An efficient orthorectification of a satellite SAR image used for monitoring occurrence of disaster

Abstract

Synthetic aperture radar (SAR), which is one of active type sensors, allows us to obtain satellite images regardless day or night. In addition it has a feature of all-weather-observation because it emits L-band microwave which is less affected by clouds and rain. The PALSAR-2 aboard the ALOS-2, which achieves high-resolution (3m) observation, is now in operation as the satellite L-band SAR in Japan. The utilization of advantage of the satellite SAR image for monitoring the occurrence of a natural disaster has been tried. The PALSAR-2 emits microwave obliquely downward and receives the back-scattered portion of the incident microwave on the surface of the earth as the observed data. Accordingly, it is quite difficult to decide the occurrence of disaster only with the satellite SAR image by just glancing unlike using a satellite photograph taken from the zenith. In addition, since the satellite SAR image is geocoded image, the actual position on the earth is affected by the location of the satellite corresponding to the time it was observed. Therefore, the process deciding the occurrence of natural disaster using plural information is required to be carried out after a process assigning a correct data based on the obtained satellite SAR image. This process is called orthorectification. However, since the size of the satellite SAR image which is used for detecting occurrence of natural disaster is quite large (larger than 1 GB), an efficient orthorectification process is required to rapidly provide an input data to the detecting process of natural disaster. In this paper we propose a method for an efficient orthorectification process.