A comprehensive RGB-D dataset for 6D pose estimation for industrial robots pick and place: Creation and real-world validation

摘要

• Analysis of data structure and dataset creation methodology: Select the objects to be included in the dataset, collect RGB-D images, and create 6D pose annotations (rotation and translation). Use annotation tools and structure the dataset with folders for images, pose annotations, and object masks. Consider both simulated and real-world data collection methods, and apply data augmentation to enhance the dataset. • Performance evaluation of Efficientpose and FFB6D on the dataset: Compare the two models in terms of pose accuracy using metrics such as ADD/ADD-S (Average Distance of Model Points), execution speed, and precision/recall. Evaluate the models on the created dataset, considering different objects, viewpoints, and environmental conditions. • Testing Efficientpose on the 6 DOF industrial robot for pick-and-place tasks: Integrate the trained Efficientpose model with a manipulator system. Use this model to estimate the 6D pose of objects in real time and pass the pose data to the robot to perform pick-and-place tasks. In the field of robotic grasping, 2D pose estimation algorithms are outdated and insufficient for modern requirements. Transitioning to 6D pose estimation of objects offers, particularly through deep learning methods, significantly improved performance. However, most high-performance 6D algorithms lack publicly disclosed methods for creating the necessary datasets, presenting challenges for researchers. This study introduces a methodology for creating a sample dataset for 6D pose estimation, addressing the challenges researchers face when applying these algorithms to specific projects. RGB and depth images are captured using Intel® RealSense™ Depth D435 camera, forming the foundation for accurately determining the pose of each object in the dataset. A CAD model along with accompanying metadata files, is generated to provide a complete dataset. The sample dataset was tested on two algorithms: EfficientPose and FFB6D, achieving accuracy rates of 97.05 % and 98.09 % respectively. These results indicate that the dataset is both effective and applicable to real-world robotic grasping tasks. Our research offers substantial practical value, enabling researchers and engineers to easily apply state-of-the-art 6D pose estimation algorithms to fields such as conveyor belt robotic picking, medical automation, and various other applications.