Epipolar Geometry Estimation and Its Application to Image Coding

This paper presents image-based rendering related techniques; (1)a computer vision technique to estimate accurate epipolar geometry and (2) a n image coding technique to compress multiple-viewpoint images, which are stored for reconstruction of any arbitrary view of 3-0 object, with the epipolar geometry. Uhlike other epipolar geometry estimation techniques based o n the correct sets of feature point correspondence, which i s hardly given in practice, the proposed method minimizes absolute luminance diflerence between two images with respect to seven epipolar geometry parameters. I n the framework of block matching, the absolute luminance difference between the two images can be represented by the seven parameters with which a nonlinear robust regression follows. W e apply the epipolar geometry estimation technique to multiple-viewpoint image coding. That practically contributes to improving the coding eficiency, where more than 1000 viewpoint images are coded by a type of predictive coding based on epapolar geometry. The experimental results show that the compression ratio attains to 1/100 and more.