Abstract

Edges, corners and vertices are strong and useful features in computer vision. This paper deals with the development of an efficient model based approach in order to detect and characterize precisely these important features. The key of our approach is first to propose some efficient models associated to each of these features and second to efficiently extract and characterize these features directly from the image. The models associated to each feature include a large number of intrinsic parameters (Grey level intensities, location, orientation of the line segments... ) but also an important parameter which is associated to the blurring effect due to the acquisition system. The important problem of the initialization phase in the minimization process is also considered and an original and efficient solution is proposed. In order to test and compare the reliability, the robustness and the efficiency of the different proposed approaches, a large number of experiments involving noisy...

Citations