Monitoring water bodies from remote sensing data is certainly an essential task to supervise the actual conditions of the available water resources for environment conservation, sustainable development and many other applications. Being Sentinel-2 images some of the most attractive data, existing traditional index-based and deep learning-based water extraction methods still have important limitations to effectively deal with large heterogeneous areas since many types of water bodies with different spatial-spectral complexities are logically expected. Note that, in this scenario, optimal feature abstraction and neighbourhood information may certainly vary from water to water pixel, however existing methods are generally constrained by a fix abstraction level and amount of land cover context. To address these issues, this paper presents a new attentional dense convolutional neural network (AD-CNN) specially designed for water body extraction from Sentinel-2 imagery. On the one hand, AD-CNN exploits dense connections to allow uncovering deeper features while simultaneously characterizing multiple data complexities. On the other hand, the proposed model also implements a new residual attention module to dynamically put the focus on the most relevant spatial-spectral features for classifying water pixels. To test the performance of AD-CNN, a new water database of Nepal (WaterPAL) is also built. The conducted experiments reveal the competitive performance of the proposed architecture with respect to several traditional indexbased and state-of-the-art deep learning-based water extraction models. The codes and data related to this paper will be accessible on https://github.com/rufernan/ADCNN.
Assessment of Landslide Susceptibility Map (LSM) is crucial to the reduction of risk of the landslides. This paper focusses on modelling LSM using two different machine learning algorithms namely Random Forest (RF), and Classification and Regression Tree (CART). Ten landslide causative factors along with an inventory of landslides containing 89 recent and historic landslide points, and 90 randomly generated nonlandslide points were used to prepare a susceptibility map. The study area; Baglung district is located in the Gandaki province of Nepal, a highly landslide susceptible zone. Frequency ratio (FR) of each class of conditioning factors were calculated. FR values of landslide and non-landslide points were extracted from normalized FR classified raster. The extracted FR values of each point (landslide and non-landslide) was randomly split into training (70%) and testing (30%) samples which were used for training and testing the model. The performance of each algorithm was evaluated using receiver operating characteristics (ROC) curves in combination with area under the curve (AUC) and error matrix. The AUC results introduced success rate of 1 and 0.88 for RF and CART respectively. Also, the rates of prediction were 0.86 and 0.96 for RF and CART respectively. Similarly, RF and CART showed accuracy of 0.88 and 0.83 from confusion matrix. Therefore, the RF algorithm was superior to CART in identifying the regions at risk for future landslides in the study area. The outcomes of this study is useful and essential for the government, planners, researchers, decision makers and general landuse planners.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.