The importance of image quality in AI visual inspection systems: Why variability is the key challenge in model training

被引：0

作者：

Dolberg, Sagi ^{[1
]}

机构：

[1] QualiSense, Building A, 126 Yigal Alon, Tel Aviv,6744332, Israel

来源：

PhotonicsViews | 2024年 / 21卷 / 04期

关键词：

Inspection;

D O I：

10.1002/phvs.202400027

中图分类号：

T [工业技术];

学科分类号：

08 ;

摘要：

The use of AI for visual inspection is currently making great strides in sectors like automotive, metals and machinery manufacturing. However, those investing in the technology not only expect a system that will be superior to manual inspection, but an error rate close to zero in defect detection. Let us look at the importance of image quality and why the variability in the production environment is the single greatest challenge that AI engineers must face. © 2024 Wiley-VCH GmbH.

引用

页码：31 / 33