Study On The Information Processing Technology Of Geostationary Orbit Large Scale Staring Camera

The geostationary orbit staring camera can perform high frame rate imaging on fixed position,so it has the advantages of high temporal resolution and high detection sensitivity,and plays an important role in short time and small target detection.As the core part of the camera,IRFPA detector has problems of nonuniformity and response drift,which will lead to the large spatial noise in the infrared image and reduce the detection ability of the camera.In addition,the increasing pixel scale and bit rate put forward requirements on information processing circuit and signal processing technology.In this paper,the staring camera system as the research object was concerned.The problems of the system's spatial nonuniformity and infrared detector response drift were deeply studied.Specifically,effective methods to suppress temporal noise and spatial noise were figured out.On this basis,a set of high speed and low noise information acquisition circuit system was designed and implemented,according to the application requirements of super large array infrared detector.The main research contents include the following aspects:The imaging model of the staring camera,the photoelectric conversion process and the signal transmission link were modeled.On this basis,the temporal and spatial noise model and the noise suppression method of the camera were analyzed.The model provided a theoretical support for the nonuniformity correction methods and the implementation of low noise information acquisition circuit system.The nonuniformity correction method can effectively restrain the nonuniformity of the system and suppress spatial noise of the system.In this paper,nonuniformity correction algorithm based on blackbody calibration was deeply studied.In order to solve the problem that the multi-point correction method needs to store a large number of correction parameters in the engineering application,an improved nonuniformity correction method was proposed.In this method,the threshold value was determined quantitatively from the correction result,which reduced the blindness of the threshold selection.The experiment results showed that this method could effectively reduce the number of correction parameters in the premise of ensuring the effect of correction.For the traditional correction methods based on blackbody calibration,the correction effect will be deteriorated when the integration time of the detector does not coincide with the integration time when the correction parameters are acquired.Aiming at this problem,this paper proposed a nonuniformity correction method that could be adapted to the change of integration time.This method performed multi-point correction on both the radiation flux and the integration time,thus effectively solving the problem.The experimental results show that the mean of local nonuniformity of image corrected by the proposed method was less than 0.2% when the integration time was different from that of the parameter acquisition,which was better than other methods mentioned in the paper.The stability of the detector response is a prerequisite for its engineering and quantification applications.The response drift will reduce the nonuniformity correction accuracy of the system.Based on experimental data,this paper analyzed the influence factors of the detector response drift.A fast screening method was proposed to solve the problem of screening pixels with unstable response rate,which did not need to change the blackbody temperature and improved the convenience of operation.Considering the pixels with larger response drift showed a discrete distribution characteristics,a correction method based on adjacent normal pixels is proposed in this paper.The method used the data of multiple scenes to modify the parameters of pixels with larger response drift.The mean of local nonuniformity of the image decreased by 10.89% after parameter modification.A set of high speed and low noise information acquisition circuit system was designed and implemented in this paper.The schematic and PCB layout and routing had been optimized in analog circuit design in order to suppress the circuit noise,and high-speed serial bus for board-level interconnection was used in digital circuit design.The test results showed that the circuit noise was less than 0.1mV,and the dynamic range was greater than 100 dB,meeting design requirements.