Key Technologies And Applicactions Of Compressed Ultrafast Photography

Ultrafast photography plays an indispensable role in biomedicine,nuclear explosion,photochemistry,applied fluidics and mechanical processing.However,the imaging speed of conventional charge-coupled device?CCD?or complementary metal oxide semiconductors?CMOS?camera has been limited due to the restrictions of electronic storing and reading velocities,which make it impossible to capture transient processes that occur on nanoseconds or even shorter time scales.In order to capture such transient processes,scientists have developed various ultrafast optical imaging technologies involving compressed ultrafast photography?CUP?.Unlike active ultrafast imaging technologies that require specific illumination light or the pump-probe technology that requires multiple measurements,CUP is a single-shot and receive-only imaging technology,its temporal resolution and number of frames can reach one hundred femtosecond and several hundreds,respectively.For some self-luminous or non-repeatable phenomena,such as laser-induced shock waves,optical rogue waves,light scattering in living tissues,irreversible crystalline photochemical reactions,etc.,CUP has great advantages,and these advantages can be attributed to the novel CUP model that combines compressed sensing theory and space-time conversion technology.Despite such advantages,CUP also has some problems,such as poor imaging quality and expensive hardware systems.In this dissertation,we took compressed sensing theory and optical imaging principle as starting points,and proposed some improved schemes in algorithms and hardware to promote the CUP imaging performance.In addition,we also analyzed the advantages of the CUP model and applied it to the fields of image information security and ultrafast electron diffraction imaging to promote its application in related research fields.Therefore,the main contents in this dissertation are summarized as follows:1,In aspect of mathematical theory,we proposed three strategies based on compressed sensing to improve the image reconstruction quality of CUP,namely reducing the coherence between the measurement operator and the dynamic scene,increasing the sampling rate,and optimizing the reconstruction algorithm.To reduce the coherence,we developd the optimizing code scheme by a genetic algorithm,which can greatly reduce the noise of the images and increase the images'correlation coefficient by approximate 1%.To increase the sampling rate,we developed a multi-coding CUP scheme,which can not only increase the spatial resolution but also break through the temporal resolution of the time deflector.To optimize the reconstruction algorithm,we developed the augmented lagrangian algorithm to replace the original two-step iterative shrinkage/thresholding algorithm,which can not only improve the quality of the reconstructed images but also is almost independent of the selection of relevant parameters,which can make image reconstruction more robust.2,In aspect of hardware,streak camera,as an important tool for space-time transformation in the CUP,because of its working principle based on photon-electron-photon conversion and the high cost,restricts the practicality of CUP.Aiming at such problem,we developed a scheme to replace the streak camera with an electro-optical deflector and a CMOS camera.This scheme achieved three-dimensional spatiotemporal imaging with an imaging speed of 5×10¹⁰ fps?frames per second?,horizontal and vertical spatial resolutions of 0.79 and 0.89 lp/mm?lp,line pairs?.In addition,aimging at CUP's incapability of spectral information,we developed a hyperspectrally compressed ultrafast photography?HCUP?technology by combining compressed sensing theory and spectral technology,obtaining four-dimensional?x-y-t-??imaging,with horizontal and vertical spatial resolution of 1.26 and 1.41 lp/mm,temporal and spectral frame interval of 2 ps and 1.72 nm,respectively,which realized the breakthrough of optical imaging technology from three to four dimensions in a single shot.3,In aspect of model's applications,we have analyzed the advantages of the novel CUP model and made some applications.First of all,CUP utilizes a piece of code to act on the three-dimensional data,which can save the amount of coding.Based on this advantage,we combined the CUP model with quantum key distribution to develop a three-dimensional image encryption technology,which can save 8.7 times of the code usage and increase the codes generation rate by about 3 times.Secondly,CUP is a single-shot technology without jitter during an exposure.Based on this advantage,we applied the CUP model to the field of ultrafast electron diffraction that is greatly affected by jitter,and theoretically proposed the compressed ultrafast electron diffraction imaging technology by using a long electron pulse as the probe,and also simulated the effects of electron density,code size and electron energy on single crystal and polycrystalline electron diffraction patterns,which provided a theoretical basis for ultrafast structural dynamic detection through a single exposure without jitter.