Study on the method of personnel accurate detection in shearer operating space

Shearer is one of the key equipment in fully mechanized working face, and its intelligence is the necessary condition to realize less-human and unmanned mining. Unfortunately, although shearer has functions including three-dimensional positioning, memory cutting, remote monitoring and etc., the detection and pre-warning function for errant personnel who straying into the shearer operating space has not yet been realized. The lack of this function may lead to serious safety accidents once underground personnel straying into the shearer operating space. Hence, it is necessary to research the key technologies of personnel detection in shearer operating space to realize the safety production when shearer operating and achieve safe and efficient production of coal mine. In view of the characteristics of low illumination and complex operating environment of fully mechanized working face, an accurate personnel detection system in shearer operating space based on infrared thermal imaging technology is proposed. By analyzing the characteristics of infrared image noise in fully mechanized working face, a multi-level guided filtering model based on Gaussian mask is proposed; Based on Lucas-Kanade optical flow and fuzzy segmentation theory, the motion information extraction for moving foreground target in dynamic background and the segmentation for infrared scene information are realized, respectively; Combined with the moving-target motion information and infrared-scene segmentation results, a weight voting method based on morphology theory is constructed to realize the accurate detection of personnel in shearer operating space. Finally, the underground industrial test was carried out in the 21208 fully mechanized working face of Gengcun Coal Mine of Henan Dayou Energy Co., Ltd. the test results show that the tracking deviation of the proposed personnel detection system in the shearer operating space in the actual fully mechanized working face is as low as 0.1065 pixel width, and the coincidence ratio is 96.10%. As accurate personnel detection in shearer operating space is the premise of effective personnel safety protection, the establishment of the personnel accurate detection system provides technical support for the safety production of intelligent fully mechanized working face.