Precision SAR processing for Radarsat

The Radarsat Precision Processor (RPP) is a project to develop a workstation-based Radarsat SAR processing system. The primary purpose of the RPP is to produce SAR images of the highest possible quality from Radarsat and also ERS-1 signal data. Image quality specifications include the following: maximum radiometric error of ±0.05 dB within a scene, impulse response width broadening of less than 3%, and phase errors of less than 5 degrees. These image quality parameters must apply for squint angles up to 8 degrees. The RPP uses a modified range-Doppler algorithm with secondary range compression performed in the azimuth-frequency domain. In this paper, descriptions of these algorithms are given. Analyses of the algorithms were performed to determine the expected throughput and image quality performance. In addition to SAR processing capabilities, the RPP also contains extensive display and analysis tools for both signal and image data. The signal display and analysis tools allow the user to assess the signal data quality through I and Q statistics and replica analysis. The image display and analysis tools allow the user to measure the effects on the image when processing parameters or algorithm choices are varied. The image analysis tools include point target analysis such as impulse response measures and distributed target statistics. Some details of these analysis and display tools are also given