Relative pose measurement of binocular vision based on feature circle

In traditional pose measurement algorithms, there are many visual measurement solutions for feature points and lines but few for feature circles. In this paper, the feature circle is described by the rotation characteristic of a point around a rotating axis in dual quaternion algebra. Dual vectors are used to describe the feature points, and nicker lines are used to describe the rotation axis; that is, the dynamic characteristic of the feature point rotating around the nicker line is used to describe the feature circle, and then the relative position and attitude information between the leader and follower rigid bodies are calculated by the principle of binocular vision measurement. The simulation results show that the algorithm can meet the requirements of the relative pose measurement accuracy of rigid bodies.