Edge detection by selection of pieces of level lines
Enric Meinhardt-Llopis · 2008
We propose an edge detector based on the selection of well contrasted pieces of level lines, following the proposal of Desolneux-Moisan-Morel (DMM) [1]. The DMM edge detector has the problem of over-representation, that is, every edge is detected several times in slightly different positions. In this paper we propose two modifications of the original DMM edge detector in order to solve this problem. The first modification is a post-processing of the output using a general method to select the best representative of a bundle of curves. The second modification is the use of Canny's edge detector instead of the norm of the gradient to build the statistics. The two modifications are independent and can be applied separately. Elementary reasoning and some experiments show that the best results are obtained when both modifications are applied together.