Three-dimensional pose measurement method based on key point tracking

By setting visible key points of varying heights on the rigid body target to be measured, and using a 3D vision recognition system to calculate the distance relationship between invisible key points, a pose coordinate system is constructed. This solves the problems of complex solution of rigid body pose transformation matrix and measurement uncertainty in existing technologies, and achieves efficient and accurate pose measurement.

CN116263327BActive Publication Date: 2026-07-07ZHEJIANG INSTITUTE OF QUALITY SCIENCES
View PDF 4 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
ZHEJIANG INSTITUTE OF QUALITY SCIENCES
Filing Date
2022-12-29
Publication Date
2026-07-07

AI Technical Summary

Technical Problem

In existing technologies, solving the rigid body pose transformation matrix is ​​too complex, which easily introduces measurement uncertainty during measurement, and real-time performance and accuracy are difficult to guarantee.

Method used

A three-dimensional pose measurement method based on key point tracking is adopted. By setting no less than three visible key points of different heights on the rigid body target to be measured, calibrating the spatial coordinates in the three-dimensional vision recognition system, calculating the distance relationship of the invisible key points, and constructing a pose coordinate system to describe the pose change of the rigid body target to be measured.

Benefits of technology

It simplifies the calculation process, reduces measurement errors, improves the accuracy and real-time performance of measurements, and can better represent the pose changes of the rigid body target under test.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN116263327B_ABST
    Figure CN116263327B_ABST
Patent Text Reader

Abstract

The application discloses a three-dimensional pose measurement method based on key point tracking. In order to overcome the problem that the solving of the rigid body pose transformation matrix in the prior art is too complex, the application comprises the following steps: S1: setting not less than three visible key points with different heights on a to-be-measured rigid body target, and calibrating space coordinates in a three-dimensional visual identification system; S2: obtaining visible key point trajectories respectively by the three-dimensional visual identification system, calculating invisible key points describing the pose according to the visible key point trajectories, and obtaining distance relationships between all the key points; S3: in a detection stage, calculating invisible key point space positions according to visible key point space positions on the to-be-measured rigid body target through the distance relationships, constructing a pose coordinate system, and obtaining pose parameters. The position of the invisible key point is determined according to the space position coordinates of the set visible key point, a pose coordinate system is constructed to describe the pose change of the to-be-measured rigid body target, and the calculation process is simple.
Need to check novelty before this filing date? Find Prior Art