AUTOMATIC REGISTRATION AND MOSAICKING SYSTEM FOR REMOTELY SENSED IMAGERY

Abstract

Image registration is an important operation in many remote sensing applications and it, besides other tasks, involves the identification of corresponding control points in the images. As manual identification of control points may be time - consuming and tiring, several automatic techniques have been developed. This paper describes a system for automatic registration and mosaic of remote sensing images under development at The National Institute for Space Research (INPE) and at The University of California, Santa Barbara (UCSB). The user can provide information to the system in order to speed up the registration process as well as to avoid mismatched control points. Based on statistical procedure, the system gives an indication of the registration quality. This allows users to stop the processing, to modify the registration parameters or to continue the processing. Extensive system tests have been performed with different types of data (optical, radar, multi-sensor, high-resolution images and video sequences) in order to check the system performance. An online demo system is available on the internet ( which contains several examples that can be carried out using web browser.