17 December 1996 Automated procedure for registering SAR and optical imagery based on feature matching
Author Affiliations +
Abstract
It is an accepted fact that combining data from different sensors can yield more information than the individual sensors will give when used on their own. This is especially true for synthetic aperture radar (SAR) data when combined with optical data. However, if performed manually, the registration of these two data sets can be time consuming and inaccurate. Due to the properties of SAR images, it is generally very difficult to select ground control points to use in the registration process. This paper covers the work being undertaken to develop an automatic system for registering SAR data to optical data using feature matching. In order to complete the registration, corresponding features have to be extracted from both images and matched with each other to generate a number of 'match points'. The parameters of the transformation can then be calculated using these match points. The successfully merged image will then be used to develop accurate change detection algorithms. Perhaps one of the most difficult aspects of this procedure is the extraction of features from SAR images, which are heavily degraded by speckle. Here we present the results of testing various speckle reduction filters and segmentation procedures in order to examine their ability to aid feature extraction. We have also developed a number of new algorithms which reduce speckle and segment SAR images.
© (1996) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Paul M. Dare, Ian J. Dowman, "Automated procedure for registering SAR and optical imagery based on feature matching", Proc. SPIE 2958, Microwave Sensing and Synthetic Aperture Radar, (17 December 1996); doi: 10.1117/12.262688; https://doi.org/10.1117/12.262688
PROCEEDINGS
12 PAGES


SHARE
Back to Top