针对复杂环境的模块化栅格地图构建算法  

A Modular Local Grid Map Building Algorithm for Complex Three-dimensional Environment

作　　者：秦玉鑫[1] 张高峰[2] 王裕清[1] 

机构地区：[1]河南理工大学机械与动力工程学院,河南焦作454000 [2]河南理工大学工程训练中心,河南焦作454000

出　　处：《控制工程》2016年第10期1627-1633,共7页Control Engineering of China

基　　金：国家自然科学基金<监测机器人数据传输与控制理论研究>(U1261115);2012年度高等学校博士学科学科点专项科研基金(20124116110003);河南省省部共建国家重点实验室培育基地开放基金项目(WS2012A06)

摘　　要：针对煤矿灾害救灾与信息探测机器人的自主导航问题,提出了一种用模块化的局部栅格地图表示复杂三维环境的地图构建算法。该算法使用TOF相机对三维环境信息进行获取,形成三维点云坐标矩阵。通过对坐标阵中每个点进行计算,判断坐标点在局部栅格地图中的位置,创建新的局部栅格地图,并对栅格单元占据概率进行实时更新,创建完整的全局环境栅格地图。根据栅格单元中坐标点的密度,对障碍物进行类型划分,确定通行区域与不可通行区域,简化路径规划的算法。实验结果和数据分析证明了本文地图构建方法的有效性和实用性。An approach for 3D environment is proposed to establish the modular local grid map of outdoor scenes. It is the problem of autonomous navigation for the coal mine rescue robot. The algorithm uses the TOF camera to obtain the 3D information to form a three-dimensional point cloud coordinate matrix. By calculating the coordinate of each point in the 3D coordinate matrix, the location coordinates of the points in the local grid map are determined. The new local grid map is created, and the grid cell occupying probability is updated in real time. Then the whole map of the global environment is created. According to the density of the point in the cell, the type of the obstacle is classified. The prevailing regional and impassable areas are determined to simplify the algorithm of path planning. Experiment results and data analysis show the validity and practicability of the proposed approach in 3D environment modeling.

关 键 词：TOF相机 模块化局部栅格地图 三维点云图像 路径规划 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]