Ultrafast Dynamic Defect Inspection With Computational Neuromorphic Imaging

利用计算神经形态成像技术进行超快速动态缺陷检测

阅读:2

Abstract

Inspecting surface defects is critical for ensuring quality in advanced manufacturing. However, current defect inspection techniques are largely limited to ideal imaging conditions and require vibration-free environments when using traditional industrial cameras. These frame-based cameras typically exhibit latency and a limited dynamic range, leading to vulnerabilities in production quality and restricting their efficiency in meeting increasing production demands. To address this challenge, a novel inspection paradigm is developed, using computational neuromorphic imaging (CNI) for ultrafast dynamic defect inspection. By leveraging the high temporal resolution and high dynamic range of event-based sensors, CNI enables ultrafast, exceptional dynamic range yet cost-effective defect inspection while minimizing perceptual and computational latency. Experimental results demonstrate event processing sampling times spanning 300-fold in fast motion and dynamic ranges exceeding 10000-fold under varying illumination. Additionally, event-driven data facilitates direct edge detection and visualization of defects, which is critical for real-time continuous diagnosis. Vibration is also utilized to enhance structural defect detection and improve robustness in practical settings. CNI offers unique advantages for industrial applications, enabling defect inspection across diverse products and complex conditions. This approach opens up a new route toward ultrafast industrial inspection in challenging environments, addressing the growing demands of mass real-time inspection and intelligent diagnosis in precision manufacturing.

特别声明

1、本页面内容包含部分的内容是基于公开信息的合理引用;引用内容仅为补充信息,不代表本站立场。

2、若认为本页面引用内容涉及侵权,请及时与本站联系,我们将第一时间处理。

3、其他媒体/个人如需使用本页面原创内容,需注明“来源:[生知库]”并获得授权;使用引用内容的,需自行联系原作者获得许可。

4、投稿及合作请联系:info@biocloudy.com。