Vision-based harmonic reducer meshing condition and backlash detection method
Through a vision-based detection method, a light field camera and a UNet network are used to detect the meshing condition and side clearance of the harmonic reducer, which solves the problems of low efficiency and susceptibility to subjective factors in traditional detection methods and achieves high-precision detection results.
CN120672945APending Publication Date: 2025-09-19BEIJING CTKM HARMONIC DRIVE CO LTD
View PDF 0 Cites 0 Cited by
Patent Information
- Application Number
- CN202510736877.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-04
- Publication Date
- 2025-09-19
Smart Images
Figure CN120672945A_ABST
Abstract
The invention discloses a vision-based harmonic reducer meshing condition and backlash detection method, and belongs to the technical field of reducers. Comprising the following steps: acquiring a multi-tooth meshing image of a meshing part of the harmonic reducer and standardizing a tooth profile area to obtain a standard image of a rigid-flexible wheel tooth profile; performing multi-layer feature decomposition to obtain a plurality of different layers of first gear feature images; carrying out feature map fusion to obtain a plurality of second gear feature images; and performing three-dimensional modeling to obtain an initial three-dimensional harmonic reducer model, and performing calculation and analysis of the backlash.
Need to check novelty before this filing date? Find Prior Art