Adhesion Ore Image Separation Method Based on Concave Points Matching

In order to segment adhesion ore image well, the adhesion ore image separation method based on concave points matching has been proposed. Firstly, adhesion ore images of conveyor belt were obtained by high-speed digital camera, and those images were preprocessed. Secondly, improved Harris corner detection operator was used to detect the corner points, and match concave point pair was determined by corner detection operator coupled with the circular template. Thirdly, sector searching was carried out in simply connected region to match concave point pair and find out the best match concave point pair by rectangular limiting. Finally, division line was determined between concave point pair, and adhesion ore image was completely segmented. Experimental results show the method has high segmentation accuracy and good segmentation effect.