Adaptive Lighting for Complex Cutting Tools in Precise Vision Measurement
Yang Liu,Wei Liu,Wenqi Wang,Ke Wang,Yang Zhang
DOI: https://doi.org/10.1109/tim.2024.3476557
IF: 5.6
2024-01-01
IEEE Transactions on Instrumentation and Measurement
Abstract:Vision-based measurement of cutting tool geometry has become prevalent in the industrial field, playing a crucial role in intelligent manufacturing. However, traditional illumination methods struggle to adapt to the complex structures of emerging cutting tools. Consequently, captured images for measurement often exhibit uneven illumination, low contrast, and missing features, resulting in inaccurate and even infeasible measurements. In this article, we propose a novel framework of adaptive lighting for precise and efficient vision measurement. First, an illuminance model for a multipartition light source is established considering multiple incident angles, enabling meticulous control of lighting patterns. Subsequently, a simplified image-based reflection estimation method is constructed to correlate lighting parameters with imaging results for various measured objects. Finally, a light control method based on convex optimization is devised to enhance image uniformity and contrast for specific interested features. The efficacy of the proposed methods is validated by a series of experiments. Compared to traditional illumination methods, experiments demonstrate a significant reduction in measurement error for complex drill bits from an average of 1.69%-0.18%, and a decrease in measurement repeatability deviation from 1.06% to 0.09%. With excellent feasibility and high robustness, our method is poised for widespread application in precise measurement and optical inspection of objects with complex geometry.