Machining Tool Wear Detection and Measurement Based on Edge Extraction and Subpixel Fitting

Tool wear has a significant impact on machining quality of products. This study proposes a novel method for detecting machining tool wear and measuring wear size based on edge extraction and subpixel fitting. First, the tool image is enhanced by homomorphic filtering and contrast-limited adaptive histogram equalization (CLAHE). Second, a maximum interclass variance method based on the Laplacian of Gaussian (LoG) operator is proposed for image segmentation. The complete edge of the tool is then extracted with a gradient method based on multiple morphological combinations. Then, image registration based on Gaussian pyramid acceleration with maximum connected domain method is used to extract tool wear region edges. Finally, the Lagrange interpolation is used for subpixel edge detection to obtain the tool wear edges. The subpixel edge points are fit by the soft k-segment principal curve to obtain the maximum wear width. The experimental results show that the proposed method can achieve micrometer-level accuracy, which is higher than those of existing methods. It can also meet the real-time requirements in industrial applications.