基于单目视觉的工件定位技术研究

doi:10.16180/j.cnki.issn1007-7820.2019.12.015

摘要/Abstract

摘要：

针对普通工件的位姿定位问题,文中提出了一种使用单目视觉平台和CAD模型的工件定位方案。为了阐明该方案的具体设计,基于单目相机的标定方法引入工件图像识别的相关技术,设计了工件的图像预处理过程,并提出基于CAD模型的单目视觉定位算法。实验测试结果表明,与基于先验性关系的定位方案相比,该定位方案具有更高的精度和较好的灵活性,可以满足普通工件的定位需要。

关键词: 机器人, 单目视觉, 图像处理, 位姿定位

Abstract:

Aiming at the problem of position and pose location of common workpiece, a workpiece location scheme using monocular vision platform and CAD model was proposed. In order to clarify the specific design of this scheme, this paper introduced the calibration method of monocular camera, the related technology of workpiece image recognition, designed the image preprocessing process of workpiece, and finally proposed a monocular vision localization algorithm based on CAD model. The experimental results showed that, compared with the location scheme based on priori relationship, the location scheme had higher accuracy and better flexibility, and could meet the needs of common workpiece location.

Key words: industrial robot, monocular vision, image processing, posture recognition

中图分类号:

TP212

杨云辉. 基于单目视觉的工件定位技术研究[J]. 电子科技, 2019, 32(12): 72-75.

YANG Yunhui. Research on Workpiece Location Technology Based on Monocular Vision[J]. Electronic Science and Technology, 2019, 32(12): 72-75.

图/表 6

图1

图2

表1

图3

表2

表3

参考文献 10

[1]	张振 . 基于单目视觉的工件位姿识别与抓取系统[D]. 杭州:中国计量大学, 2016.
	Zhang Zheng . Workpiece pose recognition and grasping system based on monocular vision[D]. Hangzhou:China University of Metrology, 2016.
[2]	严卫生, 高智, 杨小龙 , 等. 面向AUV自主回收的单目视觉定位算法[J].电子设计工程,2014(22):174-176.
	Yan Weisheng, Gao Zhi, Yang Xiaolong , et al. Monocular vision localization algorithm for AUV self-recovery[J].Electronic Design Engineering, 2014(22):174-176.
[3]	朱统帅, 徐维荣, 莫锦秋 . 基于单目立体视觉的顶层工件定位方法[J].机械设计与研究,2017(4):49-52,57.
	Zhu Tongshuai, Xu Weirong, Mo Jinqiu . Top-level workpiece location method based on monocular stereo vision[J].Mechanical Design and Research, 2017(4):49-52,57.
[4]	张驰, 廖华丽, 周军 . 基于单目视觉的工业机器人智能抓取系统设计[J]. 机电工程, 2018,35(3):283-287.
	Zhang Chi, Liao Huahua, Zhou Jun . Design of intelligent grasping system for industrial robots based on monocular vision[J]. Mechatronics Engineering, 2018,35(3):283-287.
[5]	陈宗海, 洪洋, 王纪凯 , 等. 基于循环卷积神经网络的单目视觉里程计[J]. 机器人, 2019,41(2):147-155.
	Chen Zonghai, Hong Yang, Wang Jikai , et al. Monocular visual odometer based on cyclic convolution neural network[J]. Robot, 2019,41(2):147-155.
[6]	程亚军, 任小洪, 蔡绍堂 , 等. 基于级联神经网络的单目视觉目标定位[J].自动化与仪器仪表,2018(5):14-17,23.
	Cheng Yajun, Ren Xiaohong, Cai Shaotang , et al. Monocular vision target location based on cascaded neural network[J].Automation and Instrumentation, 2018(5):14-17, 23.
[7]	李星云, 李众立, 廖晓波 . 基于单目视觉的工业机器人定位系统的设计[J]. 机床与液压, 2015,43(9):35-38.
	Li Xingyun, Li Zhongli, Liao Xiaobo . Design of industrial robot positioning system based on monocular vision[J]. Machine Tool and Hydraulic, 2015,43(9):35-38.
[8]	汤月娟, 徐晶, 司书哲 , 等. 一种基于NAO机器人的单目视觉目标定位方法[J].长春理工大学学报(自然科学版),2014(5):95-98.
	Tang Yuejuan, Xu Jing, Si Shuzhe , et al. A monocular vision target location method based on NAO robot[J].Journal of Changchun University of Technology(Natural Science Edition), 2014(5):95-98.
[9]	翟敬梅, 董鹏飞, 张铁 . 基于视觉引导的工业机器人定位抓取系统设计[J].机械设计与研究,2014(5):45-49.
	Zhai Jingmei, Dong Pengfei, Zhang Tie . Design of industrial robot positioning and grasping system based on vision guidance[J]. Mechanical Design and Research, 2014(5):45-49.
[10]	敬泽, 薛方正, 李祖枢 . 基于单目视觉的空间目标位置测量[J]. 传感器与微系统, 2011,30(3):125-127.
	Jing Ze, Xue Fangzheng, Li Zushu . Position measurement of spatial targets based on monocular vision[J]. Sensors and Microsystems, 2011,30(3):125-127.

