Abstract: The surface defects of the workpiece weld bead were detected suing a point laser displacement sensing method. Firstly, the data of weld defect surface was collected and the original data was filtered by Gaussian filtering. Then the Delaunay triangulation was applied for the processed data points and reconstruction of a three-dimensional model with the spatial coordinates. Experimental results showed that the 3D reconstruction method based on the point laser displacement sensing and the surface defect images of the weld bead can accurately judge the appearance of welds.