Three-Dimensional Reconstruction Method Based on Infrared Deflectometry

Abstract Deflectometry is a non-contact and rapid detection method with high sensitivity, which can be utilized in the areas of three-dimensional (3D) reconstruction, surface quality detection, and defect location. It has important applications in the automotive industry, optical parts manufacturing, and other fields. Classical deflectometry methods require the integration of slope data to reconstruct the absolute surface shape. However, the integration procedure is error-prone due to the accumulation of random errors into large shape deviations, resulting in inaccurate measurement results. In this paper, we extract the position of the centerline of the bright fringe from the stripe image and find the center of the fringe line point by point. Using the principle of triangulation, the 3D data for the object’s shaped surface can be obtained. In addition, infrared light enhances specular reflection, resulting in better deflectometry performance than that of visible light. Experiments show that the proposed method is sensitive to the changes in height and can roughly restore the surface topography of the object without calibration.