Abstract:In the robot finishing process of the assembly interface of large aerospace cylindrical components (short for assembly interface), to realize the high-precision and high-efficiency pose perception of the large component, an intelligent pose measurement method for the large component is proposed based on binocular vision and priori data. In this method, the global coordinate system of the robot finishing system is initially established using laser tracking measurement method and customized reference plates, giving a unified coordinate transformation datum for the interoperation of the finishing system's subsystems. Then, utilizing deep learning and digital image processing technologies, an algorithm for recognizing and locating key features of the large component is developed, which can realize the identification of key feature types and accurate localization of feature centroids. Following that, the global coordinate of the key feature centroid is determined using the proposed binocular vision three-dimensional (3D) coordinate reconstruction method. Meanwhile, by introducing the priori processing data of the large component to match the 3D reconstruction coordinates of the key feature centroids, the spatial pose of the large component can be calculated with high precision. Finally, the proposed method is experimentally validated using a case study of a large aerospace cylindrical component. Experimental results prove that the proposed method can achieve high-precision pose measurement of the large component, which can provide pose data support for the adjustment or modification of the cutting path of the robot that is generated by the as-designed model of the large component, to ensure the correctness of the robotic machining of the assembly interface, and thus the proposed method can meet the robot finishing needs of the large component.

High-precision Visual Localization Based on an Improved Shape-Based Matching Algorithm.

Leveraging Local Planar Motion Property for Robust Visual Matching and Localization.

An Improved Supervoxel Clustering Algorithm of 3D Point Clouds for the Localization of Industrial Robots

Long-Term Map-Based Visual Localization: Analysis of Individual Components of a Hierarchical Pipeline

Robust Indoor Localization and Map Matching Algorithm with Visual SLAM

End-to-end Pattern Optimization Technology for 3D Shape Measurement.

A High-Precision Template Localization Algorithm Using SIFT Keypoints

A Machine Vision Based Method of High-precision Measurement of the Size of Workpiece

A Feature Matching Algorithm Applied to Welding Trajectory Revising

High-precision point cloud registration system of multi-view industrial self-similar workpiece based on super-point space guidance

Precise and Robust Binocular Camera Calibration Based on Multiple Constraints

A High-Precision Localization Algorithm By Improved Sift Key-Points

A High-Precision Mark Positioning Algorithm Based on Sub-Pixel Shape Template Matching in Wafer Bonding Alignment

A System and Vision Localization Method for the Opening of Railway Oil Tank Wagon Based on Shape Matching

Vision-Driven Dynamic Texture Recognition for Light Bar Intelligent Assembly Process

Improving Feature-based Visual Localization by Geometry-Aided Matching

Optimization Scheme of Measuring Accuracy of Bearing Seat Size Based on Machine Vision

Structured Light Based High Precision 3D Measurement and Workpiece Pose Estimation

Linear Laser Scanning Measurement Method Tracking by a Binocular Vision

Binocular vision and priori data based intelligent pose measurement method of large aerospace cylindrical components

A three-dimensional visual localization system based on four inexpensive video cameras