Font Size: a A A

Research On Low Light Level High Speed Imaging Technology Based On Coding Perception

Posted on:2022-03-13Degree:MasterType:Thesis
Country:ChinaCandidate:L Y RenFull Text:PDF
GTID:2558307070956499Subject:Optical engineering
Abstract/Summary:PDF Full Text Request
In night vision imaging system,the sensitivity and frame frequency are mutually restricted.In target detection under low-light-level(LLL)environments,the motion blurring caused by low frame rate greatly increases the difficulty of target tracking and recognition.This makes it difficult for low-light-level device/assembly to be deployed on high-speed reconnaissance platforms such as uav.In this paper,a Photon-limited coded aperture compressive temporal imaging(PL-CACTI)method is proposed based on the combination of low-light-level imaging and coded aperture compressive temporal imaging technology.By means of coding awareness,the frame frequency of low-light-level imaging can be improved without reducing the sensitivity of imaging system.In this paper,a high sensitivity and low frame rate camera is used to measure the high-speed low-light-level image modulated by spatial light modulator.The high-speed low-light-level image is reconstructed from aliasing sampling based on compressed sensing.Due to the translation of the coded aperture during each exposure time means that each temporal plane in the video stream is modulated by a shifted version of the code.Thus,each pixel is modulated without the additional bandwidth of low-light-level sensor.The Photon-limited coded aperture compressive temporal imaging of four coded aperture types is simulated and analyzed,and verified by the mainstream GAP-TV algorithm De SCI algorithm.Based on the frame-wise peak signal-to-noise ratio(PSNR)and structural similarity(SSIM)of reconstructed low-light-level high-speed images,the performance of CACTI is quantitatively analyzed and compared by four groups of video data sets with different compression ratios(Cr).Finally,according to the experimental requirements,random Gaussian matrix was selected for encoding,and Cr is 4 and 8,and the GAP-TV algorithm was used for reconstruction.In this paper,it is constructed that a hybrid measurement model of LLL device/assembly with Poisson noise and Gaussian white noise.The performance of PL-CACTI is analyzed by simulation.The simulation results show that CACTI performs well in LLL environments where PSNRs are more than 20 d B and SSIMs are more than 0.8.On this basis,the prototype of PL-CACTI is built.The experimental results show that PL-CACTI can achieve the time resolution of 100 fps from an electron multiplying CCD with 25 fps,and the reconstructed images can meet the practical requirements.The research results of this paper can provide theoretical basis and engineering guidance for the research of high-speed low-light-level imaging device/assembly.
Keywords/Search Tags:Compressive sensing, Coded aperture, low-light-level environments, Temporal imaging
PDF Full Text Request
Related items