AI-driven surface roughness measurement

Surface roughness measurement is essential for ensuring optical components, semiconductor substrates, and similar surfaces meet quality and performance standards. Precise characterization and control of surface irregularities enhance overall efficiency and performance. This paper introduces an AI-driven solution for surface roughness measurement in the image space, offering a versatile optical metrology tool. It offers a rapid, cost-effective integration, eliminating the need for scanning in most applications. AI algorithms initially determine the fundamental surface figure. They then extract higher spatial frequencies related to surface roughness using AI-driven phase retrieval techniques. This approach enables comprehensive surface roughness profiling and estimation of statistical parameters like Ra and Rq, crucial for quality control. The solution is ideal for high-throughput applications, ensuring precise surface quality assessment. In conclusion, this AI-driven solution revolutionizes surface roughness measurement, benefiting industries working with optical components, semiconductor. This paper introduces the method and results using a Point Source Microscope (PSM) from OPG.