视觉伺服对准方法、装置、系统、电子设备和程序产品
By decomposing visual servo alignment into multi-stage control, and utilizing shape features for attitude convergence and precise alignment of flexible and slender targets, the problem of angular feature failure when flexible and slender targets approach the target area in existing technologies is solved, achieving high-precision and stable alignment results.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- LINGXIN QIAOSHOU (BEIJING) TECH CO LTD
- Filing Date
- 2026-06-16
- Publication Date
- 2026-07-17
AI Technical Summary
In automated precision assembly, existing technologies struggle to meet the stability and accuracy requirements of high-precision alignment control between flexible, slender targets and target workpieces, especially when approaching the target area, where angular features are prone to failure.
Visual servo alignment is decomposed into three stages: coarse position alignment, shape convergence alignment, and fine position alignment. Multi-degree-of-freedom decoupled control is achieved by utilizing shape features. By acquiring the binary mask image of the workpiece and calculating the normalized shape descriptor using the second-order central moment, the attitude convergence and precise alignment of flexible and slender targets can be realized.
It improves the final alignment accuracy and the certainty and stability of control, solves the problem of angular feature failure, and ensures the success rate of contact action.
Smart Images
Figure CN122401063A_ABST