The Research And Application Of Mine Roadway3D Scene Reconstruction

Coal has been the largest energy consumer in China for decades. However, theefficiency of underground coal mining is rather low due to underground heat, humidity,dust and other harsh environments. And in underground coal mining, the roadwayexcavation plays a very critical role, which is also efficiency-low with dependence onmanual command to work. What’s more, manual command to work inevitably causesdeviations, which needs to check many times manually. So improving the performance ofroadway excavation is one of the hot spot in researches of coal production safety.In this paper, firstly the main registration algorithms of3D point cloud aresummarized and analyzed at home and abroad, secondly RANSAC (Random SampleConsensus) algorithm and the ICP (Iterative Closest Point) algorithm are deeply discussed.For the problem that the accuracy of searching closest point in traditional ICP algorithm isno high, the angle between the normal vector of the two closest points as a constraint isapplied when searching closest points. For the problem that the initial value of the unitquaternion is related to the result and convergence rate of traditional ICP algorithm, acoarse-to-fine registration method based on RANSAC algorithm and ICP algorithm isproposed. Through the experiment in the library when given a same experiment object, theresults show that the performance of the proposed method in the convergence rate and thebad accuracy of searching closest point are improved, which proves the effectiveness ofthe proposed method.On the basis of the above methods, a set of mine roadway3D scene reconstructionsystem is designed and implemented, which is targeted at the tunneling face, in this paper.It can show the3D sence of downhole roadway dynamically. We complemented thefunction of heading machine cutting window detection, the function of displaying themine roadway exetension direction as well as improving the nacigation parameter of heading machine to meet the need of working in downhole roadway. This system providessome technical basis for establishing a system software platform, which can monitor themachine working conditions and scan the3D model of job site, so that it is possible tocomplement remote visual monitoring of thin coal seam mining.