一种板卡的异物检测方法、装置、设备及存储介质
By combining semantic segmentation and object alignment methods with depth and texture image data, and using template fusion objects for matching, the problem of low accuracy in foreign object detection on the board is solved, achieving high accuracy and low false alarm rate in foreign object detection.
CN117291892BActive Publication Date: 2026-07-17GUANGZHOU LEICHEN INTELLIGENT EQUIP TECH CO LTD
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- GUANGZHOU LEICHEN INTELLIGENT EQUIP TECH CO LTD
- Filing Date
- 2023-09-28
- Publication Date
- 2026-07-17
AI Technical Summary
Technical Problem
Existing technologies have low accuracy and a high false alarm rate when detecting foreign objects on circuit boards, making it difficult to meet the needs of high-quality production.
Method used
A foreign object detection method combining depth image data and texture image data is adopted. Through semantic segmentation and object alignment, template fusion objects are used for matching to identify foreign objects on the board.
Benefits of technology
It improves the accuracy of foreign object detection, reduces the false alarm rate, enhances the stability and versatility of detection, and adapts to the effects of changes in color, shape, and light.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN117291892B_ABST
Abstract
本发明公开了一种板卡的异物检测方法、装置、设备及存储介质，该方法包括：对目标板卡分别采集目标深度图像数据与目标纹理图像数据；在目标深度图像数据中执行语义分割，得到语义为元件或底板的目标深度对象；在目标纹理图像数据中执行语义分割，得到语义为元件或底板的目标纹理对象；将目标深度对象与目标纹理对象对齐，得到目标融合对象；将目标融合对象与模板融合对象进行匹配；若目标融合对象与任一模板融合对象均匹配失败，则检测目标融合对象是否属于目标板卡上的异物。减小了检测的对象，且丰富了特征的信息量，从而将检测的注意力集中到较小的对象上，可有效检测出目标板卡上的异物，提高检测异物的精确度，降低误报率。
Need to check novelty before this filing date? Find Prior Art