Research On Cell Image Super-resolution Technology Based On On-chip Cell Detection System

Cell testing has a key guiding significance in the diagnosis of people's health.Lensless cell microscopic imaging and analysis technologies provide new technologies for modern medicine and life sciences.In view of the problem of the low pixel count of the lensless cell image microscopy system,it is of great significance to study the super-resolution reconstruction technique based on the on-chip lensless cell detection system.Aiming at the problem of too low pixels in the cell image collected by the lens-free array sensor image acquisition system,this paper proposes and builds an on-chip lensless cell image acquisition system based on line-array CMOS image sensor for super-resolution scanning imaging.Acquisition and image reconstruction.In order to analyze the factors affecting the quality of scanned reconstructed images,firstly,a spatial image transformation and a mean-based point-spread-function downsampling method were used to establish the acquisition imaging model based on a line-array image sensor super-scanning mode.This model was used to analyze the effects of three factors,such as the pixel fill factor of the linear array image sensor,the tilt angle of the linear array image sensor,and the speed of the target flowing in the microfluidic channel,on the super-resolution scan imaging quality.The results show that with the decrease of the fill factor,the peak signal to noise ratio of the reconstructed image increases.When the velocity decreases and the angle decreases,the peak signal-to-noise ratio of the reconstructed image increases and the imaging quality improves.On this basis,there is Fresnel diffraction problem for the imaging system with lens-free cell detection system.The diffraction model is established by the transfer function in the frequency domain.Combined with the super-scan imaging model,the image after super-resolution scanning is adopted.Blind deconvolution restoration method,interpolation-based distortion correction and diffraction recovery were implemented to establish a lens-free cell detection system model with diffraction function.After completing the entire super-resolution system,20-microscopically collected red blood cells were used for experiments to verify the accuracy of the system imaging model and super-resolution reconstruction.Anticoagulant blood collected from human veins was used for experiments.The cells were imaged by systematic super-scanning to obtain distorted diffracted cell images.Normal destructive cell images were obtained after blind deconvolution,distortion correction,and diffraction recovery.The obtained cell images were compared with those obtained by 4x and optical microscopes,respectively,to illustrate the feasibility and effectiveness of the super-resolution method.