In recent years, remote sensing researchers have investigated the use of different modalities (or combinations of modalities) for classification tasks. Such modalities can be extracted via a diverse range of sensors and images. Currently, there are no (or only a few) studies that have been done to increase the land cover classification accuracy via unmanned aerial vehicle (UAV)-digital surface model (DSM) fused datasets. Therefore, this study looks at improving the accuracy of these datasets by exploiting convolutional neural networks (CNNs). In this work, we focus on the fusion of DSM and UAV images for land use/land cover mapping via classification into seven classes: bare land, buildings, dense vegetation/trees, grassland, paved roads, shadows, and water bodies. Specifically, we investigated the effectiveness of the two datasets with the aim of inspecting whether the fused DSM yields remarkable outcomes for land cover classification. The datasets were: (i) only orthomosaic image data (Red, Green and Blue channel data), and (ii) a fusion of the orthomosaic image and DSM data, where the final classification was performed using a CNN. CNN, as a classification method, is promising due to hierarchical learning structure, regulating and weight sharing with respect to training data, generalization, optimization and parameters reduction, automatic feature extraction and robust discrimination ability with high performance. The experimental results show that a CNN trained on the fused dataset obtains better results with Kappa index of~0.98, an average accuracy of 0.97 and final overall accuracy of 0.98. Comparing accuracies between the CNN with DSM result and the CNN without DSM result for the overall accuracy, average accuracy and Kappa index revealed an improvement of 1.2%, 1.8% and 1.5%, respectively. Accordingly, adding the heights of features such as buildings and trees improved the differentiation between vegetation specifically where plants were dense.
Predicting landslide occurrences can be difficult. However, failure to do so can be catastrophic, causing unwanted tragedies such as property damage, community displacement, and human casualties. Research into landslide susceptibility mapping (LSM) attempts to alleviate such catastrophes through the identification of landslide prone areas. Computational modelling techniques have been successful in related disaster scenarios, which motivate this work to explore such modelling for LSM. In this research, the potential of supervised machine learning and ensemble learning is investigated. Firstly, the Flexible Discriminant Analysis (FDA) supervised learning algorithm is trained for LSM and compared against other algorithms that have been widely used for the same purpose, namely Generalized Logistic Models (GLM), Boosted Regression Trees (BRT or GBM), and Random Forest (RF). Next, an ensemble model consisting of all four algorithms is implemented to examine possible performance improvements. The dataset used to train and test all the algorithms consists of a landslide inventory map of 227 landslide locations. From these sources, 13 conditioning factors are extracted to be used in the models. Experimental evaluations are made based on True Skill Statistic (TSS), the Receiver Operation characteristic (ROC) curve and kappa index. The results show that the best TSS (0.6986), ROC (0.904) and kappa (0.6915) were obtained by the ensemble model. FDA on its own seems effective at modelling landslide susceptibility from multiple data sources, with performance comparable to GLM. However, it slightly underperforms when compared to GBM (BRT) and RF. RF seems most capable compared to GBM, GLM, and FDA, when dealing with all conditioning factors.
Landslides are type of natural geohazard interfering with many economical and social activities and causing serious damages on human life. It is ranked as a great disaster, threatening life, property and environment. Therefore, early prediction of landslide prone areas is vital. Variety of causative factors such as glaciers melting, excessive raining, mining, volcanic activities, active faults, earthquake, logging, erosion, urbanization, construction, and other human activities can trigger landslide occurrence. Then, identification of factors that directly influences the slide events is highly in demand. Some topographical, geological, and hydrological datasets (e.g., slope, aspect, geology, terrain roughness, vegetation index, distance to stream, distance to road, distance to fault, land use, precipitation, profile curvature, plan curvature) are considered to be effective conditioning factors. However, the importance of each factor differs from one study to another. This study investigates the effectiveness of four sets of landslide conditioning variable(s). Fourteen landslide conditioning variables were considered in this study where they were duly divided into four groups G1, G2, G3, and G4. Three machine learning algorithms namely, Random Forest (RF), Naive Bayes (NB), and Boosted Logistic Regression (LogitBoost) were constructed based on each dataset in order to determine which set would be more suitable for landslide susceptibility prediction. In total, 227 landslide inventory datasets of the study area were used where 70% was used for training and 30% for testing. To this end, in the present research, the two main objectives were: 1) Investigation on effectiveness of 14 landslides conditioning factors (altitude, slope, aspect, total curvature, profile curvature, plan curvature, Stream Power Index (SPI), Topographic Wetness Index (TWI), Terrain Roughness Index (TRI), distance to fault, distance to road, distance to stream, land use, and geology) by analyzing and determining the most important factors using variance-inflated factor (VIF), Pearson's correlation and Chi-square techniques. Consequently, 4 categories of datasets were defined; first dataset included all 14 conditioning factors, second dataset included Digital Elevation Models (DEM) derivatives (morphometrice factors), third dataset was only based on 5 factors namely lithology, land use, distance to stream, distance to road, and distance to fault, and last dataset was included 8 factors selected using factor analysis and optimization. 2) Evaluate the sensitivity of each modeling technique (NB, RF and LogitBoost) to different conditioning factors using the area under curve (AUC). Eventually, RF technique using optimized variables (G4) performed well with AUC of 0.940 followed by LogitBoost (0.898) and NB (0.864).
