| Spectral imaging and depth imaging have been explored by more and more research institutions and manufacturers due to their great application potential in detection and autonomous driving fields.In order to break the bottleneck of hyperspectral data collection and transmission,hyperspectral image data is collected by means of coding compression.In order to collect spectral and depth information simultaneously in a system,the optical path structure of the system needs to be cleverly designed.In this paper,two systems are proposed and constructed,namely,the Miniaturized Coded Snapshot Spectral Imager(CASSI),and the binocular Spectral depth imaging system composed of a gray camera combined with the miniaturized Coded Snapshot Spectral Imager(CASSI).These two systems in turn realize hyperspectral data compression and simultaneous acquisition of spectral and depth information.By means of theoretical analysis and experimental verification,the reconstruction effects of the two systems on spectral accuracy,spatial accuracy and depth information are demonstrated.The main work and achievements of this paper are as follows:1.A miniaturized CASSI system with dual dispersion was proposed and experimentally implemented.Spectral data were encoded by prism and mask compression,and hyperspectral information was carried out by single compression measurement,and reconstructed by TWIST algorithm.The system has high spatial fidelity and high spectral accuracy.In experimental results,the correlation coefficient between the reconstructed spectra of feature points and the spectral truth value can reach more than 0.85.In comparison with CASSI system with single dispersion,spatial and spectral results are better.The system is miniaturized and its volume is reduced by nearly half.2.The experimental spectral depth imaging system was built.By combining the miniaturized dual-dispersion CAS SI system with another gray camera,a binocular cross-mode spectral depth imaging system was constructed.The spectral data block is reconstructed by TWIST algorithm.The depth information is reconstructed by semi-global block matching algorithm.The system has good spectral accuracy and spatial accuracy,and the correlation coefficient between reconstructed spectra and truth value is about 0.85.In terms of depth information,scene texture features and depth information can be better captured.The reconstructed depth image is compared with the reference depth image,and the point cloud is compared with the reference image,which shows a good reconstruction effect.The reconstruction results of spectral image and depth image verify that the experimental system has the ability of collecting spectral information and depth information simultaneously. |