数字全息显微中常见重建算法比较

Comparison of Commonly Used Numerical Reconstruction Algorithms in Digital Holographic Microscopy

基于理论分析和实验验证相结合的方法,对数字全息显微术中常见的三种重建算法即菲涅耳变换算法、角谱算法和卷积算法做了比较研究。结果表明:利用菲涅耳变换算法对离轴无透镜傅里叶变换数字全息进行重建时,无重建距离的限制;采用卷积重建法只能在最佳再现距离附近一个非常小的范围内才能获得高分辨率再现像;而采用角谱重建法在略小于最佳再现距离及大于最佳再现距离较大范围内重建,均能获得高分辨率的再现像。角谱重建法总体上优于卷积重建法。菲涅耳变换重建法简单、快捷,是优化的重建算法。

Based on therotical analysis and experimental demonstration, three commonly used reconstruction algorithms in digital holographic microscopy, Fresnel transform algorithm, angular spectrum algorithm, and convolution algorithm, are investigated. It shows that the lensless Fourier transform digital hologram can be reconstructed at any distance by Fresnel transform algorithm. For convolution algorithm, the whole reconstructed image with high resolution can be acquired in an optimal reconstruction range or a small area around it. When the reconstruction distance is slightly smaller or larger than the optimal one, a high quality image can also be obtained by using angular spectrum algorithm. Angular spectrum algorithm is better than convolution algorithm as a whole. The Fresnel transform algorithm is time-efficient and simpler than the other two.