Design of saw blade contour positioning and external dimension measurement method based on machine vision
Abstract
In view of the problems of low efficiency and large subjective errors in traditional manual measurement of saw blade external dimensions, this paper proposes a method for saw blade contour localization and external dimension measurement based on machine vision. An image acquisition system is built using a high-resolution linear array camera and dual telecentric lenses to obtain saw blade contour images. Based on the Halcon platform, threshold segmentation, morphological opening operation, connected component analysis, and shape filtering are performed in sequence to achieve precise localization of the saw blade target. Then, Boolean subtraction and convex hull transformation are used to extract the feature region of the bottom edge. The pixel size is calculated through minimum bounding rectangle fitting based on second-order moment analysis, and converted to actual physical dimensions by combining the calibrated pixel equivalent. Experimental results on five groups of saw blades with different specifications show that the average absolute error of this method is -0.015 cm, and the average relative error is -0.20%. The accuracy is superior to traditional manual measurement and can meet the online, non-contact, and automated detection requirements of industrial assembly lines.