基于机器视觉的井下搜救机器人路径规划  

Path Planning for Rescue Robot Based on Machine Vision in Underground

作　　者：徐斌[1] 卜祥丽[1] 张玉洁[1] 

机构地区：[1]华北科技学院机电工程学院,北京101601

出　　处：《煤矿机械》2015年第6期247-248,共2页Coal Mine Machinery

基　　金：中央高校基本科研业务费专项资金资助(3142013066)

摘　　要：针对井下复杂矿难环境,传统方法建立环境地图需要大量信息,处理复杂搜索问题,提出基于机器视觉建立环境地图。算法从帧差背景中获取环境障碍物位置信息,依据机器人加速鲁棒特征定位,离散建模运动空间。实验仿真表明,通过自动机演化规则实现细胞状态分析,可实现任意起点到终点的最优路径搜索。The traditional way to create environmental map is necessary to obtain great amount environmental information, which will lead to the trouble of search difficulty. A new algorithm to build the map by machine vision is proposed. The method makes use of monocular vision to achieve the position of obstacles from the frames. And the SURF with the robot will be found in order to recognize the position of robot in the scene. Then according to the enviromnent model build by computer vision, the optimal path will be found from the start to the end using a cellular automation algorithm.

关 键 词：单目视觉 移动机器人 自动机 路径规划 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]