Research on high-precision workpiece size measurement based on grayscale moment algorithm

In response to the issues of low accuracy and significant errors in manual measurements of high-precision workpieces, this paper proposes a high-precision, low-error dimension measurement algorithm. Focusing on triangular workpieces as the detection target, the proposed algorithm first applies threshold segmentation to the workpiece image, and then it employs an improved grayscale moment algorithm to identify sub-pixel edges. Subsequently, it uses an eight-neighborhood algorithm combined with Hough transform for rough line fitting and, finally, employs the RANSAC algorithm along with the least squares method for precise line fitting. After multiple measurements of the workpiece, experimental results demonstrate that the proposed algorithm achieves high-precision dimension measurement with a minimum error of 0.467 μm and an average error of 1.439 μm.