Huang Feng, Huang Duanhao, Yang Shumin, Wang Pengfei
Opt Express. 2024 Feb 12;32(4):5444-5459. doi: 10.1364/OE.513502.
The Risley-prism imaging system (RPIS) is a powerful way to achieve bionic human eye imaging with great advantages on large field of view (FOV) and variable resolution imaging owing to the autonomous controlled deflection of light. But the imaging dispersion originating from nonlinear and uneven light deflection results in limited imaging wavelength that seriously hinders its application. The existing solutions for imaging dispersion mainly rely on the hardware, which generally has bulky structure and limited improvement on image. Besides, the existing image evaluation methods for dispersion are not suitable for RPIS due to inhomogeneous dispersion. Herein, this paper systematically analyzes the mechanism and characteristics of dispersion in the RPIS, and proposes a cooperative correction method for image distortion and dispersion of multiple-color imaging, achieving the elimination of distortion and dispersion simultaneously without changing the optical structure. A dispersion evaluation index based on Pearson's correlation coefficient (PCC) is also established, and the objectivity and validity of the index are proved by experiments. Furthermore, a kind of compact RPIS based on an RGB camera is built, and both indoor and outdoor experiments are conducted. The experimental results demonstrate that proposed algorithm has strong universality and robustness for various scenes and targets.
里斯利棱镜成像系统（RPIS）是实现仿生人眼成像的一种强大方式，由于光的自主控制偏转，在大视场（FOV）和可变分辨率成像方面具有很大优势。但是，由非线性和不均匀光偏转引起的成像色散导致成像波长受限，严重阻碍了其应用。现有的成像色散解决方案主要依赖硬件，硬件结构通常庞大，对图像的改善有限。此外，由于色散不均匀，现有的色散图像评估方法不适用于RPIS。在此，本文系统地分析了RPIS中色散的机制和特性，提出了一种多色成像的图像畸变和色散协同校正方法，在不改变光学结构的情况下同时消除畸变和色散。还建立了基于皮尔逊相关系数（PCC）的色散评估指标，并通过实验证明了该指标的客观性和有效性。此外，构建了一种基于RGB相机的紧凑型RPIS，并进行了室内和室外实验。实验结果表明，所提出的算法对各种场景和目标具有很强的通用性和鲁棒性。
