Computer Vision - ECCV 90: First European Conference on Computer Vision. Antibes, France, April 23-27, 1990. Proceedings [Paperback]

A collection of papers on computer vision research in Euro-pe, with sections on image features, stereo and reconstruc-tion, optical flow, motion, structure from motion, tracking,stereo and motion, features and shape, shape description,and recognition and matching.On scale and resolution in the analysis of local image structure.- Optimal filter for edge detection methods and results.- Biased anisotropic diffusionA unified regularization and diffusion approach to edge detection.- Hierarchical image analysis using irregular tessellations.- A bit plane architecture for an image analysis processor implemented with P.L.C.A. gate array.- Scale-space singularities.- Recursive filtering and edge closing: Two primary tools for 3D edge detection.- Edge contours using multiple scales.- Stereo integration, mean field theory and psychophysics.- A pyramidal stereovision algorithm based on contour chain points.- Parallel and deterministic algorithms from MRFs: Surface reconstruction and integration.- Parallel multiscale stereo matching using adaptive smoothing.- On the motion of 3D curves and its relationship to optical flow.- Structure-from-motion under orthographic projection.- Parallel computation of optic flow.- Obstacle detection by evaluation of opticalflow fields from image sequences.- Extending the Oriented smoothness constraint into the temporal domain and the estimation of derivatives of optical flow.- A comparison of stochastic and deterministic solution methods in Bayesian estimation of 2-D motion.- Motion determination in space-time images.- Ambiguity in reconstruction from image correspondences.- An optimal solution for mobile camera calibration.- Analytical results on error sensitivity of motion estimation from two views.- On the estimation of depth from motion using an anthropomrphic visual sensor.- The derivation of qualitative information in motion analysis.- Road following algorithm using a panned plan-view transformation.- 3D-vision-based robotlÃ0