The Study Of Radiometric Calibration And Error Estimation Of Space-based Infrared Camera

This paper is based on the processing requirements for infrared atmosphericbackground measurement data, which were downloaded from the atmosphericbackground measurement satellite to the ground application system. According to thespecific applications of the space-based infrared camera, a solution for full-pixelradiometric calibration and the corresponding processing methods for calibration datawere proposed. Based on the calibration data, a linear model of calibration coefficient, adouble exponential model of gray-scale decaying for space-based cameras and acalibration error model were built. The main contents of the paper and the researchresults are as follows:In the aspect of the full-pixel radiometric calibration method, the principle, systemcomposition and testing program were proposed. By using the processing methodsproposed above, the data from pre-launch laboratory calibration and on-boardcalibration were processed. The data contained various combinations of spectrum, gainand exposure time by laboratory calibration and numerous tests by on-board calibrationin half a year. Based on the methods of pre-processing and non-effective pixeldiscrimination, the full-pixel calibration method was built and used.In the aspect of full-pixel calibration results analyzing, by using the calibrationcoefficients, the radiance images were inversed and evaluated. By analyzing thelaboratory calibration data, a linear model between the coefficients and the cameraparameters was put forward. By comparing the on-board calibration data, the problemscaused by space environment were noticed.In the aspect of the space contaminate, aimed at the phenomenon of increasingabnormal calibration data, a solution of automatic processing was proposed on frameand pixel information. To deal with the problem of the gray-scale decaying, a doubleexponential model was built on space contaminate damage theory, by transforming thetotal working hours into optical damage factor. The model parameters were corrected byon-board calibration data.In the aspect of the calibration errors modeling and analyzing, error estimationmodels for laboratory and on-board calibration were built based on calibration data.Using this model, calibration errors were calculated and analyzed. The performance ofthe space-based camera was also evaluated.In the aspect of designing and realizing, the software was written by the need of theground application system. With the idea of modularization, this paper developed afull-pixel radiometric calibration processing and analysis software under the build environment of MATLAB. The software included calibration data processing, resultsbrowsing and error analyzing part. This software was incorporated into the groundapplication system.The analysis and research of this paper were designed to provide a reliable andeffective radiometric calibration solution and the corresponding methods for processingdata. All above were some meaningful theories and models, which were helpful to thequantitative and engineering application for the infrared remote sensing data.