A Machine Vision Based In-Line Quality Assessment Method for the Fabrication of Structured Surfaces
A Machine Vision Based In-Line Quality Assessment Method for the Fabrication of Structured Surfaces
复制标题
DOI：
10.1109/icac57885.2023.10275199
复制
发表时间：
2023-08
期刊：
影响因子：
--
通讯作者：
Maomao Wang;Wenbin Zhong;Guoyu Yu;Xiangqian Jiang;W. Zeng
中科院分区：
文献类型：
--
作者：
Maomao Wang;Wenbin Zhong;Guoyu Yu;Xiangqian Jiang;W. Zeng
结构表面工程因其独特的光学扩散和散射等功能而备受关注。然而，由于质量评估过程的复杂性，以低成本和高精度在微尺度上制造这些表面仍然具有挑战性。用表面形貌测量仪评定结构表面的传统方法效率低、成本高，不适合于表面缺陷的在线检测。针对这些挑战，本文提出了一种新的在线质量评估方法，该方法将Hough变换和图割算法相结合，以实现结构化表面的在线快速质量评估。所提出的方法可以在几秒钟内评估关键质量因素，如加工微结构的数量、尺寸和潜在缺陷，而不需要从机器上取出零件。该方法已成功集成到超精密钻石机中。通过对典型结构表面的在线评估，验证了该方法的有效性。该方法为高性能结构化表面的在线质量控制提供了一种快速、经济有效的解决方案。
The engineering of structured surfaces has gained significant attention due to their unique functionalities such as optical diffusion and scattering. However, the fabrication of these surfaces at the microscale with low cost and high accuracy remains challenging, primarily due to the complexity of the quality assessment processes. Traditional methods for assessing structured surfaces with surface topography measurement instruments are inefficient and costly and are not suitable for inline detection of surface defects. In light of these challenges, this paper presents a novel in-line quality assessment method that integrates the Hough transform and graph cut algorithms to enable fast on-machine quality assessment for structured surfaces. The proposed method enables the evaluation of key quality factors, such as the number, dimensions, and potential defects of the machined microstructures in seconds, without requiring the removal of the part from the machine. This method has been successfully integrated into an ultra-precision diamond machine. The effectiveness of the proposed method has been demonstrated through the in-line assessment of typical structured surfaces. The proposed method provides a potential solution to the fast and cost-effective in-line quality control of high-performance structured surfaces.