摘要: We report on a refinement of our technique for determining the orientation textured surface from two-point autocorrelation function its image. \Ve replace previous assumptions isotropic texture by knowledge moment matrix ofthe when viewed head on. The is then deduced effects foreshortening these moments. This applied to natural images planar surfaces and gives significantly improved results anisotropic textures which under assumption isotropy mimic projective foreshortening. potential practicality this method higher level image understanding systems discussed. *This work was supported in part DARPA grant # N00039-84-C-0165.
columbia.edu
sci-hub.st 参考文章(8)
John Ronald Kender, Shape from Texture ,(1986)
Yu-ichi Ohta, Kiyoshi Maenobu, Toshiyuki Sakai, Obtaining surface orientation from texels under perspective projection international joint conference on artificial intelligence. pp. 746- 751 ,(1981)
Michael J. Swain, John Aliomonos, Shape from texture international joint conference on artificial intelligence. pp. 926- 931 ,(1985)
Mark Laurence Moerdler, Shape from textures: a paradigm for fusing middle level vision cues Department of Computer Science, Columbia University. ,(1990) , 10.7916/D883415Q
Ruzena Bajcsy, Lawrence Lieberman, Texture gradient as a depth cue Computer Graphics and Image Processing. ,vol. 5, pp. 52- 67 ,(1976) , 10.1016/S0146-664X(76)80005-6
Ken-ichi Kanatani, Detection of surface orientation and motion from texture by a stereoogical technique. Artificial Intelligence. ,vol. 23, pp. 213- 237 ,(1984) , 10.1016/0004-3702(84)90010-9
Larry S. Davis, Ludvik Janos, Stanley M. Dunn, Efficient Recovery of Shape from Texture IEEE Transactions on Pattern Analysis and Machine Intelligence. ,vol. PAMI-5, pp. 485- 492 ,(1983) , 10.1109/TPAMI.1983.4767427
Lisa Gottesfeld Brown, Haim Shvaytser, None, Surface orientation from projective foreshortening of isotropic texture autocorrelation IEEE Transactions on Pattern Analysis and Machine Intelligence. ,vol. 12, pp. 584- 588 ,(1990) , 10.1109/34.56194