Machine vision-based recognition of elastic abrasive tool wear and its influence on machining performance

This study presents a novel Hunter-Prey Optimization (HPO)-optimized Otsu algorithm in tool wear assessment and machining process quality control. The algorithm is explicitly tailored to address the challenges conventional image recognition methods face when identifying the unique wear patterns of elastic matrix abrasive tools. The proposed HPO-optimized Otsu algorithm was validated through machining experiments on silicon carbide workpieces, demonstrating superior performance in wear identification, image segmentation, and operational efficiency when compared to both the conventional 2-Dimensional (2D) Otsu algorithm and the Genetic Algorithm (GA)-optimized Otsu algorithm. Notably, the proposed algorithm reduced the average runtime by 36.99% and 28.39%, and decreased the mean squared error by 24.78% and 20.52%, compared to the 2D Otsu and GA-optimized Otsu algorithms, respectively. Additionally, this study investigates the influence of elastic tool wear on abrasive machining performance, offering valuable insights for assessing tool status and life expectancy, and predicting machining quality. The high level of automation, accuracy, and fast execution speed of the proposed algorithm makes it an attractive option for wear identification, with potential applications extending beyond the manufacturing industry to any sector that requires automated image analysis. Consequently, this study contributes to both the theoretical comprehension and practical application of tool wear assessment, providing significant benefits to industries striving for enhanced production efficiency and product quality.