点击次数：
影响因子：5.343
DOI码：10.1109/LGRS.2021.3129776
所属单位：Beihang Univ, Image Proc Ctr, Sch Astronaut
发表刊物：IEEE Geoscience and Remote Sensing Letters
关键字：Image segmentationSemanticsTrainingAdaptation modelsRemote sensingGeometryDecodingGenerative-adversarial learninggeometry consistency (GC)remote-sensing images (RSIs)semantic segmentationunsupervised
摘要：In remote-sensing image (RSI) semantic segmentation, the dependence on large-scale and pixel-level annotated data has been a critical factor restricting its development. In this letter, we propose an unsupervised semantic segmentation network embedded with geometry consistency (UGCNet) for RSIs, which imports the adversarial-generative learning strategy into a semantic segmentation network. The proposed UGCNet can be trained on a source-domain dataset and achieve accurate segmentation results on a different target-domain dataset. Furthermore, for refining the remote-sensing target geometric representation such as densely distributed buildings, we propose a geometry-consistency (GC) constraint that can be embedded in both image-domain adaptation process and semantic segmentation network. Therefore, our model could achieve cross-domain semantic segmentation with target geometric property preservation. The experimental results on Massachusetts and Inria buildings datasets prove that the proposed unsupervised UGCNet could achieve a very comparable segmentation accuracy with the fully supervised model, which validates the effectiveness of the proposed method.
论文类型：期刊论文
一级学科：控制科学与工程
卷号：19
ISSN号：10.1109/LGRS.2021.3129776
是否译文：否
发表时间：2022-01-21
收录刊物：SCI
发布期刊链接：https://ieeexplore.ieee.org/document/9623453/