Assessment of the most appropriate groundwater conditioning factors (GCFs) is essential when performing analyses for groundwater potential mapping. For this reason, in this work, we look at three statistical factor analysis methods-Variance Inflation Factor (VIF), Chi-Square Factor Optimization, and Gini Importance-to measure the significance of GCFs. From a total of 15 frequently used GCFs, 11 most effective ones (i.e., altitude, slope angle, plan curvature, profile curvature, topographic wetness index, distance from river, distance from fault, river density, fault density, land use, and lithology) were finally selected. In addition, 917 spring locations were identified and used to train and test three machine learning algorithms, namely Mixture Discriminant Analysis (MDA), Linear Discriminant Analysis (LDA) and Random Forest (RF). The resultant trained models were then applied for groundwater potential prediction and mapping in the Haraz basin of Mazandaran province, Iran. MDA has been successfully applied for soil erosion and landslide mapping, but has not yet been fully explored for groundwater potential mapping (GPM). Although other discriminant methods, such as LDA, exist, MDA is worth exploring due to its capability to model multivariate nonlinear relationships between variables; it also undertakes a mixture of unobserved subclasses with regularization of non-linear decision boundaries, which could potentially provide more accurate classification. For the validation, areas under Receiver Operating Characteristics (ROC) curves (AUC) were calculated for the three algorithms. RF performed better with AUC value of 84.4%, while MDA and LDA yielded 75.2% and 74.9%, respectively. Although MDA performance is lower than RF, the result is satisfactory, because it is within the acceptable standard of environmental modeling. The outcome of factor analysis and groundwater maps emphasizes on optimization of multicolinearity factors for faster spatial modeling and provides valuable information for government agencies and private sectors to effectively manage groundwater in the region.
This paper presents an application of data-driven Dempster-Shafer theory (DST) of evidence to fuse multisensor data for land-cover feature extraction. Over the years, researchers have focused on DST for a variety of applications. However, less attention has been given to generate and interpret probability, certainty, and conflict maps. Moreover, quantitative assessment of DST performance is often overlooked. In this paper, for implementation of DST, two main types of data were used: multisensor data such as Light Detection and Ranging (LiDAR) and multispectral satellite imagery [Satellite Pour l'Observation de la Terre 5 (SPOT 5)]. The objectives are to classify land-cover types from fused multisensor data using DST, to quantitatively assess the accuracy of the classification, and to examine the potential of slope data derived from LiDAR for feature detection. First, we derived the normalized difference vegetation index (NDVI) from SPOT 5 image and the normalized digital surface model (DSM) (nDSM) from LiDAR by subtracting the digital terrain model from the DSM. The two products were fused using the DST algorithm, and the accuracy of the classification was assessed. Second, we generated a surface slope from LiDAR and fused it with NDVI. Subsequently, the classification accuracy was assessed using an IKONOS image of the study area as ground truth data. From the two processing stages, the NDVI/nDSM fusion had an overall accuracy of 88.7%, while the NDVI/slope fusion had 75.3%. The result indicates that NDVI/nDSM integration performed better than NDVI/slope. Although the overall accuracy of the former is better than the latter (NDVI/slope), the contribution of individual class reveals that building extraction from fused slope and NDVI performed poorly. This study proves that DST is a time-and cost-effective method for accurate land-cover feature identification and extraction without the need for a prior knowledge of the scene. Furthermore, the ability to generate ot-er products like certainty, conflict, and maximum probability maps for better visual understanding of the decision process makes it more reliable for applications such as urban planning, forest management, 3-D feature extraction, and map updating.
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.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.