摘要
为解决产品表面质量视觉检测中因图像几何特征不足而导致的相机位姿无法校准问题,提出了一种基于图像灰度分布的相机位姿自动校准方法.首先将相机安装于机械臂末端,标定出机械臂末端坐标系与相机坐标系间的转换关系;然后采集相机处于不同位姿下的产品表面图像,在图像感兴趣区域内建立极坐标系,并根据图像灰度与极角之间的分布关系计算出灰度分布轴线的位置,以引导机械臂带动相机作姿态调整;最后再根据图像清晰度评价值控制机械臂沿相机光轴方向平移至正焦位置,最终实现相机位姿校准.以汽车涡轮壳零件为被测对象进行了相机位姿校准试验,图像中感兴趣区域两侧的灰度差值由119降低至4,图像清晰度评价值达到极大值150.90,相机经位姿校准后采集到了准确、清晰的零件表面图像.
To solve the problem that the camera pose cannot be calibrated due to insufficient geometric features of images in the visual inspection of the surface quality,an automatic calibration method of the camera pose is proposed based on the image gray distribution.Firstly,with the camera installed at the end of the manipulator,the transformation relationship between the coordinate system of the manipulator end and the camera coordinate system is calibrated.Then,the product surface images with different camera positions are taken,and the polar coordinate system is established in the region of interest.According to the distribution relationship between image gray and polar angle,the position of the gray distribution axis is calculated to guide the manipulator and drive the camera to make attitude adjustment.Finally,according to the evaluation value of image clarity,the manipulator is controlled to move along the optical axis of the camera to the positive focus position,and the calibration of camera pose is completed.Tests were carried out on automobile turbine housing.The gray difference between two sides of the region of interest is reduced from 119 to 4,and the maximum value of image clarity reaches 150.90.A clear and accurate image of the part surface is obtained after the calibration.
作者
余厚云
王慧青
张辉
胡玉坤
Yu Houyun;Wang Huiqing;Zhang Hui;Hu Yukun(College of Mechanical and Electrical Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China;Wuxi Institute,Nanjing University of Aeronautics and Astronautics,Wuxi 214187,China;School of Instrument Science and Engineering,Southeast University,Nanjing 210096,China)
出处
《东南大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2020年第5期942-949,共8页
Journal of Southeast University：Natural Science Edition
基金
国家自然科学基金资助项目(51975293).
关键词
表面质量
相机位姿校准
灰度分布
视觉检测
自动对焦
手眼标定
机械臂
surface quality
calibration of camera pose
gray distribution
vision detection
autofocus
hand-eye calibration
manipulator
作者简介
余厚云(1975—),男,博士,讲师,meehyyu@nuaa.edu.cn.
