基于衍射理论的哈达玛编码光谱成像数据重构

Reconstruction of Hadamard coded spectral data based on diffraction theory

哈达玛变换光谱成像技术是一种在不牺牲分辨率的情况下,通过多通道复用的方法增加光学系统光通量,使得系统信噪比显著高于传统成像的新型计算光谱成像技术.本文针对系统成像过程中数字微镜器件衍射造成的图像降质问题,建立了基于标量衍射理论的哈达玛编码光谱成像图谱数据退化模型,提出以Lucy-Richardson(L-R)算法为核心的重构数据修正算法,提高重构数据立方体的图像质量和光谱精度.通过对成像过程的模拟和重构算法的仿真实验,数字化生成了系统成像衍射降质的物理过程,并验证了本文所采用修正方法的有效性.经过L-R修正之后的复原数据立方体的光谱角距离评价结果为0.1296,图像相似度评价因子优于0.85,相比修正前的重构数据质量有较大提升,表明算法对哈达玛编码光谱成像数据的重构及修正具有较好的效果.

Hardmard transfer imaging spectrometer(HTIS)is a novel computationally optical system.Its characteristic of multi-channel multiplexing increases the luminous flux of the optical system without sacrificing spatial resolution,thereby enabling the system’s signal-to-noise ratio to be significantly higher than traditional spectrometer’s.Encoding with digital mirror devices(DMD)in the system causes a serious diffraction effect that gives rise to the apparent degradation of the imaging formation.For improving the image quality and spectral accuracy of the reconstructed data cube,the Hadamard coded spectral imaging data degradation model is established based on the scalar diffraction theory.A data reconstruction algorithm is proposed based on the Lucy Richardson(L-R)algorithm.Through the simulation experiment,the process of image degradation is revealed.On the one hand,it proves that the degradation of system imaging diffraction is the main reason for the distortion of reconstructed data.On the other hand,it verifies the effectiveness of the correction method adopted in this paper.The evaluation result of the spectral angle distance of the restored data cube after L-R correction is 0.1296,and the image similarity evaluation factor is better than 0.85.Compared with the reconstructed data before being corrected,the corrected data is greatly improved in quality.The experimental results show that the algorithm has a good correction effect on the data cube reconstruction of HTIS.