TY - JOUR
T1 - Robust Space-Frequency Joint Representation for Remote Sensing Image Scene Classification
AU - Fang, Jie
AU - Yuan, Yuan
AU - Lu, Xiaoqiang
AU - Feng, Yachuang
N1 - Publisher Copyright:
© 1980-2012 IEEE.
PY - 2019/10
Y1 - 2019/10
N2 - Remote sensing image scene classification is a fundamental problem, which aims to label an image with a specific semantic category automatically. Recent progress on remote sensing image scene classification is substantial, benefitting mostly from the powerful feature extraction capability of convolutional neural networks (CNNs). Even though these CNN-based methods have achieved competitive performances, they only construct the representation of the image in location-sensitive space-domain. As a result, their representations are not robust to rotation-variant remote sensing images, which influence the classification accuracy. In this paper, we propose a novel feature representation method by introducing a frequency-domain branch to the traditional only-space-domain architecture. Our framework takes full advantages of discriminative features from space domain and location-robust features from the frequency domain, providing more advanced representations through an additional joint learning module, a property that is critically needed to perform remote sensing image scene classification. Additionally, our method produces satisfactory performances on four public and challenging remote sensing image scene data sets, Sydney, UC-Merced, WHU-RS19, and AID.
AB - Remote sensing image scene classification is a fundamental problem, which aims to label an image with a specific semantic category automatically. Recent progress on remote sensing image scene classification is substantial, benefitting mostly from the powerful feature extraction capability of convolutional neural networks (CNNs). Even though these CNN-based methods have achieved competitive performances, they only construct the representation of the image in location-sensitive space-domain. As a result, their representations are not robust to rotation-variant remote sensing images, which influence the classification accuracy. In this paper, we propose a novel feature representation method by introducing a frequency-domain branch to the traditional only-space-domain architecture. Our framework takes full advantages of discriminative features from space domain and location-robust features from the frequency domain, providing more advanced representations through an additional joint learning module, a property that is critically needed to perform remote sensing image scene classification. Additionally, our method produces satisfactory performances on four public and challenging remote sensing image scene data sets, Sydney, UC-Merced, WHU-RS19, and AID.
KW - Frequency domain
KW - joint representation
KW - remote sensing image classification
KW - robust
KW - space domain
UR - http://www.scopus.com/inward/record.url?scp=85078250690&partnerID=8YFLogxK
U2 - 10.1109/TGRS.2019.2913816
DO - 10.1109/TGRS.2019.2913816
M3 - 文章
AN - SCOPUS:85078250690
SN - 0196-2892
VL - 57
SP - 7492
EP - 7502
JO - IEEE Transactions on Geoscience and Remote Sensing
JF - IEEE Transactions on Geoscience and Remote Sensing
IS - 10
M1 - 8720267
ER -