一种基于极大似然估计的半导体Mark对称性识别方法
By using a maximum likelihood estimation method, the symmetry axis of semiconductor Mark points is automatically identified, solving the problems of high labor costs, cumbersome debugging, and high hardware requirements in existing technologies. This enables fast and simple Mark point identification, which is suitable for multi-site deployment.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHANGHAI GLORYSOFT CO LTD
- Filing Date
- 2023-04-27
- Publication Date
- 2026-07-17
AI Technical Summary
Existing Mark point recognition technologies suffer from high labor costs, cumbersome parameter tuning, long model training cycles, high hardware requirements, and low deployment efficiency, making it difficult to quickly identify Mark points of different types, shapes, and image features.
The method based on maximum likelihood estimation is adopted. Through image denoising, autocorrelation symmetry axis identification and calculation, and maximum likelihood estimation, the symmetry axis of the Mark point is automatically identified, including the identification and calculation of the horizontal and vertical symmetry axes, and outliers are removed to obtain accurate angle values.
It enables rapid and automatic identification of Mark points of different types and shapes. The identification of the axis of symmetry is simple, the hardware requirements are low, and it is suitable for rapid deployment at multiple sites, thus improving the identification efficiency and accuracy.
Smart Images
Figure CN117203666B_ABST