3D inspection technology combining passive stereo matching and active structured light for steel plate surface sample

Three-dimensional (3D) inspection technology is a new important research hotspot in the field of steel surface defect inspection. However, current 3D information acquisition methods (i.e., the passive stereo vision method and active structured light method) still suffer from several issues. In order to solve these issues, a three-dimensional inspection system is designed in this paper. In this system, the passive stereo vision method and active structured light method are combined to obtain the surface sample height information of high temperature steel plate. Furthermore, the Hexagonal grid Census (Hg_Census) transform is proposed to improve the robustness of census transform stereo matching. Moreover, the passive stereo disparity map is adopted as a constraint condition to realise phase matching using wrapped phase. In addition, the local phase matching and sub-pixel disparity refinement are proposed to obtain high measuring accuracy. Three high temperature steel plate surface samples are picked to verify the effectiveness of the proposed method. The actual experimental results present the average error is less than 1 mm.