A machine vision–based geometric dimension measurement system for multitarget industrial workpieces is proposed to improve the efficiency and accuracy of conventional manual inspection. The system integrates image acquisition, dimensional calibration, image preprocessing, subpixel edge extraction, morphological refinement, and parameter measurement into a complete noncontact inspection framework. High-resolution workpiece images are captured by an industrial camera, followed by grayscale transformation and guided filtering to suppress noise while preserving edge details. Subpixel edge extraction based on gradient analysis and Zernike moments is then applied to improve contour localization accuracy. Morphological processing and connected-domain labeling are further used to separate multiple targets and calculate geometric parameters including edge length, contour perimeter, and aperture size. A calibration strategy is established to convert pixel coordinates into physical dimensions through a standard reference block. Experimental results demonstrate that the proposed system achieves stable repeatability, high consistency with manual measurement, and significantly improved inspection speed. Compared with conventional manual inspection, the proposed framework provides a practical and efficient solution for multi-target dimensional measurement in industrial environments.
Wang et al. (2026) studied this question.