Change Detection Based on Supervised Contrastive Learning for High-Resolution Remote Sensing Imagery

被引:31
|
作者
Wang, Jue [1 ]
Zhong, Yanfei [1 ]
Zhang, Liangpei [1 ]
机构
[1] Wuhan Univ, State Key Lab Informat Engn Surveying Mapping & Re, Wuhan 430079, Peoples R China
基金
中国国家自然科学基金;
关键词
Task analysis; Remote sensing; Feature extraction; Decoding; Image resolution; Data mining; Buildings; Change detection (CD); deep learning; multitemporal image analysis; pretraining; supervised contrastive learning (CL); NETWORK;
D O I
10.1109/TGRS.2023.3236664
中图分类号
P3 [地球物理学]; P59 [地球化学];
学科分类号
0708 ; 070902 ;
摘要
Change detection (CD) is a challenging task on high-resolution bitemporal remote sensing images. Many recent studies of CD have focused on designing fully convolutional Siamese network architectures. However, most of these methods initialize their encoders by random values or an ImageNet pretrained model, without any prior for the CD task, thus limiting the performance of the CD model. In this article, the novel supervised contrastive pretraining and fine-tuning CD (SCPFCD) framework, which is made up of two cascaded stages, is presented to train a CD network based on a pretrained encoder. In the first supervised contrastive pretraining stage, the encoder of the Siamese network is asked to solve a joint pretext task introduced by the proposed CDContrast pretraining method on labeled CD data. The proposed CDContrast pretraining method includes land contrastive learning (LCL), which is based on supervised contrastive learning, and proxy CD learning. The LCL focuses on learning the spatial relationships among the land cover from bitemporal images by solving a land contrast task, while the proxy CD learning performs a proxy CD task on the top of the upsampling projector to avoid local optima for the LCL and learn features for the CD. Then, in the second fine-tuning stage, the whole Siamese network initialized with the pretrained encoder is fine-tuned to perform the CD task in an end-to-end manner. The proposed SCPFCD framework was verified with three CD datasets of high-resolution remote sensing images. The extensive experimental results consistently show that the proposed framework can effectively improve the ability to extract change information for Siamese networks.
引用
收藏
页数:16
相关论文
共 50 条
  • [21] High-Resolution Remote Sensing Bitemporal Image Change Detection Based on Feature Interaction and Multitask Learning
    Zhao, Chunhui
    Tang, Yingjie
    Feng, Shou
    Fan, Yuanze
    Li, Wei
    Tao, Ran
    Zhang, Lifu
    IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2023, 61
  • [22] A Survey on Deep Learning-Based Change Detection from High-Resolution Remote Sensing Images
    Jiang, Huiwei
    Peng, Min
    Zhong, Yuanjun
    Xie, Haofeng
    Hao, Zemin
    Lin, Jingming
    Ma, Xiaoli
    Hu, Xiangyun
    REMOTE SENSING, 2022, 14 (07)
  • [23] Self-supervised multimodal change detection based on difference contrast learning for remote sensing imagery
    Hou, Xuan
    Bai, Yunpeng
    Xie, Yefan
    Zhang, Yunfeng
    Fu, Lei
    Li, Ying
    Shang, Changjing
    Shen, Qiang
    PATTERN RECOGNITION, 2025, 159
  • [24] Knowledge Distillation-Based Lightweight Change Detection in High-Resolution Remote Sensing Imagery for On-Board Processing
    Wang, Guoqing
    Zhang, Ning
    Wang, Jue
    Liu, Wenchao
    Xie, Yizhuang
    Chen, He
    IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING, 2024, 17 : 3860 - 3877
  • [25] Fault-Tolerant Building Change Detection From Urban High-Resolution Remote Sensing Imagery
    Tang, Yuqi
    Huang, Xin
    Zhang, Liangpei
    IEEE GEOSCIENCE AND REMOTE SENSING LETTERS, 2013, 10 (05) : 1060 - 1064
  • [26] An automatic shadow detection method for high-resolution remote sensing imagery based on polynomial fitting
    Xue, Li
    Yang, Shuwen
    Li, Yikun
    Ma, Jijing
    INTERNATIONAL JOURNAL OF REMOTE SENSING, 2019, 40 (08) : 2986 - 3007
  • [27] CHANGE DETECTION FOR HIGH-RESOLUTION REMOTE SENSING IMAGERY USING OBJECT-ORIENTED CHANGE VECTOR ANALYSIS METHOD
    Li, Liang
    Li, Xue
    Zhang, Yun
    Wang, Lei
    Ying, Guowei
    2016 IEEE INTERNATIONAL GEOSCIENCE AND REMOTE SENSING SYMPOSIUM (IGARSS), 2016, : 2873 - 2876
  • [28] Land Cover Change Detection Based on Vector Polygons and Deep Learning With High-Resolution Remote Sensing Images
    Zhang, Hui
    Liu, Wei
    Niu, Hao
    Yin, Pengcheng
    Dong, Shiling
    Wu, Jialin
    Li, Erzhu
    Zhang, Lianpeng
    Zhu, Changming
    IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2024, 62 : 1 - 18
  • [29] Fine Object Change Detection Based on Vector Boundary and Deep Learning With High-Resolution Remote Sensing Images
    Shi, Jiacheng
    Liu, Wei
    Zhu, Yihu
    Wang, Shengli
    Hao, Sibao
    Zhu, Changming
    Shan, Haoyu
    Li, Erzhu
    Li, Xing
    Zhang, Lianpeng
    IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING, 2022, 15 : 4094 - 4103
  • [30] SEMI-SUPERVISED SPARSE RELEARNING REPRESENTATION CLASSIFICATION FOR HIGH-RESOLUTION REMOTE SENSING IMAGERY
    Li, Jiayi
    Huang, Xin
    Zhang, Liangpei
    2016 IEEE INTERNATIONAL GEOSCIENCE AND REMOTE SENSING SYMPOSIUM (IGARSS), 2016, : 2618 - 2621