激光扫描数据的获取与处理是煤矿智能工作面的关键技术之一。作为监测煤矿采煤机截割轨迹的重要手段，三维激光扫描技术，因其具有非接触、精度高、受烟尘影响小等特点日益受到重视，但由于激光点云数据存在的海量性、离散性、冗余多的特点，实际生产中，目前只能将远程获取工作面三维点云信息，无法利用激光点云数据的空间特征直接提取采煤机割煤顶板线，即无法提供当前刀采煤机截割轨迹的反馈信息。为实时获取和预测采煤机当前刀割煤顶板线位置数据，在基于物探、钻探、巷道素描和激光扫描等数据构建初始透明化工作面的基础上，利用激光扫描仪，实时感知获取煤矿井下综采工作面的激光点云数据，在充分理解激光扫描装置运行原理和激光点云数据特征的前提下，针对三维激光点云存在的数据离散、信息量大、特征提取困难等问题，通过移除离群点、点云滤波、点云切片和基于空间形态的特征点提取等算法，建立了一套完整的工作面三维激光点云的特征提取流程，实现了综采工作面激光点云的割煤顶板线的自动提取，并与基于目视解译法提取的割煤顶板线数据进行对比验证，误差小于0.04 m的点数量占84%，误差小于0.08 m的点数量占96%，验证了该提取采煤机当前刀割煤顶板线算法的可行性，从而为下一刀或下几刀采煤机截割行程提供数据参考。

The acquisition and processing of laser scanning data is one of the key technologies for the intelligent working face of coal mines. As an important mean to monitor the cutting trajectory of coal shearers,3D laser scanning technology has been increasingly valued because of its characteristics of non-contact,high precision and less effect by smoke and dust. However,due to the massive,discrete and redundant characteristics of laser point cloud data,in actual production,at present,only the 3D point cloud information of the working face can be obtained remotely and the spatial characteristics of the laser point cloud data cannot be used to directly extract the coal cutting roof line of the shearer,namely feedback information of the current shearer cutting trajectory cannot be provided. For real-time acquisition and prediction of current shearer cutting coal roof line position data，on the basis of constructing the initial transparent working face based on data such as geophysical prospecting,drilling,roadway sketching and laser scanning,the laser scanner is used to perceive and obtain the laser point cloud data of the fully mechanized coal mining face in real time so as to fully understand the laser scanning device. Under the premise  of the operation principle and the characteristics of laser point cloud data,aiming at the problems of data dispersion,large amount of information and difficult feature extraction of the three-dimensional laser point cloud,a complete set of feature extraction process of 3D laser point cloud of working face is established by removing outliers,filtering point cloud,slicing point cloud and feature point extraction based on spatial morphology,which realizes the automatic extraction of coal cutting roof line of laser point cloud in fully-mechanized mining face,and is combined with the visual interpretation method. The extracted coal cutting roof line data are compared and verified,and the points with error less than 0.04 m account for 84%,and the points with error less than 0.08 m account for 96%,which verifies the feasibility of extracting the current algorithm of the cutting coal roof line of the shearer. The cutting stroke of the shearer with the next or next cuts provides data reference.