CUDA-based Acceleration Algorithms for heterodyne multiple frequency phase shifting technology

摘要

Heterodyne multiple frequency phase-shifting technology is a widely used optical 3-D shape profile measurement technique. In this paper, we introduce a GPU-based acceleration method into heterodyne multiple frequency phase-shifting technology using Compute Unified Device Architecture (CliDA). We first introduce background of corresponding parallel platform CUDA and analyze algorithms used in heterodyne multiple frequency phase-shifting technology, including phase unwrapping, phase-based matching and 3-D reconstruction. And we present a CUDA-based parallel implementation of integrating all these algorithms on GPU. Furthermore, various CUDA optimization techniques are also applied to maximize the utilization of the GPU. Finally, we compare the experimental results of CUDA programs on Gl'U with traditional programs on CPU. Experiments show that it obtains a speedup of approximately 10.3 times.