Compressed Sensing Angorithm And Technology Research For Super-Resolution Imaging

As a cross-frontier area of nature science and technology development,nanotechnology plays a very important role in a number of research areas,including construction of new electronics on the nano-scale aspects of new photonics,synthesis of new surface of the catalyst,as well as new nano fluorescent markers,etc.The development of nanotechnology is inseparable with nanoscopy imaging technology.In this regard,the electronic scanning microscope and atomic force microscope have been as a representative.However,with the continuous development of nanotechnology,researchers want to know not only the appearance of the structure and morphology in nano-scale,but also the electronic structure,phonon and plasmon and other microscopic internal state in the same scale.However,since the resolution of conventional optical microscope is limited by the diffraction limit,a resolution of no less than ?/(2NA),so that the imaging resolution is generally greater than 180 nm.In recent years,as the representative of the super-resolution imaging technology,the development of stimulated emission depletion(STED)and stochastic optical reconstruction imaging(STORM)have completely break the optical diffraction limit restrictions,make the optical microscopy resolution go to the nanometer scale.STORM technology has the feature of low excitation intensity and wide field of super-resolution imaging,so it has been used to do imaging analysis of nanoscale living biological tissues,cells and their sub-structures,which cannot be observed in the original optical microscopy,and it can do extremely sensitive analysis of interactions between molecules in very low concentrations of sample,so it has been rapidly applied to bio-medical field,achieved an further research on nano-sized biological tissue structure.To achieve fast super resolution imaging for STORM,this paper proposes a new type of super-resolution imaging algorithm based on compressed sensing,and compare it with the existing reconstruction algorithms,discuss the relationship between different molecular density and fluorescence recall rate,the localization error,the results showed that with the increase of the number of iterations,block CS algorithm is superior to CVX algorithm;in the aspect of analyzing complete analog STORM image,the reconstruction time of block CS increased by 3 to 4 orders of magnitude;this paper also designs different measurement matrix to compare the different remodeling results between different measurement matrices,the results showed that random matrix as the measurement matrix can get a better effect on the reconstruction.Related laboratory equipment has been purchased.This article provides a theoretical basis to the build of an experimental imaging equipment.