28 October 2024 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Semantic change detection (SCD) represents a challenging task in the interpretation of remote sensing images (RSIs), with the goal of identifying change regions and extracting semantic information from bitemporal RSIs simultaneously. The recent integration of deep neural networks leveraging multitask learning has shown promise in enhancing SCD performance. However, there is still a challenge in improving SCD performance, specifically in designing a fine-grained network structure that can handle the two subtasks of change region localization and semantic information recognition in parallel. In this context, a novel multitask Siamese network, termed EGMS-Net, is proposed to boost the performance of SCD, which consists of three core components. First, a coarse-to-fine multitask Siamese network is constructed to obtain semantic information and change information at multiple levels. Second, an adaptive change information enhancement method based on spatial-spectral collaborative attention mechanism is proposed, which can assist the accurate localization of change regions without significantly increasing the model parameters. Third, a change information guidance module is developed to strengthen the interaction between multitask branches and reduce the difficulty of network training. Experiments on three benchmark datasets demonstrate that the proposed EGMS-Net outperforms existing state-of-the-art methods in the SCD community.
4 July 2025 in IEEE Transactions on Geoscience and Remote Sensing
Implementing change detection (CD) with bi-temporal remote sensing images (RSIs) is essential for gaining insights into the dynamic evolution of the Earth’s surface. Recent advancements in deep learning methodologies have demonstrated considerable success in CD applications. However, the efficacy of supervised CD networks is significantly constrained by their dependence on high-quality change labels and precisely registered bi-temporal RSIs, which imposes significant limitations in real-world scenarios. To address this issue, we propose a novel single-temporal unsupervised CD method, STU-SAMI, which integrates the Segment Anything Model (SAM) with instance-level change generation. This method aims to facilitate the training of high-performance CD networks using unpaired and unlabeled single-temporal RSIs. The proposed approach comprises three main steps. First, the SAM is introduced to extract morphologically complete object instances from single-temporal RSIs. Second, an online instance-level change generation method is proposed, which transforms single-temporal RSIs into bi-temporal pseudo change samples based on the object instances obtained by SAM. Third, a streamlined and efficient deep Siamese CD network is constructed to support the training and inference processes. Extensive experiments conducted on three benchmark datasets demonstrate that the proposed method outperforms the existing state-of-the-art unsupervised CD methods. The code is available at https://github.com/IceStreams/STU-SAMI.
18 March 2025 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Currently, satellite remote sensing image acquisition systems typically include two forms of panchromatic and multispectral images, both of which have complementary advantages in spatial and channel dimensions. However, translating advantageous information into a deciphering function in road extraction tasks remains a challenge. This study, therefore, proposes a road-extraction method combining spectral information and spatial details. First, a multibranch network framework was built based on an encoding-decoding structure. The encoding layers of the panchromatic and multispectral image branches were constructed from the residual modules. Fusion branches were then constructed during the decoding phase. The spectral information of the multispectral image and spatial details of the panchromatic image were then obtained using the HIS color transform and Haar wavelet transform, respectively, and injected into the fusion branch. A polarized self-attention mechanism was finally introduced at different levels of the fusion branch to reduce information loss during feature extraction, and operations such as connected convolution and nonlinear activation were later connected to complete the road prediction. The implementation of the proposed method on the GF2-FC and CHN6-CUG datasets revealed a superior performance compared to comparative methods in terms of quantitative evaluation metrics. The proposed method performed the strongest in several scenarios, particularly in difficult road extraction areas such as shadows and vegetation cover.