A precision metal part edge positioning method based on gray evolution trend continuity
By using an edge localization method based on grayscale evolution trends, combined with gradient calculation and weighted centroid method, the accuracy and stability issues of edge localization for highly reflective metal parts are solved, achieving sub-pixel-level edge localization, which is suitable for highly reflective metal parts in precision manufacturing.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SUZHOU FIELD TECHNOLOGY GROUP CO LTD
- Filing Date
- 2026-05-20
- Publication Date
- 2026-07-17
AI Technical Summary
Traditional edge detection methods struggle to achieve stable and accurate edge positioning on highly reflective metal parts, especially in chamfered transition areas, failing to meet micron-level accuracy requirements and being affected by the irregularity of scattered light and grayscale changes.
The edge localization method based on the continuity of grayscale evolution trend uses gradient calculation and sliding window processing of one-dimensional grayscale profile sequence, combined with weighted centroid method and AI pseudo-edge removal, to identify vertical surface feature intervals with consistent gradient direction and stable amplitude, and calculate grayscale energy center to achieve sub-pixel level edge localization.
It significantly improves the accuracy and stability of edge positioning, enabling sub-pixel-level positioning accuracy on highly reflective metal parts, adapting to different reflective conditions, reducing scattered light interference, and is suitable for high-precision applications in the field of precision manufacturing.
Smart Images
Figure CN122415580A_ABST