Portable Machine Vision System for Measuring Polished Metallic Surface Roughness

Traditional methods for inspecting polished metallic surfaces such as molds rely on human vision, which leads to inconsistent results and inefficiency. Here, a portable machine vision system is proposed that optimizes key lighting parameters including the illumination angle, light source distance, and height to accurately measure the surface roughness of polished metal molds. Sixty configurations with different design parameters were tested, and the optimal setup was determined by using a cost function balancing the gray-scale average (GSA) contrast and system size. Validation experiments with polished metal samples of known roughness values revealed a strong correlation between the GSA and surface roughness. The proposed system demonstrated an accuracy comparable to that of traditional contact-based methods while offering enhanced portability, faster measurements, and suitability for industrial environments.