An Accelerated Algorithm for 3D Reconstruction of Groove Structure Based on Parallel Light and White Light Interference

摘要

We propose a three-dimensional image reconstruction acceleration algorithm based on parallel light white light interference system. We move the sample or mirror to be tested longitudinally to adjust the optical path, and then, collect and number the interference images in the 2um range near the position of the sample surface where the interference fringes are obvious. The group’s images around the upper and lower surfaces which are the interference images with the largest contrast in each group of images are partially reconstructed in 3D. Use the contrast information to segment the interference images with the largest contrast in each group of images collected, ensure that each sub-region can contain the maximum and minimum values of interference fringes on the iso-optical path surface. The contrast is compared with the contrast of the entire image, and the position of the iso-optical path surface is extracted, and the local 3D reconstruction results are spliced using the obtained position of the iso-optical path surface to complete the overall 3D shape detection. The reconstruction result of the algorithm in this paper is significantly improved compared with the reconstruction result directly using the differential centroid method. In terms of speed, the reconstruction time of the algorithm in this paper is 0.0429s, and the reconstruction time directly using the differential centroid method. is 1.540s.