Application of Volume Detection Based on Machine Vision in Coal and Gangue Separation

In order to improve the separation rate of coal and gangue, aiming at sorting of large-size coal and gangue based on density characteristics, a coal and gangue separation system based on machine vision was designed, and a side height calibrated shape from shading algorithm was proposed for volume detection. The density is calculated by measuring and analysing mass and volume detection in this system. The gray level information and side height of the ore contour obtained by two Angle cameras were used to detect the volume. Firstly, the depth information is calculated from the image gray level by using the linear shape from shading algorithm. Secondly, the side height is obtained from the image contour. Then, the depth information is corrected by combining the side height to get the height. The results show that the height correction shape from shading algorithm can effectively extract the three-dimensional information of ore and reduce the error of volume detection. The volume change rate under different light intensities is verified that the proposed algorithm can reduce the interference of light intensity on volume detection and enhance robustness, which provides an important basis for subsequent separation based on density.