Katkovnik Vladimir, Astola Jaakko
Department of Signal Processing, Tampere University of Technology, P. O. Box 553, Tampere 33101, Finland.
J Opt Soc Am A Opt Image Sci Vis. 2012 Aug 1;29(8):1556-67. doi: 10.1364/JOSAA.29.001556.
The computational ghost imaging with a phase spatial light modulator (SLM) for wave field coding is considered. A transmission-mask amplitude object is reconstructed from multiple intensity observations. Compressive techniques are used in order to gain a successful image reconstruction with a number of observations (measurement experiments), which is smaller than the image size. Maximum likelihood style algorithms are developed, respectively, for Poissonian and approximate Gaussian modeling of random observations. A sparse and overcomplete modeling of the object enables the advanced high accuracy and sharp imaging. Numerical experiments demonstrate that an approximative Gaussian distribution with an invariant variance results in the algorithm that is efficient for Poissonian observations.
考虑使用用于波场编码的相位空间光调制器（SLM）进行计算鬼成像。从多个强度观测值重建透射掩模幅度物体。使用压缩技术以便通过比图像尺寸小的观测次数（测量实验）获得成功的图像重建。分别针对随机观测的泊松和近似高斯建模开发了最大似然风格算法。对物体进行稀疏和超完备建模可实现先进的高精度和清晰成像。数值实验表明，具有不变方差的近似高斯分布导致的算法对于泊松观测是有效的。
