Research on ore segmentation algorithm based on improved watershed algorithm

Abstract In mining enterprises, ore particle size is the main technical indicator to measure the degree of ore fragmentation and is also an important parameter for mineral processing automation. In response to the problems of low segmentation accuracy and severe undersegmentation in traditional algorithms for processing ore images in complex environments, this paper proposes a segmentation algorithm that combines improved distance transformation with watershed. Firstly, we preprocess the ore image appropriately; Then, we obtain the seed point region using the local maximum algorithm proposed in this article; Finally, we combine morphological operations with optimized seed point regions to perform watershed transformation on the ore image. The comparative experiments showed that the ore segmentation algorithm proposed in this paper has the advantages of fewer limiting factors, higher efficiency, and higher accuracy and has a good detection effect and could meet the actual operation needs of mining enterprises.