Automatic separation of coal mine and waste rock is a non contact and real time screening process,which is an important step in coal mine production,and can effectively improve the energy utilization rate and reduce environmental pollution.However,the existing coal gangue image recognition and sorting technologies mostly use the Laplacian operator method,which leads to small differences in the texture of the extracted images and poor recognition results.This paper takes X ray image coal as the core for image collection.The collected images of coal mine and gangue are filtered and binarized to reduce image noise.The texture extraction is performed on the pre processed image to obtain the texture characteristics of coal gangue.The Prewitt operator method is used to calculate the texture degree and grayscale peak value of the target image.Based on this,the identification value of coal and gangue is proposed.The gray threshold value of coal sorting is determined.The sorting of coal gangue is realized according to the gray threshold value.The experimental results prove that the gray scale difference between coal and gangue extracted by the Prewitt operator method used in this paper is greater,which improves the recognition effect of coal and gangue.