Normalized iterative denoising ghost imaging based on the adaptive threshold
Normalized iterative denoising ghost imaging based on the adaptive threshold
复制标题
基于自适应阈值的归一化迭代去噪鬼影成像
DOI：
10.1088/1612-202x/aa555e
复制
发表时间：
2017-02-01
影响因子：
1.7
通讯作者：
Liu, Baolei
中科院分区：
文献类型：
--
作者：
Li, Gaoliang;Yang, Zhaohua;Liu, Baolei
提出了一种提高鬼像质量的方法。本文通过理论分析，建立了基于归一化GI的迭代模型。在迭代模型中选择自适应阈值。估计迭代模型的初始值作为去除相关噪声的步骤。仿真和实验结果表明，该策略在不增加复杂度的情况下，比传统的和归一化的GI重建更好的图像。NIDGI-AT方案不需要关于对象的先验信息，并且还可以自适应地选择阈值。更重要的是，重建图像的信噪比（SNR）大大提高。因此，这种方法是迈向实际应用的又一步。
An approach for improving ghost imaging (GI) quality is proposed. In this paper, an iteration model based on normalized GI is built through theoretical analysis. An adaptive threshold value is selected in the iteration model. The initial value of the iteration model is estimated as a step to remove the correlated noise. The simulation and experimental results reveal that the proposed strategy reconstructs a better image than traditional and normalized GI, without adding complexity. The NIDGI-AT scheme does not require prior information regarding the object, and can also choose the threshold adaptively. More importantly, the signal-to-noise ratio (SNR) of the reconstructed image is greatly improved. Therefore, this methodology represents another step towards practical real-world applications.