High performance two-dimensional phase unwrapping on GPUs

Abstract

Phase unwrapping is an important procedure in digital image and signal processing, and has been widely used in many fields, such as optical and microwave interferometry, magnetic resonance imaging, synthetic aperture radar, adaptive optics. Phase unwrapping is a time consuming process with large amount of calculations and complicated data dependency. A number of algorithms with different features have been developed to solve this problem. Among all of them, Goldstein's algorithm is one of the most widely used algorithms, and has been included in some standard libraries (such as MATLAB). In this paper we propose an innovative implementation of Goldstein's algorithm on GPU. Several important approaches and optimizations are proposed for the GPU algorithm. For example, by introducing a localmatching step, we were able to parallelize the branchcut step efficiently, getting much better performance than existing work. With a cascaded propagation model, another important operation in the algorithm, f loodf ill, is able to make good use of the computing power of GPU. We tested our GPU algorithm on NVIDIA C2050 and K20 GPUs, and achieved speedup of up to 781 and 896 over the CPU implementation respectively. To the best of our knowledge, this is the best performance of unwrap ever achieved on GPUs.

A Parallel Implementation of The Rotational Compensator Phase Unwrapping Algorithm

“…The speedup of this method is 7.3 over the CPU. In the method by Wu et al the GPU is used to accelerate Goldstein's phase unwrapping algorithm on a machine with an Intel Xeon X5550 CPU and an NVIDIA Tesla C2050 GPU [37]. The GPU has 14 multiprocessors with 448 cores; each of the multiprocessors has 64KB cache/shared memory.…”

“…The methods of Karasev et al and Mistry et al [35,36] used the least-squares algorithm with discrete cosine transforms. Moreover, the methods [32][33][34]37] used the Goldstein and the branch-cut algorithms. The performance evaluation of the RC phase unwrapping algorithm among Goldstein and Least-Squares algorithms shows that RC has higher accuracy [6]; though it has high time cost than them.…”

“…Hence, we choose the RC phase unwrapping algorithm to be the first contribution accelerating its process. Furthermore, the methods implemented by López et al, Huang [35][36][37] have a GPU to accelerate the phase unwrapping algorithms. Therefore, multicore systems have become more commercially prevalent; we prefer using CPU in our work.…”