Research On Modulation Reconstruction Of Improved FP Micro-array Based On Compressed Sensing

With the continuous development of optical information acquisition,various types of optical information acquisition adopt integrated technology to achieve simultaneous acquisition of different target information.Spectral imaging technology is a new type of multidimensional imaging technology,which combines imaging technology and spectral analysis technology,which can simultaneously acquire the spectral and spatial two-dimensional information of the target scene.Therefore,spectral imaging technology is widely used in satellite imaging,military exploration,atmospheric environment monitoring,medical inspection and other fields.Aiming at the problem that high-throughput and high spectral resolution cannot be achieved simultaneously in existing space-borne scanning imaging,the paper proposes a combination of Fabry Perot(FP)micro-array and compressed sensing spectral imaging to achieve modulation of input spectral signals Obtain different spectral responses and finally obtain a hyper-spectral data cube.This thesis will mainly discuss the research from the following aspects.(1)Starting from the theory of compressed sensing in spectral imaging,the basic framework of compressed sensing is introduced in detail.It consists of three parts,namely the sparse representation of the signal,the selection of the measurement matrix and the signal reconstruction.In-depth analysis of the working principle of compressed sensing spectral imaging provides a solid theoretical basis for further realizing the original spectral reconstruction.(2)The paper proposes a fast scanning spectral imaging system based on Fabry Perot(FP)micro-array.The system puts an FP micro-array in front of the imaging detector.Each unit of the FP micro-array modulator corresponds to a different height,and then modulates the incident light signal to obtain different spectral responses,and the surface detector collects the modulated spectrum.The system scans the same location multiple times,which means that the system has a higher signal-to-noise ratio and luminous flux.Analyze the incident light signal and design a suitable measurement matrix,and perform theoretical analysis on the original spectrum reconstruction,and select the TWIST two-step iterative shrinking threshold algorithm to reconstruct the original spectrum information.(3)MATLAB is used to simulate and simulate the original spectrum reconstruction of two different lasers,including the spectrum modulation process and the spectrum compression and reconstruction process.And carried out a series of experiments to reconstruct two different wavelengths of monochromatic light.The simulation and experimental results show that the compressed sensing spectral imaging method has a higher compression ratio,improves the information collection speed and luminous flux,and has higher spatial resolution and spectral resolution,and avoids data waste.It proves the feasibility of this system,provides convenience for the spectrum imaging under the weak illumination environment on board,and has a wide range of application prospects.