Recognition Of Coal-rock Interface In Infrared Image Based On Active Excitation

The background of this paper is to realize mine automation and intelligent mining,research on an effective method for accurate identification of coal and rock interface based on active infrared technology,which provide important numerical basis and technical support for automatic height adjustment of shearer and cutting mining along coal and rock trajectory.By analyzing the parameters that affect the recognition accuracy of infrared image at the interface of coal and rock,the boundary conditions of each factor are determined.The infrared image acquisition experiment of active thermal excitation at coal-rock interface is carried out by orthogonal test.The FCM image segmentation algorithm is used to process the infrared images of each group of coal and rock collected by orthogonal test,and the recognition accuracy of each segmentation result is calculated.By using the methods of extreme difference analysis and variance analysis to determine the optimal combination of the parameters of each influencing factor and the significant difference of the influence of each influencing factor on the infrared image accuracy of coal and rock.Finally,use the global optimization genetic algorithm to verify the optimal combination of the influencing factors determined by the orthogonal test,and the optimization parameters are basically consistent with the orthogonal test results,and the reliability of the orthogonal test results is verified.Based on the optimal parameter combination of various influencing factors obtained by orthogonal test,the identification experiment of random coal and rock interface is carried out.The experimental results show that within the range of parameters set in this paper,that is,the light distance is 0.8m,the light time is 30 min,the light intensity is 500 W,the shooting angle of infrared thermal imager is 30°,and the position of light source is placed horizontally.The recognition accuracy of coal and rock interface is between 97.69%and 98.19%,and the accuracy of recognition results is high.It can provide accurate coal-rock interface trajectory for automatic mining of coal shearer,and has very good practical value and popularization value.The paper has 39 diagrams,14 tables and 54 references.