Research On Data Processing And Image Fusion Algorithms Of Photon Counting Imaging System

Solar-blind ultraviolet photon counting imaging technology is an ultra-high sensitivity ultraviolet detection technology.In this technology,Avalanche Photo-Diode(APD)working in Geiger mode is usually used as the detection device,supplemented by optical and displacement equipment to realize the detection of the surrounding environment,and its sensitivity can reach the limit of optical detection.Due to the special "sun-blindness" characteristics of the solar-blind ultraviolet band,this photoelectric detection technology plays an important role in many fields such as national defense,police forensics,and high-voltage power grid security detection.Solar-blind ultraviolet photon counting imaging system is usually composed of optics and detection system,photon counting readout circuit system and image processing system.Among them,the optical and detection system and the photon counting readout circuit system complete the detection and conversion of the weak solar-blind ultraviolet singlephoton signal,and generate a digital signal which is convenient for subsequent processing.The image processing system collects the detected digital signals,designs image processing algorithms according to actual application requirements,and completes the detection of solar-blind ultraviolet target.With the rapid development of front-end hardware systems and the diversification of application scenarios,the accuracy and quality requirements of solarblind ultraviolet photon counting imaging are increasingly required.However,because it is a new type of imaging system,the existing image processing algorithms cannot be fully applied to solar-blind ultraviolet photon counting imaging,so it is of great significance to carry out research on related algorithms.Based on the hardware of 4H-SiC APD and photon counting readout circuit independently developed by the research group,combined with the practical application requirements of high-voltage power grid safety detection,this paper designs an image processing system applied to the solar-blind ultraviolet photon counting imaging system.Firstly,combined with the working principle and performance characteristics of the hardware equipment,the corresponding data acquisition scheme and image preprocessing algorithm are designed to extract the effective information of the solar-blind ultraviolet target from the complex noise.Secondly,combined with the wavelength characteristics of solar-blind ultraviolet band,this paper designs an image registration algorithm based on the binocular model to achieve precise registration between solar-blind ultraviolet detection images and visible images.Finally,on the basis of precise registration,an image fusion algorithm based on multi-scale decomposition is designed to achieve high-quality fusion between heterogeneous images.The effective information in the solar-blind ultraviolet band and the visible light band is merged into one image,and the solar-blind ultraviolet target which is invisible to the human eyes is located in the visible light image to realize the detection and imaging of the solarblind ultraviolet target.All digital image processing algorithms designed in this paper are simulated and verified on the MATLAB R2020 a software platform.This paper collects multiple sets of representative and comparative real detection data,and validates and analyzes the designed image preprocessing algorithm,image registration algorithm and image fusion algorithm.When the detection distance is between 5 meters and 30 meters,the registration error of the image registration algorithm based on the binocular model designed in this paper can be controlled within 3 pixels.The image fusion algorithm based on multi-scale decomposition designed in this paper is superior to common image fusion algorithms in terms of information entropy,average gradient,standard deviation and spatial frequency.The running time of the fusion algorithm is about 1s,which meets the needs of practical applications.In the laboratory testing environment,the image processing system designed in this paper can better complete the detection and location of the solar-blind ultraviolet target,output images for subsequent observation and analysis,and possess certain solar-blind ultraviolet photon counting detection and imaging capabilities.