Edge detection based on mathematical morphology theory

摘要

Edge detection is an essential pre-processing step in image segmentation and object recognition. Conventionally, mathematical morphological edge detection methods adopted a single structure element so that they are difficult to extract complex edge features. In this paper, a novel edge detection algorithm about multi-scale and multi-structure elements morphological is proposed. First, the noise can be filtered by an alternative-order, morphological open-close filter with a multi-scale element. Then, we gained different edge detection results by using a multi-structure element morphological algorithm, and the final edge result was obtained by using a synthetic weighted method. The experimental results showed that the proposed algorithm can filter noise successfully and it was efficient for complex border detection.