离焦光源的傅里叶单像素成像及其优化

Fourier single-pixel imaging with defocused light source and its optimization

傅里叶单像素成像是一种新型光学成像技术,在三维成像等领域具有重要应用前景。为了促进傅里叶单像素成像的应用,本文分析研究光源的离焦对傅里叶单像素成像的影响并探索离焦成像的优化策略。首先基于描述光源离焦的高斯模型,理论分析离焦光源的傅里叶单像素成像,接着实验上以傅里叶基底图案调制的数字光投影仪为光源,通过改变光源的离焦参量,分析研究光源离焦对傅里叶单像素成像的峰值信噪比(peak signal-to-noise ratio,PRSN)的影响。结果表明:光源偏离焦点越多,成像质量越差,具体为重建图像PRSN从28.4 dB下降至20.5 dB。为此,本文探索实现了一种融合自适应滤波和盲去卷积的成像质量优化策略,优化后的图像PRSN提高了1.4—4.0 dB。该结果对于傅里叶单像素成像应用于远距离成像等复杂环境具有重要意义。

As a novel imaging technique, Fourier single-pixel imaging has potential applications in many fields such as 3 D imaging.In order to promote the application of Fourier single-pixel imaging, we investigate the impact of light source defocusing on Fourier single-pixel imaging and its optimization strategy in this paper.Applying Gaussian model to describe the defocusing, we analyze theoretically Fourier single-pixel imaging with defocused light source.Then employing digital light projector modulated by Fourier basis pattern as the light source, we investigate the influence of the defocusing on the peak signal-to-noise ratio PRSNof Fourier single-pixel imaging by changing the defocused parameter of the light source.The results show that the more the light deviates from the focus of light field, the worse the imaging quality.Correspondingly, the PRSNof the reconstructed image decreases from 28.4 dB to 20.5 dB.In terms of this problem, we explore an adaptive strategy to optimize the defocused image which combines adaptive filter with blind deconvolution, the PRSNof the optimized image improved by 1.4—4.0 dB.It is helpful to the application of Fourier single-pixel imaging in complex environments such as long-distance imaging.