内参	数值	外参	数值
f	16.008	t₁	-97.656
d_x	2.300	t₂	-104.007
d_y	2.300	t₃	832.167
u₀	1 306.640	r₁	2.240
v₀	1 186.765	r₁	2.210
K₁	0.075	r₁	0.069
K₂	-6.450	-	-
K₃	89.589	-	-

编号	实际坐标/mm	定位坐标/mm
1	(463.32,63.24,85.04)	(462.26,62.78,84.47)
2	(463.88,-34.95,83.37)	(464.33,-35.57,82.35)
3	(482.06,63.38,84.52)	(480.38,64.04,82.27)
4	(435.75,53.65,84.51)	(436.32,55.21,86.31)

编号	实际坐标/mm	定位坐标/mm
1	(435.64,25.23,83.41)	(434.87,24.95,83.49)
2	(463.70,23.51,79.18)	(465.84,25.32,82.17)
3	(427.99,-32.44,83.42)	(428.48,-31.38,84.16)
4	(465.48,-19.12,84.22)	(464.76,-18.74,84.32)

[1]	宗圣康,程建鹏,张西良. 基于光斑位置的起重机轨道高度差自动检测方法[J]. 电子科技, 2022, 35(1): 21-28.
[2]	宋章明,贺慧勇,黄跃俊. 扁平漆包线表面缺陷在线检测研究[J]. 电子科技, 2021, 34(5): 72-78.
[3]	闫超,孙占全,田恩刚,赵杨洋,范小燕. 基于深度学习的医学图像分割技术研究进展[J]. 电子科技, 2021, 34(2): 7-11.
[4]	陈文佑,章伟,史晓帆,宋芳. 一种改进ORB特征匹配的半稠密三维重建ORB-SLAM算法[J]. 电子科技, 2021, 34(12): 62-67.
[5]	胡琼琼,伞红军,陈久朋,谢飞亚,陈中平,李鹏宇. 一种四自由度并联机器人的运动性能分析及仿真[J]. 电子科技, 2021, 34(11): 46-54.
[6]	陈明方,张凯翔,陈久朋,熊彬洲,李奇,姚国一,李鹏宇. 基于MATLAB Robotics Toolbox的四足机器人轨迹仿真与优化[J]. 电子科技, 2020, 33(9): 31-37.
[7]	韩强,李昌昊,程建鹏,张西良,李伯全,尹经天. 起重机轨道检测机器人激光姿态角变结构双闭环控制研究[J]. 电子科技, 2020, 33(7): 17-21.
[8]	陈久朋,伞红军,张道义,胡琼琼,张凯翔,熊彬州. 一种四足机器人对角步态仿真分析[J]. 电子科技, 2020, 33(5): 33-38.
[9]	史伟忠,曹卫卫,范彦铭,董佳筠,陈纾,肖昊. 基于FPGA的深空图像实时边缘检测算法与实现[J]. 电子科技, 2020, 33(5): 45-49.
[10]	章天昊,李郝林. 一种基于高斯拟合法的激光光束中心提取方法[J]. 电子科技, 2020, 33(5): 72-76.
[11]	秦兴,高晓琪,陈滨. 基于压缩卷积神经网络的图像超分辨率算法[J]. 电子科技, 2020, 33(5): 1-8.
[12]	张道义,伞红军,陈久朋,李鹏飞,熊彬州. 5-DOF混联雕刻机速度运动学分析及仿真[J]. 电子科技, 2020, 33(4): 42-49.
[13]	谢智烜,姚红兵,范宁,陈枫. 面向多目标透镜的连通域标记检测算法[J]. 电子科技, 2020, 33(4): 50-54.
[14]	詹雁,张娟. 一种结构感知损失的域适应深度估计方法[J]. 电子科技, 2020, 33(12): 12-16.
[15]	张凯翔,陈久朋,熊彬州,李奇,胡琼琼. 工业机器人空间轨迹仿真分析[J]. 电子科技, 2020, 33(10): 26-32.