Oblique incidence lasertriangulation was used to detect medium and heavy plate flatness in this paper.The article mainly focused on point laser spot center extraction algorithm. Infielddetection of medium and heavy plate flatness, according to the problem of thepoint laser spot center extraction difficulty in interference condition, andfor enhancing lightness of spot center filed, an extraction algorithm of pointlaser spot center is put forward in this paper which combines threshold methodand square-weighted gray level based on floor board effect and traditionalcentroid method. Under no interference condition, point laser spot centerextraction using traditional centroid method has high accuracy. In the fielddetection with dusts, oil drops, water drops or other interferences,measurement precision has since fallen sharply. On this background, accordingto a kind of point laser spot center with the interference of water drops oroil drops, classification extraction algorithm of point laser spot center wasput forward in this paper. At the first level, the spot and the drop contourwas extracted with Sobel Operators. At the second level, the spot area wasextracted. And then, this spot area center was extracted by a point laser spotcenter extraction algorithm which combines threshold method and square-weightedgray level. And the center we get is point laser spot center. The validationresults show that with the water drop interference, the accuracy with a pointlaser spot center extraction algorithm which combines threshold method andsquare-weighted gray level is obviously improved by 94.17% than the traditionalcentroid method.