Monocular 6D pose estimation method based on regular dodecahedron

Abstract

The invention discloses a monocular 6D pose estimation method based on a regular dodecahedron. According to the method, an ArUco code is combined with a regular dodecahedron, a monocular camera is used for collecting a regular dodecahedron image marked by the ArUco code, spatial three-dimensional coordinates of a center point of the ArUco code are calculated, sparse point cloud of an inscribed sphere of the ArUco code is obtained, and then multivariate nonlinear fitting and an iterative closest point algorithm are used for comparing virtual regular dodecahedrons at camera coordinates. And the spatial displacement and the rotation angle of the current regular dodecahedron are calculated, and the pose of the measured object is indirectly calculated. The invention overcomes the high requirements of an existing method for illumination, environment and equipment, detection is rapid, and stability is improved; a large amount of energy does not need to be input in the earlier stage for template collection or model training, the pose estimation cost is remarkably reduced, the pose estimation precision is guaranteed, the universality is enhanced, and pose estimation of large-amplitude movement of an object can be conducted.

Description

technical field [0001] The invention belongs to the field of image recognition pose detection, in particular to a monocular 6D pose estimation method based on a regular dodecahedron. Background technique [0002] The pose estimation of the target object is one of the key technologies to ensure the random operation of the robot in the workspace. At present, the industrial automation industry is developing rapidly, and the requirements for real-time detection and feedback of the state of the robot are also increasing, and the estimation of the target pose based on vision is of great significance to the improvement of the performance of the robot. The monocular vision system uses only one camera, has a simple structure and low cost, and is widely used. At present, the main methods of target pose estimation are: [0003] The pose is estimated using a feature point matching method. The specific method is to first extract the feature points in the image, and match the feature p...

AI Technical Summary

Problems solved by technology

However, the accuracy of the template depends on the size of the template library, and one template is suitable for one scene, which lacks versatility

Images