TY - GEN
T1 - Remote sensing image registration based on feature points of global edge
AU - Liu, Siying
AU - Jiang, Jie
N1 - Publisher Copyright:
© 2017 IEEE.
PY - 2017/7/1
Y1 - 2017/7/1
N2 - In this paper, in view of the grayscale information and morphological changes of remote sensing images, a registration algorithm based on the distribution information of the global edge and the local information of the feature points is proposed. Lifting wavelet transform is used for multi-scale edge extraction of stable global edge, such as river and coastline. On the basis of SIFT feature point detection, the Equivalent DoG (EDoG) pyramid is constructed and the scale-invariant feature points on edge are extracted. The SIFT descriptor is combined with the global descriptor to construct Global-SIFT (G-SIFT) descriptor. The algorithm is tested on the remote sensing images before and after severe natural disasters such as earthquakes, floods and storms. Compared with the performance of SIFT algorithm, the algorithm has advantages in terms of algorithm stability, matching accuracy and registration precision. It can meet the registration accuracy requirement in remote sensing image with large changes.
AB - In this paper, in view of the grayscale information and morphological changes of remote sensing images, a registration algorithm based on the distribution information of the global edge and the local information of the feature points is proposed. Lifting wavelet transform is used for multi-scale edge extraction of stable global edge, such as river and coastline. On the basis of SIFT feature point detection, the Equivalent DoG (EDoG) pyramid is constructed and the scale-invariant feature points on edge are extracted. The SIFT descriptor is combined with the global descriptor to construct Global-SIFT (G-SIFT) descriptor. The algorithm is tested on the remote sensing images before and after severe natural disasters such as earthquakes, floods and storms. Compared with the performance of SIFT algorithm, the algorithm has advantages in terms of algorithm stability, matching accuracy and registration precision. It can meet the registration accuracy requirement in remote sensing image with large changes.
KW - Remote sensing images
KW - feature points on edge
KW - global feature description
KW - image registration
UR - https://www.scopus.com/pages/publications/85049411239
U2 - 10.1109/IST.2017.8261564
DO - 10.1109/IST.2017.8261564
M3 - 会议稿件
AN - SCOPUS:85049411239
T3 - IST 2017 - IEEE International Conference on Imaging Systems and Techniques, Proceedings
SP - 1
EP - 6
BT - IST 2017 - IEEE International Conference on Imaging Systems and Techniques, Proceedings
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2017 IEEE International Conference on Imaging Systems and Techniques, IST 2017
Y2 - 18 October 2017 through 20 October 2017
ER -