Research Of Image Processing Algorithm Based On Mathematic Morphology

This thesis presents the basic theory, mathematic properties, basic algorithm and the structure element of mathematic morphology as well as its application in image processing.On the one hand, in the field of binary image, the concepts of edge-detection was presented in this thesis, and the differences between the traditional algorithm and the morphological gradient edge-detection algorithm were compared according to the three Canny's edge-detection principles. Then the improved morphology algorithm, the controlled-morphology algorithm, was seriously researched and compared with the original as well as the other algorithms of image processing. At last, the gradient edge-detection and the anti-noise edge-detection algorithm were studied in the field of grayscale image.On the other hand, in the problem of multi-structure elements, the most important multi-scale and omni-direction structure elements were illustrated in this issue. The principles of structure element selection were also deeply studied.In the end, the experiment results showed the feasibility of the complicated morphological filter designed in this thesis.