基于机器视觉的工件位姿估算方法研究  

Research on Pose Estimation of Workpiece Based on Machine Vision

作　　者：彭静 田军委[1] 张震 李锦涛 Peng Jing;Tian Junwei;Zhang Zhen;Li Jintao(School of Mechanical and Electrical Engineering,Xi'an University of Technology,Xi'an 710021,China)

机构地区：[1]西安工业大学机电工程学院,陕西西安710021

出　　处：《科学技术创新》2022年第34期99-102,共4页Scientific and Technological Innovation

摘　　要：零件装配中目标物位姿估算问题,使用图像处理获取工件端面椭圆参数,本研究提出一种基于机器视觉的工件位姿估算方法,首先使用机械臂自带相机采集不同腕关节角度下的样本图,然后利用ROI提取、轮廓提取、椭圆拟合等算法获取工件端面参数,最后将上述参数与断面的最小外接圆参数对比,获取偏差样本和对应的腕关节角度偏差,将腕关节角度作为校正量,然后通过验证,可成功实现将工件位姿调整至水平位置。For the problem of object pose estimation in part assembly, image processing is used to obtain the elliptical parameters of the end face of the workpiece. This paper proposes a method for estimating the position and pose of the workpiece based on machine vision. First, use the camera of the manipulator to collect sample images at different wrist joint angles, then use ROI extraction, contour extraction, ellipse fitting and other algorithms to obtain the end face parameters of the workpiece. Finally, compare the above parameters with the minimum circumscribed circle parameters of the section, Obtain the deviation sample and the corresponding wrist angle deviation, take the wrist angle as the correction value, and then, through verification, successfully adjust the workpiece position and posture to the horizontal position.

关 键 词：位姿估算 图像处理 机械臂 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]