Compressive Hyperspectral Imaging Mask Optimization

With the development of imaging theory and charge coupled device manufacturing technology,people are not satisfied with obtaining grayscale information but prefer to obtaining the information of additional dimensions in the view field.The spectrometer can obtain spectral information of the field of view,and the current spectral resolution can reach the nanometer level.Different from traditional eameras which you can only get the integral of brightness of one point in the view field,in this paper,a spectral acquisition device based on coded aperture is proposed,which can obtain the Integral curve of spectral in the view field.The values of different spectral bands are obtained,thereby obtaining the spectral information of the view field.Snapshot imaging technology based oa coded aperture is characterized by compact structure and fast imaging speed,which has become a research hotspot in the field of Multimedia imaging.In this paper,we have detailed research on compressive hyperspectral imaging technology based on coded aperture from the following aspects:First,the compressive hyperspectral imaging described in this paper combines traditional 2D imaging with spectral imaging techniques,such ihaX dozens or hundreds of spectral bands continuously imaging the objects in the view within the spectral coverage,make sure obtain spatial and spectral infonnation at the same time.Secondly,the method of compressing hyperspectral imaging which is placed mask in front of the sensor is discussed,and the method of improving the quality of spectral reconstruction by optimizing the mask is discussed,the method described in this paper can effectively improve the PSNR of the reconstructed spectral image and improve the accuracy of spectral data.Finally,the effect of the blur caused by the defocus of the mask on the spectral coding in the actual imaging system is proposed,we also analyzed this effect and quantified the relationship between the degree of defocus and the accuracy of the reconstructed data.A compressive hyperspectral imaging based on defocus mask coding is proposed,and compare the results of spectral reconstruction without considering mask defocus.