Monocular Image Position and Attitude Measurement Method Based on Circle Feature and Non-Surface Feature Points

Abstract

The invention relates to a monocular image position and attitude measurement method based on circle features and different-surface feature points. The ellipse detection method based on orientation transformation is used to locate the imaging ellipse corresponding to the circle feature on the target in the image, and the circle feature center coordinates, normal and Radius information to obtain relative position and attitude parameters; introduce different-surface point features, use the spatial relationship between different-surface points and circle features, and eliminate the ambiguity based on monocular image circle feature position and attitude parameters; introduce global-based Constrained search strategy, automatic tracking of circular feature imaging ellipse in sequence images, and continuous measurement of target position and attitude parameters. The invention utilizes the target circle feature and different-surface feature points to complete the target position and attitude measurement based only on the monocular image, the algorithm steps are simple and clear, the calculation complexity is low, and the application is easy.

Description

technical field [0001] The invention mainly relates to the fields of computer vision, camera measurement, and mechanical automation, in particular to a monocular image target position and attitude measurement method using target circle features and different-surface feature points. [0002] technical background [0003] Automated operations are more and more commonly used in the fields of aerospace and industrial robots, and the measurement of the relative position and attitude relationship between equipment and targets is the key to realize automated operations. With the development of computer vision and camera measurement technology, more and more attention has been paid to the relative position and attitude measurement method based on vision. The measurement technology based on vision has the advantages of high precision and low cost. [0004] There are two main steps in the existing visual correlation measurement method: feature extraction and position and attitude param...

AI Technical Summary

Problems solved by technology

Existing test results show that the parameter solution accuracy of the first type of method is lower than that of the second type of method. In addition, deep learning technology relies on data-driven and requires a large amount of labeled training data, which limits its application.

Images