Research On The Target Tracking And Early-warning Method Of Underground

Safe production is an eternal topic of mine safety,while safe transport is an important part of it.It is emphasized that the focus of technology development in mine safe production and monitor is solving the key technical problems in the process of mine disaster prevention and improving the ability of mine disaster comprehensive prevention by the means of using advanced and novel technology.So rail safe transport faces a severe challenge.Firstly, this paper combining background subtraction and frame difference method to detect the moving object for the underground complex environment. Then we research the object recognition method based on the detection result, which is divided into the recognition of locomotive and mine worker. For the apparent shape difference of locomotive and mine worker, we propose a method combining shape feature and texture feature to detect the moving object.Second, this adopt a underground warning method which is based on time safe model.For the underground complex environment of similarity in gray of object and background, low lighting, and gray simple of object, the multi-object tracking algorithm was proposed and achieve the goal of tracking multi-object by improving the particle filter algorithm finally.Last, research the underground warning method. We calculate the distance between objects based on monocular vision principle and then make the safe warning decision according to the value which is got by comparing the distance between object and mine worker and the safe warning value calculated by safety warning model based on time. Then,according to the compared value, we judge whether the distance between object and mine worker is less than the safe warning value. It will issue a warning voice to protect the safety ofmine workers when there is a security risk.This paper researches and proposes a underground moving object tracking and warning method.All in all, the method of underground moving object tracking and warning proposed in this paper can accomplish real-time automatic monitoring to underground moving object by making full use of video monitor information in mine industry and has a critical meaning to mine monitor and management and safe prevention.