基于图像和激光点云融合的智能采面煤岩识别  

Coal-Rock Recognition of Intelligent Mining Face Based on the Fusion of Image and Laser Point Cloud

作　　者：司垒[1] 王忠宾[1] 李嘉豪 魏东 梁斌[1,2] 肖俊鹏[3] SI Lei;WANG Zhongbin;LI Jiahao;WEI Dong;LIANG Bin;XIAO Junpeng(School of Mechatronic Engineering,China University of Mining and Technology Xuzhou,221116,China;Xuhai College,China University of Mining and Technology Xuzhou,221008,China;Inner Mongolia Zhahanur Coal Industry Co.,Ltd.Tongliao,029100,China)

机构地区：[1]中国矿业大学机电工程学院,徐州221116 [2]中国矿业大学徐海学院,徐州221008 [3]内蒙古扎鲁特旗扎哈淖尔煤业有限公司,通辽029100

出　　处：《振动．测试与诊断》2023年第2期254-262,407,共10页Journal of Vibration,Measurement & Diagnosis

基　　金：国家自然科学基金面上资助项目(52074271,52174152);江苏省自然科学基金面上资助项目(BK20211245);中国博士后科学基金特别资助项目(2020T130696);江苏省科协青年科技人才托举工程资助项目;江苏高校优势学科建设工程资助项目。

摘　　要：为了提高采煤工作面的智能化水平,提出了基于图像和激光点云融合的煤岩识别方法。首先,利用三维重建构建了蕴含煤岩颜色信息及截割纹理特征的图像点云;其次,提出了基于改进迭代最近点(iterative closest point,简称ICP)算法的煤岩点云配准方法,提高了点对间的搜索速度和配准精度;然后,设计了基于改进区域生长算法的煤岩识别方法,通过仿真分析验证了改进措施的有效性;最后,搭建了采煤机煤岩截割实验系统,并对相关改进算法进行了实验对比分析。结果表明,所提方法的点云数据分割效果最好,煤岩识别准确率达92.95%。在煤矿井下采煤工作面进行了现场测试,进一步证明了所提煤岩识别方法的可行性和实用性。In order to improve the intelligent level of coal mining face,this paper proposes a novel coal-rock recognition method based on the fusion of image and laser point cloud.Firstly,the three-dimensional reconstruction is used to realize the construction of coal-rock image point cloud containing color information and cutting texture features.Secondly,a coal-rock point cloud registration algorithm based on improved iterative closest point(ICP)algorithm is proposed to improve the search speed and accuracy between point pairs.Then,a coal-rock recognition method based on improved regional growth algorithm is designed.The effectiveness of the improved measures is verified by simulation analysis.The self-established coal-rock cutting experimental system is set up,and the experimental comparison and analysis of the improved ICP and region growth algorithm are carried out.The results indicate that the point cloud data segmentation effect of the proposed method is the best and the accuracy of coal-rock recognition can reach to 92.95%.The field test is carried out in the underground coal face,which further proves the practicability and feasibility of proposed coal-rock recognition method.

关 键 词：煤岩识别 图像点云 激光点云 点云配准 点云分割 

分 类 号：TH6[机械工程—机械制造及自动化] TD42[矿业工程—矿山机电]