A Machine Vision Based In-Line Quality Assessment Method for the Fabrication of Structured Surfaces

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.