Orthorectification of SAR Data using Digital Elevation Model (DEM) Data and Image Registration Technique based on Fast Fourier Transform (FFT)

摘要

The need for remote sensing data to monitor the vast area of Indonesia, from year to year has increased by a variety of spatial, spectral and temporal resolution. In addition to optical remote sensing satellite data, radar data which acquired using active SAR (Synthetic Aperture Radar) sensors must be optimized because it has several advantages that can penetrate clouds and can operate day and night in all weather conditions. Nevertheless, the SAR satellite imagery basically still contains many distortions/errors caused by data acquisition system with a sideway direction (side looking imaging geometry) and the effect of various height of objects on the earth surface. To correct/eliminate such distortions, orthorectification correction should be conducted as a pre-processing before the data is used in further applications, especially in applications that require high geometric accuracy. This research aims to make development of orthorectification method with a consideration that the process can be conducted automatically. First step, forming of simulation image is conducted by using SRIM-DEM (Shuttle Radar Topography Mission - Digital Elevation Model). The next step is image-to-image registration process between original SAR image and the simulation image. In this research, automatic technique based on Fast Fourier Transformation (FFT) is used to find the optimum tie-points. And the last step is image reconstruction from each pixel in the original SAR image to its new position in the specified map projection system. The method have been tested using ALOS PALSAR data on the area around Merapi - Merbabu Mountain, Central Java, Indonesia and will be packaged in the computer modules/programs that can be used by many users particularly in Indonesia. With the pre-processing method, it is expected that the remote sensing image data with higher quality and standardized can be obtained and then can be used in many remote sensing applications.