一种多波长梯度加速相位恢复迭代算法

发布时间：2019-03-12 12:58

【摘要】：在传统多波长迭代算法的基础上,引入角谱传输理论和梯度加速函数,提出一种快速收敛的相位恢复迭代算法——多波长梯度加速迭代算法。该算法利用不同波长的光源经过相同光路在固定位置处所探测的光场强度信息,通过迭代逼近,并在迭代过程中引入梯度加速方法,提高收敛速度,从而恢复出输入面的相位信息。构建仿真模型,随机选取初始迭代值,对已知相位分布的输入场进行相位恢复,并与传统多波长迭代算法进行了比较,该算法表征相位面复原精度的相对均方根值均达到10-3数量级,收敛速率提高了两倍以上。在635~785nm波段内选取多组不同波长数量的入射光利用该算法进行对比实验,结果表明当波长数量为7、8时该算法具有良好快速的收敛特性及高精度的相位恢复能力。
[Abstract]:Based on the traditional multi-wavelength iterative algorithm, the angular spectrum transmission theory and gradient acceleration function are introduced, and a fast convergence phase recovery iterative algorithm, multi-wavelength gradient accelerated iterative algorithm, is proposed. In this algorithm, the intensity information of the light field with different wavelengths detected by the same light path in a fixed position is used, and the gradient acceleration method is introduced in the iterative process to improve the convergence rate. Thus the phase information of the input plane can be recovered. The simulation model is constructed, the initial iteration values are selected randomly, and the phase recovery of the input field with known phase distribution is carried out, and compared with the traditional multi-wavelength iterative algorithm. The relative root mean square (RMS) values of the proposed algorithm are all in the order of 10 ~ 3, and the convergence rate is more than two times higher. The experimental results show that when the number of wavelengths is 7,8, the algorithm has good convergence performance and high-precision phase recovery ability when the number of wavelengths is 7,8. The experimental results show that the proposed algorithm has good convergence performance and high-precision phase recovery ability when the number of wavelengths is 7,8 in the 635~785nm band.
【作者单位】：南京理工大学电子工程与光电技术学院;
【基金】：国家自然科学基金(61377015,61505080) 江苏省自然科学基金(BK20150788)
【分类号】：O436.1

【相似文献】