Advances in International Computer Science

In the calculation process of rainfall intensity, the appearance of adhesive raindrops will affect the calculation result of rainfall intensity, so we need to segment the adhesive raindrops. In this paper, according to the characteristics of outer convex of raindrops, the raindrop particles are first classified into single raindrop and adhesive raindrop, and then according to the characteristics of adhesive raindrop, the contour of adhesive raindrop is studied. Based on the found concave points, the core operation of the effective segmentation of the adhesive raindrop particles is to find the concave points accurately. The proposed segmentation algorithm based on convex defect concave point, study the adhesion after rain outline, per the above points as the center, establish a square area, and then respectively in each contour mobile establish above the square, then the statistics of each square area within the black spots number of points, the point as a concave point, and according to the rules of the corresponding concave points matching, Finally, the segmentation of adhesive raindrops is realized. Experimental results show that the proposed concave point segmentation algorithm based on convex defects can achieve effective segmentation of adhesive raindrops.

concave point search; Adhesion region segmentation; Convex defect