Curvature Estimation Using Shape-from-Texture
This paper shows how the apparatus of robust statistics can be used to extract consistent estimates of surface orientation using shape-from-texture. We make initial estimates of surface orientation by measuring the affine distortion of neighbouring spectral peaks. We show how the initial orientation estimates can be refined using a process of robust smoothing and subsequently used for reliable curvature estimation. We apply the method to a variety of real-world and synthetic imagery. Here it is demonstrated to provide useful estimates of curvature.
KeywordsTilt Angle Texture Surface Spectral Peak Curvature Estimation Surface Orientation
- 1.David Marr. Vision: A computational investigation into the human representation and processing of visual information. Freeman, 1982.Google Scholar
- 2.John Krumm and Steve A. Shafer. Shape from periodic texture using spectrogram. In IEEE Conference on Computer Vision and Pattern Recognition, pages 284–289, 1992.Google Scholar
- 4.John R. Kender. Shape from texture: an aggregation transform that maps a class of texture into surface orientation. In 6th IJCAI, Tokyo, pages 475–480, 1979.Google Scholar
- 5.Jonas Garding. Shape from texture for smooth curved surfaces in perspective projection. J. of Mathematical Imaging and Vision, 2:329–352, 1992.Google Scholar
- 6.J. Malik and R. Rosenholtz. Recovering surface curvature and orientation from texture distortion: a least squares algorithm and sensitive analysis. Lectures Notes in Computer Science-ECCV’94, 800:353–364, 1994.Google Scholar