Search

To address the problems in remote sensing image change detection such as missed detection of features at different scales and incomplete region detection, this paper proposes a high-resolution remote sensing image change detection model (Multi-scale Attention Siamese Network, MASNet) based on a Siamese network and multi-scale attention mechanism. The MASNet model took the Siamese structure of the ResNet-50 network to extract features of different simultaneous images and then applied the attention module to feature maps of different scales to generate multi-scale feature representations. Meanwhile, an improved contrastive loss function was adopted to enhance the learning of change features and improving the imbalance problem between unchanged and changed samples. Furthermore, to address the current time-consuming and laborious phenomenon of manually annotating datasets, we provided a change detection dataset from Yunnan Province in China (YNCD) that contains 1540 pairs of 256 × 256 bi-temporal images with a spatial resolution of 1 m. Then, model training and change detection applications were studied by expanding a small number of experimental area samples into the existing public datasets. The results showed that the overall accuracy of the MASNet model for change detection in the experimental area is 95.34%, precision rate is 79.78%, recall rate is 81.52%, and F1 score is 80.64%, which are better than those of six comparative models (FC-EF, FC-Siam-Diff, FC-Siam-Conc, PAN, MANet, and STANet). This verifies the effectiveness of the MASNet model as well as the feasibility of change detection by expanding existing public datasets. [ABSTRACT FROM AUTHOR]

Change detection in high resolution (HR) remote sensing images faces more challenges than in low resolution images because of the variations of land features, which prompts this research on faster and more accurate change detection methods. We propose a pixel-level semantic change detection method to solve the fine-grained semantic change detection for HR remote sensing image pairs, which takes one lightweight semantic segmentation network (LightNet), using the parameter-sharing SiameseNet, as the architecture to carry out pixel-level semantic segmentations for the dual-temporal image pairs and achieve pixel-level change detection based directly on semantic comparison. LightNet consists of four long–short branches, each including lightweight dilated residual blocks and an information enhancement module. The feature information is transmitted, fused, and enhanced among the four branches, where two large-scale feature maps are fused and then enhanced via the channel information enhancement module. The two small-scale feature maps are fused and then enhanced via a spatial information enhancement module, and the four upsampling feature maps are finally concatenated to form the input of the Softmax. We used high resolution remote sensing images of Lake Erhai in Yunnan Province in China, collected by GF-2, to make one dataset with a fine-grained semantic label and a dual-temporal image-pair label to train our model, and the experiments demonstrate the superiority of our method and the accuracy of LightNet; the pixel-level semantic change detection methods are up to 89% and 86%, respectively. [ABSTRACT FROM AUTHOR]

Mining for rare earth elements is rapidly increasing, driven by current and projected demands for information and energy technologies. Following China's Central Government's 2012 strategy to shift away from mining in favor of value-added processing, primary extraction has increased outside of China. Accordingly, changes in mineral exploitation in China and Myanmar have garnered considerable attention in the past decade. The prevailing assumption is that mining in China has decreased while mining in Myanmar has increased, but the dynamic in border regions is more complex. Our empirical study used Google Earth Engine (GEE) to characterize changes in mining surface footprints between 2005 and 2020 in two rare earth mines located on either side of the Myanmar–China border, within Kachin State in northern Myanmar and Nujiang Prefecture in Yunnan Province in China. Our results show that the extent of the mining activities increased by 130% on China's side and 327% on Myanmar's side during the study period. We extracted surface reflectance images from 2005 and 2010 from Landsat 5 TM and 2015 and 2020 images from Landsat 8 OLI. The Normalized Vegetation Index (NDVI) was applied to dense time-series imagery to enhance landcover categories. Random Forest was used to categorize landcover into mine and non-mine classes with an overall accuracy of 98% and a Kappa Coefficient of 0.98, revealing an increase in mining extent of 2.56 km2, covering the spatial mining footprint from 1.22 km2 to 3.78 km2 in 2005 and 2020, respectively, within the study area. We found a continuous decrease in non-mine cover, including vegetation. Both mines are located in areas important to ethnic minority groups, agrarian livelihoods, biodiversity conservation, and regional watersheds. The finding that mining surface areas increased on both sides of the border is significant because it shows that national-level generalizations do not align with local realities, particularly in socially and environmentally sensitive border regions. The quantification of such changes over time can help researchers and policymakers to better understand the shifting geographies and geopolitics of rare earth mining, the environmental dynamics in mining areas, and the particularities of mineral extraction in border regions. [ABSTRACT FROM AUTHOR]