光学学报, 2021, 41 (2): 0210001, 网络出版: 2021-02-27
改进最小费用流相位解包裹算法 下载: 942次
Improved Minimum Cost Flow Algorithm for Phase Unwrapping
图像处理 电子散斑干涉 最小费用流 相位解包裹 残差点 image processing digital speckle pattern interferometry minimum cost flow phase unwrapping residues
摘要
在不同解包裹算法中,最小费用流(MCF)解包裹法可以限制残差点误差远程扩散,并将误差优先限制在低相干区域,有利于保证高相干区域解包裹结果不受干扰,精度较高,但残差点数量较多时计算效率很低。为缩短解包裹时间,提出一种残差点预处理方法。该方法将残差点视作正负电荷,通过电场力,引导距离较近的异号残差点互相抵消,大幅减少残差点数量,从而提高解包裹计算效率。仿真数据和实验数据表明,残差点预处理对MCF解包裹精度影响很小,在残差点数量超过3000时可以大幅提高解包裹计算效率。
Abstract
In different unwrapping algorithms, the minimum cost flow (MCF) unwrapping method can limit the long-range diffusion of the residue error. Further, it can first limit the error to the low coherent region, which ensures that the unwrapping results in the high coherence region are not disturbed, and its accuracy is high. However, when the number of residues is large, the calculation efficiency is very low. To decrease the unwrapping time, a residue preprocessing method is proposed. In this method, the residues are regarded as positive and negative charges. Moreover, the electric field force is used to guide the residues with different signs close to each other to mutually offset, thus significantly reducing the number of residues and improving the efficiency of unwrapping calculations. Simulated and experimental data show that the residue preprocessing has little influence on the phase unwrapping accuracy. When the number of residues exceeds 3000, the residue preprocessing significantly improves the phase unwrapping efficiency.
邵珩, 周勇, 聂中原, 祁俊峰. 改进最小费用流相位解包裹算法[J]. 光学学报, 2021, 41(2): 0210001. Heng Shao, Yong Zhou, Zhongyuan Nie, Junfeng Qi. Improved Minimum Cost Flow Algorithm for Phase Unwrapping[J]. Acta Optica Sinica, 2021, 41(2): 0210001.