Research On Key Techniques Of Frequency Upconversion Imaging Based On Coding

Infrared imaging based on frequency upconversion is an indirect infrared imaging that the shortwave-infrared(SWIR)or the mid-infrared(MIR)is converted to visible or near-infrared(NIR)before detecting,which has advantages of the low noise of detectors,smaller pixels and the room-temperature working condition.The period poled nonlinear crystal and the quasi-phase-match theory contribute to breaking the limitation of effective nonlinear coefficient and improving the upconversion efficiency.Thus,the upconversion imaging based on the quasi-phase-match develops.The doctoral dissertation focuses on solving the problem that it is hard to optimize the point spread function,the upconversion efficiency and the imaging speed at the same time in the upconversion imaging based on the quasi-phase-match by combining the upconversion imaging technology and the single-pixel imaging theory.The main contents of this dissertation are as follows:To solve the problem that the point spread function and the upconversion efficiency at the same time can not be optimized,the frequency upconversion imaging based on coding(UIBC)is proposed after research the factors limiting the the point spread function of frequency upconversion imaging system.Firstly,the infrared is coded by the spatial light modulator according to the single-pixel imaging.Secondly,the modulated infrared interacts with the pump beam in the nonlinear crystal and the wavelength of infrared is converted.Then,the single-pixel detector measures and records the converted infrared.In the end,the converted image is reconstructed with the measurements and the measurement matrix.In order to optimize the upconversion efficiency,the pump beam is focused to increase it’s power density in the nonlinear crystal.To improve the imaging speed,the temperature of nonlinear crystal maintains a constant.The experimental results shows the object at 1550 nm is converted with 2W pump beam and the image has a sharp edge.To ruduce the measurements times and optimize the imaging time,the multi-resolution frequency upconversion imaging based on coding(MR-UIBC)is proposed.Firstly,the low-resolution image is acquired by UIBC.Then,the positions of wavelet coefficients of the object detail are predicted according to the multi-resolution analysis(MRA)and measured at higher resolution.In the end,the converted image is reconstructed via low resolution converted image and the converted infrared of the measured wavelet coefficients.Thus,imaging time is decreased by reducing the measurement time.The experimental results shows the measurement times of a 128×128 converted image is reduced 73.68% comparing UIBC.To reduce the noise,the frequency upconversion photon-counting imaging based on coding is proposed.The signal-noise-ratio is improved by analysis the noise resource and use the time gate divide the converted signal from the measurement according to the temporal characteristics of the converted signal.The experimental results shows the peak signal-noise-ratio is improved 30.91%.The peak of the dark noise is less than 0.34 photons/spatial/s element for an image with 64×64 pixels by taking into account 10 ms measurement time for each pattern.