Abstract: In order to solve the problem of extraction accuracy of brightness supersaturated laser stripe centerline in three-dimensional scanning of coal flow, a centerline extraction algorithm is proposed. According to the principle of oblique laser triangulation measurement, a visual measurement system platform was built to obtain the digital image of laser stripes with oversaturated brightness on the surface of coal piles and preprocess it. The distribution law of laser light intensity is explored, and the reason for the supersaturation of laser stripe brightness is analyzed. It is found that the traditional structured light centerline extraction method is not suitable for this study. Therefore, a center point extraction algorithm based on Gaussian curve fitting method is used for partial supersaturation, and a center point extraction algorithm based on neighborhood block idea is applied for total supersaturation. The centerline extraction strategy based on the number of supersaturated sampling points on the laser stripe cross section is proposed to obtain the fringe centerline with sub-pixel coordinate accuracy. By setting the contrast simulation experiment with the gray center of gravity method and curve fitting method, it is found that the mean absolute deviation of the proposed algorithm is smaller, and the extraction accuracy is better than the traditional algorithm. Finally, the experimental verification is carried out on the laser stripe image of coal heap, and the experimental results indicate that the proposed algorithm has stronger anti-noise interference ability in the actual working environment, which can reduce the influence of brightness supersaturation on the extraction of laser stripe centerline on the surface of coal heap.