作者: D.M. Wuescher , K.L. Boyer
DOI: 10.1109/34.67629
关键词: Mathematics 、 Curvature 、 Nonlinear system 、 Grayscale 、 Signal processing 、 Artificial intelligence 、 Algorithm 、 Edge detection 、 Robustness (computer science) 、 Geometry 、 Constant curvature 、 Image processing
摘要: The problem of decomposing an extended boundary or contour into simple primitives is addressed with particular emphasis on Laplacian-of-Gaussian zero-crossing contours. A technique introduced for partitioning such contours constant curvature segments. nonlinear 'blip' filter matched to the impairment signature computation process, overlapped voting scheme, and a sequential contiguous segment extraction mechanism are used. This insensitive reasonable changes in algorithm parameters robust noise minor viewpoint-induced distortions shape, as those encountered between stereo image pairs. results vary smoothly data, local perturbations induce only result. Robustness insensitivity experimentally verified. >