摘要：The storage of hazardous chemicals is closely related to the safety and well-being of the people and the city. Currently, the evenly lay the warehouse in a stacking way is widely used. Hence, the safety conditions especially the “5-distances” have become a key factor during safety monitoring. In this paper, safe storage distance detection based on binocular vision is studied, and a contour-based rapid corner detection method for stacking hazardous chemicals is proposed. This method uses a pre-processed image. First, an edge-thinning algorithm is used to reduce noise. The image then undergoes a hierarchical search and a two stage edge classification, the corners are determined based on the combination of the threshold and the angular coefficient. Experiments have shown that the accuracy of corner detection by this algorithm is 73%, and the time for single image processing is less than 250ms.
