Research on Accurate Grasping Method of Steel Shaft Parts Based on Depth Camera

Aiming at the precise grasping problem of steel shaft parts in the automatic loading and unloading process of the composite mobile robot for the guide roller production line, a part grasping pose estimation method based on depth camera is proposed. Firstly, the transformation relationship between the camera coordinate system and the robotic arm end coordinate system is established through camera calibration, depth map registration, and hand-eye calibration. Secondly, based on OpenCV, the RGB image and depth image are preprocessed, and the maximum contour information of the steel shaft part is extracted by combining the Canny operator and morphological processing. A gripping pose estimation method based on contour geometric moments and PCA is proposed. Finally, combined with the loading and unloading process of the steel shaft part, the automatic loading and unloading operation tasks of the parts are completed. Experimental results show that the overall success rate of the gripping operation of the compound robot visual grasping system can reach over 90%, and its accuracy and stability meet the requirements of the loading and unloading gripping tasks.