A Monocular 6D Pose Estimation Method Based on Icosahedron

Abstract

The invention discloses a monocular 6D pose estimation method based on a regular icosahedron. This method combines the ArUco code with the icosahedron, and then uses the monocular camera to collect the icosahedron image marked by the ArUco code, calculates the spatial three-dimensional coordinates of the center point of the ArUco code, and obtains the sparse point cloud of its inscribed sphere , and then use multivariate nonlinear fitting and iterative closest point algorithm to compare the virtual icosahedron at the camera coordinates, and then calculate the spatial displacement and rotation angle of the current icosahedron, and indirectly calculate the pose of the measured object . This method overcomes the high requirements of the existing methods on lighting, environment and equipment, detects quickly, and improves stability; it does not need to invest a lot of energy in the early stage for template collection or model training, which significantly reduces the cost of pose estimation and ensures Pose estimation accuracy and enhanced versatility, and can perform pose estimation of large-scale motion of objects.

Description

Technical field [0001] The present invention belongs to the field of image recognition detecting position and orientation, in particular a monocular icosahedron. 6D pose estimation method. Background technique [0002] Pose target estimate whether in production or aspects of life actor an important role, is a prerequisite to achieve flexible and efficient robot motion. Currently the industrial automation industry is growing, the state of the robot real-time detection and feedback is also rising, and vision-based target pose estimation is important for improving robot performance. Monocular vision system using only one camera, simple structure and low cost, used widely. The main objective method at present pose estimation are: [0003] Document "LEPETIT V, PILET J, FUA P.Point matching as a classificationproblem for fast and robust object pose estimation [C] / / Proceedings of the2004IEEE Computer Society Conference on Computer Vision and PatternRecognition, 2004.CVPR 2004.:volume ...

AI Technical Summary

Problems solved by technology

In the document "REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: Unified, real-time object detection [C] / / Proceedings of the IEEE conference on computer vision and pattern recognition.2016:779-788." The YOLO network restores its pose through the two-dimensional frame of the target in the image, but the method based on deep learning requires a large number of sample resources to construct a large amount of sample data and mark the real pose, and requires a large amount of computing resources for network training. It is difficult to control the cost, and most of them are targeted development, which cannot meet the versatility

Images