A coal gangue separation method based on depth vision

Abstract

The invention discloses a coal gangue separation method based on depth vision, and relates to the technical field of machine vision. The method includes taking the ore color image of the ore sorting production line as input, and using the ore segmentation convolutional neural network to perform segmentation to obtain the ore area Segmentation results, according to the ore region segmentation results, the ore region is clipped from the ore color image to obtain several ore region slices, and then the ore classification convolutional neural network is used to output the ore type as coal or coal gangue, and the image analysis of the ore region slices is determined ore information, and finally realize the automatic separation of coal and gangue; this method can realize rapid detection and accurate identification, and provide accurate guidance information for the automatic separation of coal gangue by estimating the volume level and relative position of the ore. High efficiency; two convolutional neural networks are specially customized for ore separation scenarios, and have the characteristics of small size, strong real-time performance, and high detection accuracy.

Description

technical field [0001] The invention relates to the technical field of machine vision, in particular to a method for separating coal gangue based on depth vision. Background technique [0002] Coal gangue is a solid waste discharged from coal mining and washing and processing, and its output accounts for about 15% of raw coal output. Coal gangue separation is a necessary process in the coal mine production process. Removal of gangue is also the basis for coal production of clean energy. Reducing the emission of gangue particles can reduce the emission of pm2.5 units. Sorting gangue can reduce the cost of washing, improve the grade of finished coal, and improve the economic benefits of coal mining enterprises. Strengthening the research on the comprehensive treatment and utilization of coal gangue and improving the clean coal technology are the important research contents of the current coal-burning country, and also the needs of environmental protection and development. ...

AI Technical Summary

Problems solved by technology

[0003] At present, most coal preparation plants use manual visual inspection to sort coal blocks, and the sorting speed and sorting accuracy are not ideal. Some coal preparation plants also use deep vision technology to realize the identification of coal blocks and gangue for automatic sorting. However, there is a large amount of flying coal ash in the coal preparation process, which will interfere with the machine vision detection equipment. Therefore, after a period of use, the recognition accuracy will be reduced, the reliability will not be high, or regular manual cleaning is required, and manual intervention is still required, resulting in difficulties. Complete automation, the actual sorting effect is not ideal

Images