Embedded Files
Long-term landslides detection and susceptibility mapping with multi-source optical remote sensing in the eastern Himalayan syntaxis
Long-term landslides detection and susceptibility mapping with multi-source optical remote sensing in the eastern Himalayan syntaxis
In this study, we applied machine learning method: Random Forest, to predict distribution of landslides based on calculated feature maps. Other than that, we build a landslide dataset based on long-term time series remote sensing imagery to figure out the changing pattern of the spatial and temporal distribution of landslides.
Digital Twins of Youyi Community in Beijing
Digital Twins of Youyi Community in Beijing
By collecting real measuring data of buildings in community, we conducted digital 3D models in order to help the reconstruction planning. Basically, we first construct 3D models in professional softwares like Theft, 3Ds Max. After that, 3D models are imported to ArcGIS Pro for data format conversion and then display.
Semantic Segmentation of High Resolution Remote Sensing Image
Semantic Segmentation of High Resolution Remote Sensing Image
We proposed a novel method which integrating multi-head attention module into GLnet, a network that extracting both general and local features from processed remote sensing images in order to segment different landscapes. The multi-head attention module here helps to extract the relations between features of different levels in order to have a both comprehensive and detailed understanding of the whole image.
Page updated
Google Sites
Report abuse