一种压缩编码孔径成像光谱偏振技术的研究

One hot topic in the optics field is acquiring the image, spectral and polarization information of objective simultaneously, accurately and effectively. In the project, the compressive coded aperture imaging spectropolarimetry(CCAISP), which is based on the coded aperture spectral imaging and channeled spectropolarimetry, is proposed to acquire the image, spectra and the full Stokes polarization information of the static or dynamic scenes by one shot or multi-shot snapshot measurements. The theoretical study includes the imaging modeling and information decoded reconstruction. The numericalexperiment is implemented to analyze the effect of the factors on the system and the optical experiment is implemented to verify the plan of the project. There are many advantages of CCAISP. CCAISP has higher throughput and SNR comparing with the dispersive imaging spectropolarimetry (ISP) based on the slit. CCAISP is feature with without pushbrooming and snapshot comparing with the traditional pushbroom ISP. Comparing with the computed tomography ISP, the large area focal plane array is unnecessary and there is no the “missing cone” problem. Comparing with the coded aperture spectral imager(CASI), the full stokes polarization information is acquired besides the image and spectra information taken by CASI. CCAISP proposed us an alternative approach to the optical information sensing and the realization of compressive sensing by hardware. It is a meaningful and potential technology for the remote sensing, biomedical sciences, and so on.

同时、精确、高效地获取图像、光谱和偏振信息是目前光学领域的研究热点。本项目通过研究一种基于编码孔径光谱成像技术和通道光谱偏振技术的压缩编码孔径成像光谱偏振技术，实现一次或者多次曝光的快照成像同时获取静态或者动态目标的图像、光谱和全Stokes偏振信息。在理论方面，主要研究系统成像模型和信息解码重构算法；在实验研究方面，通过数值实验对各种误差对系统的影响进行分析，通过光学实验对于系统可行性进行验证。该技术具有以下优点：相比基于狭缝的色散型成像光谱偏振技术，具有高通量和高信噪比；相比传统推扫型成像光谱偏振技术，无需推扫，具有快照成像的特点；相比计算层析型成像光谱偏振技术，无需大面阵焦平面阵列且没有“信息失锥问题”；相比编码孔径成像光谱技术，除了得到图像、光谱信息之外，还有全Stokes偏振信息。为光学信息探测和压缩感知的硬件化实现提供了一种新方法，在遥感探测、生物医学等领域具有广阔的应用前景。

空间、光谱和偏振信息是人们进行目标识别的三类主要的应用信息。空间信息直观地提供了目标的外形信息、光谱信息提供了物质的成分信息、偏振信息则告知我们目标的表面粗糙度和边缘信息。在精准目标识别方面，光谱信息和偏振信息的作用愈加重要。传统基于通道光谱偏振技术的成像光谱偏振技术中，在数据处理环节中首先对于调制后的光谱进行傅里叶变换，获得通道化的干涉图，在干涉图域进行空间通道滤波，得到子通道的干涉图，再通过子通道干涉图的逆傅里叶变换获得目标的Stokes偏振光谱信息。在解调过程中存在通道混叠问题，从而影响光谱精确度；另一方面由于进行了干涉图域的空间滤波，会导致光谱分辨率降低。为了克服这些问题，本项目研究了一种压缩感知通道成像光谱偏振技术，完成的主要内容有：通过将压缩感知成像机制和通道光谱偏振技术相结合，建立了系统的压缩感知线性模型，运用压缩感知重构算法直接重构得到了目标的空间、光谱和偏振信息。克服了传统通道光谱偏振技术中存在的通道混叠、光谱分辨率降低、光谱通道数减少的问题。该技术可以实现快照式的成像，可以应用于动态目标的探测。我们在实验室搭建了实验系统，实现了12个波段的全Stokes偏振信息的静态和动态目标探测。在生物医学、刑侦监测、环境保护等领域具有广泛的应用前景。