Automatic remote sensing imagery registration based on improved SIFT

Huawei Li, Chongguang Zhu, Fengping DiProceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE 2007 · DOI

Abstract

With the rapid development of remote sensing technology and the current diversity of new sensors, we can get more and more multi-source remote sensing data in the same area. It is the key to match those remote sensing data in wide application. And most current methods of registering remote sensing imagery require human involvement, so, there is a need for automated techniques in the remote sensing area. Especially the multi-source and multi-temporal remote sensing data call for new demands for automated techniques. In this paper, the authors proposed a feature matching approach based on improved SIFT (Scale Invariant Feature Transform) for remote sensing imagery registration. In this method, the feature mainly is the specific characteristic of the remote sensing imagery. The experiment results using this automatic remote sensing registration method indicate that the improved descriptors are more distinctive and robust, which results in a higher matching performance and a better matching efficiency.
BibTeX
@inproceedings{li2007automatic,
  author = {Huawei Li and Chongguang Zhu and Fengping Di},
  title = {{Automatic remote sensing imagery registration based on improved SIFT}},
  booktitle = {Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE},
  year = {2007},
  doi = {10.1117/12.774762},
}

Loading related content...

Figure