基于视觉的挖掘机工作装置位姿测量  被引量:2

Pose Measurement of Excavator Device Based on Vision

在线阅读下载全文

作  者:马伟 宫乐[1,2] 冯浩 殷晨波 周俊静[1] 曹东辉[3] MA Wei; GONG Yue;; FENG Hao; YIN Chenbo; ZHOU Junjing ;CAO Donghui(Institute of Automobile and Construction Machinery, Nanjing Tech University, Nanjing 211816,China;Atlas Copco (Nanjing) Construction & Mining Equipment Co. Ltd. Nanjing 210046,China; Sanyi Heavy Machinery Co. , Ltd. , Suzhou Jiangsu 215300, China)

机构地区:[1]南京工业大学机械与动力工程学院,南京211816 [2]阿特拉斯科普柯南京建筑矿山设备有限公司,南京210046 [3]三一重机有限公司,江苏苏州215300

出  处:《机械设计与研究》2018年第5期173-176,182,共5页Machine Design And Research

基  金:江苏省科技计划前瞻性联合研究项目(BY2015005-15)

摘  要:针对接触式传感器在恶劣环境中测量位姿不可靠这一问题,研究了通过非接触式的视觉摄像机来识别挖掘机工作装置位姿信息的方法。首先,采用摄像机拍摄挖掘机工作装置的2维图,将其经过图像处理得到二值化图像。然后,对图像进行轮廓提取,获得了工作装置关节点的像素坐标,并对比了不同算法的精确性。最后,通过理论计算推导出坐标系之间的转换矩阵,结合Open CV软件将像素坐标转换为世界坐标,并利用实验验证了其正确性。实验表明:该方法测量所得结果与全站仪测得结果的误差在30 mm内,两者吻合较好,可以应用于挖掘机位姿的实际测量中。For the problem that the contact sensors measure position is not reliable,a method of identifying pose information of excavator working device by non-contact visual camera is studied. First,the camera takes a two-dimensional map of the excavator's working device,which is subjected to image processing to obtain a binarized image.Then,the image is extracted and the pixel coordinates of the joint of the working device are obtained,and the accuracy of the different algorithms is compared. Finally,the transformation matrix between coordinate systems is deduced by theoretical calculation,and the coordinates of the pixel are converted into world coordinates with Open CV software,and the correctness is verified by experiments. Experimental results show that: the error between measurement results obtained with this method and Total Station is within 30 mm and the two are in good agreement.They can be applied to the actual measurement of excavator's position and orientation.

关 键 词:挖掘机 位姿信息 视觉 图像处理 

分 类 号:O357.1[理学—流体力学] S275.6[理学—力学]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象