classic Iterative Back Projection (IBP) SR algorithm to estimate a SR color image from two color LR views. We virtually modeled the proposed lens attachment and demonstrate that, by using only two overlapped views, our approach achieves qualitatively better results than other multi-frame SR techniques.
%@language pt
%3 2017___Nacif__Fernandes___SIBGRAPI___WUW.pdf