Abstract
The registration of multiresolution optical remote sensing images has been widely used in image fusion, change detection, and image stitching. However, traditional registration methods achieve poor accuracy in the registration of multiresolution remote sensing images. In this study, we propose a framework for generating deep features via a deep residual encoder (DRE) fused with shallow features for multiresolution remote sensing image registration. Through an L2 normalization Siamese network (L2-Siamese) based on the DRE, the multiscale loss function is used to learn the attribute characteristics and distance characteristics of two key points and obtain the trained feature extractor. Finally, the DRE is used to extract the deep features of the key points and their neighbors, which are concatenated with the shallow features into a fusion feature vector to complete the image registration. We performed comprehensive experiments on four sets of multiresolution optical remote sensing images and two sets of synthetic aperture radar images. The results demonstrate that the proposed registration model can achieve subpixel registration. The relative registration accuracy improved by 1.6%–7.5%, whereas the overall performance improved by 4.5%–14.1%.
Highlights
W ITH rapid development of remote sensing technology in recent years, remote sensing images are advancing toward multiresolution and multispectrum
The algorithm framework presented in this article comprises three main sections: an L2-Siamese model used to train the deep residual encoder (DRE) network; a DRE network that extracts the deep features of key points and their neighborhoods; and a combination of deep features and scale-invariant feature transform (SIFT) descriptors to generate feature vectors and the multiresolution remote sensing image registration
It shows that the deep feature distribution of an image generated by the auto-encoder differs significantly after transformation, and its features are unsuitable for image registration
Summary
W ITH rapid development of remote sensing technology in recent years, remote sensing images are advancing toward multiresolution and multispectrum. Improved ground observation requires the integration of heterogeneous remote sensing data, and multiresolution remote sensing image registration is fundamental in the field of remote sensing image processing. The goal is to make any pair of pixels in a multiresolution remote sensing image at the same location represent the same geographic location [1], [2]. Manuscript received June 28, 2020; revised August 21, 2020 and October 19, 2020; accepted November 3, 2020. Date of publication November 19, 2020; date of current version January 6, 2021.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
