中文摘要：

依据MDL VS150空区探测仪工作原理及点云数据空间拓扑关系的特点,基于弦高比理论、最大夹角理论及周长比理论,提出一种点云去噪及数据精简的点云数据完整预处理算法和实施流程,并在广西某矿山的空区探测中开展了实际应用。结果表明：该预处理方法在实现去除噪声点、降低点云数据总量的同时,有效保留了采空区边界三维信息的完整性,最终得到的精确空区三维模型与实际情况符合较好,是一种高效可靠的采空区点云数据预处理方法。

英文摘要：

According to the working principle of MDL VS150 detector and topological spatial relations of point cloud data, a complete set of algorithm for preprocessing point cloud data that combines point cloud denoise with data reduction, was proposed based on the Chord Height Ratio, maximum angle and ratio of perimeters. The application of such preprocessing into the cavity detection for a Guangxi mine showed that the removal of noise points and the reduction of god＇s point cloud data can be realized with the integrity of 3D information about goal＇s boundary retained. The finally obtained 3D model of cavity is in conformity with the actual situation, indicating such algorithm is an efficient andreliable method for preprocessing goaf＇s point cloud data.