Research And Implementation Of Remote Sensing Data Radiometric And Geometric Correction Algorithm

The on-board remote sensing data preprocessing becomes particularly important,since the technical advance of remote sensing image processing.In space environment,the onboard processing system had some limited conditions,so it blocked to achieving real-time processing for the remote sensing data preprocessing.In order to realize remote sensing data preprocessing on the DSP and FPGA platform and build real-time satellite remote sensing data preprocessing system,this dissertation was focus on studying the relative radiometric algorithm optimization,Mapping data block storage method and data selective read and write methods and for parallel processing optimized geometric correction block algorithm.All of main works are as follows:1,For real-time processing system,designed the relative radiometric correction method.This method combines the characteristics of scaling and statistical methods commonly used,takes into account the correction of real-time and accuracy.2,Geometric correction process large amount of data read in random positions.To improve the data access efficiency,designed optimized geometric correction block algorithm and redesigned the block resample method.3,The relative radiometric and geometric correction processes have to repeat read and write data blocks.To effectively improve the data read and write rates,designed data block storage method and data selective read and write methods.4,Through the study of remote sensing data preprocessing algorithm for system hardware and software,designed a system based on DSP and FPGA architectures for data processing.The research results can be served as foundation for the subsequent development of on-board real-time processing for remote sensing data preprocessing.5,The traditional FPGA development process takes a long time and a high degree of difficult.To improve the efficiency of implementation,used HLS to designed some algorithm modules.