Method for establishing improved defect-detection model based on YOLO model, defect-detection method and defect-detection system
TW202630261APending Publication Date: 2026-07-16TRIPOD TECHNOLOGY CORPORATION
View PDF 0 Cites 0 Cited by
Patent Information
- Authority / Receiving Office
- TW · TW
- Patent Type
- Applications
- Current Assignee / Owner
- TRIPOD TECHNOLOGY CORPORATION
- Filing Date
- 2025-01-06
- Publication Date
- 2026-07-16
Smart Images

Figure TWG2TA001067763_001 
Figure TWG2TA001067763_002 
Figure TWG2TA001067763_003
Abstract
A method for establishing an improved defect-detection model, a defect-detection method and a defect-detection system are provided. The defect-detection system includes a computer host that provides multiple improvement modules and uses a processing module to perform the method for establishing the improved defect-detection model based on YOLO model. In the method, bullseye chart images of a high-density interconnect printed circuit board (HDI-PCB) are retrieved, and various types of defects on the bullseye chart images are labeled. The labeled defects are matched with tags of defect types so as to establish a defect image dataset. The YOLO is selected to be a target object detection model that can be improved by the improvement modules. The defect image dataset is used to train the YOLO model so as to establish the improved defect-detection model for HDI-PCBs.
Need to check novelty before this filing date? Find Prior Art