Search

AbstractBuilding an oil spill segmentation model is very challenging because of the limited available information on oil spill accidents. Therefore, this paper proposes a custom data generator based on Segmentation Network (Seg-Net) model implemented in Conditional Generative Adversarial Network (CGAN). The proposed model is trained for oil spill segmentation using 50 Sentinal-1 Synthetic Aperture Radar (SAR) images. The proposed model employs a modified Seg-Net as a generator to produce high-quality oil spills’ images and a Patch-GAN as discriminator. This architecture results in a significant improvement of the final oil segmentation results, in comparison with Seg-Net model, while using relatively small training dataset. For performance assessment, the paper presents the oil spills segmentation results of four suggested models using Sentinel-1 SAR images. The presented models are U-Net, Seg-Net, CGAN, and a Seg-Net-based CGAN the performance assessment reveals that the proposed model produces oil spill segmentation images with an average accuracy of 99.04%, Intersection over Union (IoU) index of 96.59%, and a precision of 85.24%. In addition, the training time required for the proposed model is 3 h 20 min per 50 epochs, while it is nearly 10 h and 55 min for training a CGAN model.

A dependable transportation infrastructure is an essential local resource, linking communities and facilitating economic expansion. The Nile Delta, renowned for its high population density, is experiencing subsidence as a result of natural forces exerting pressure on its sediments. Land subsidence exerts a substantial influence on infrastructure, economic, and social progress. This study examines the evaluation of land subsidence in roads experiencing various issues in the Nile Delta. Each year, subsidence results in significant damage and, in certain instances, complete disruption of numerous road networks, with potential durations ranging from a few days to several months. Conventional methods for identifying and monitoring subsidence on a large and intricate road network can be challenging due to the substantial time and resource requirements. Satellite-based monitoring is essential as it can effectively detect ground displacements caused by human activities (such as subterranean diggings and disruption of other infrastructures) or environmental dangers, specifically those associated with crucial hydrogeological incidents. Moreover, it possesses the benefit of efficiently encompassing extensive regions with precision. This study utilized the persistent scatterer interferometry (PSI) of SNAP-StaMPS workflow to employ differential interferometry synthetic aperture radar (DInSAR) processing methods for investigating the area. By utilizing Sentinel-1 data, we successfully acquired a comprehensive understanding of the deformation occurring in the study area, specifically in the roads where subsidence is persistently taking place. The data has demonstrated its value in enhancing our understanding of the safety status of infrastructure and facilitating maintenance actions. The accuracy of the radar interferometry results was confirmed through the use of in situ methods, including Global Navigation Satellite System (GNSS), Global Positioning System (GPS), smart phone, and unmanned aerial vehicle. This approach has demonstrated its capability to consistently detect vertical displacement in critical areas of roads. Furthermore, there is a strong correlation between the PSI interferometry and the in situ measurements when absolute values are taken into consideration. The findings of this study also demonstrate that open-source solutions are more effective, which encompass freely available data and software, offered through the COPERNICUS program, possess significant potential for monitoring vertical displacements in roads.

The digital elevation model (DEM) is crucial in many global and regional scientific studies in civilian and military applications. The aim of this research is to develop and test a new DEM approach for correcting the various errors in the Shuttle Radar Topography Mission (SRTM) digital elevation model. Firstly, the DEMs with the feature attributes from Sentinel-2 multispectral imagery are generated. Secondly, SRTM DEM with one band and attributes of a sentinel-2 image with eight bands are used as input data in supervised max-like hood, an artificial neural network (ANN), and support vector machine (SVM) classification models. Thirdly, ANN, supervised max-like hood, and SVM classification models, which have various properties, are fused by fuzzy majority voting (probability fusion). Finally, the fused probability is assigned for each pixel of the image, which has 12 fixed ground control points (GCPs), which is considered new input data for the inverse probability weighted interpolation (IPWI) approach to create the corrected SRTM elevations. The results were contrasted with a reference DEM (RD) created by image matching with Worldview-1 stereo satellite images, which had a 1-m vertical accuracy. The results of this study demonstrated that the RMSE of the original SRTM DEM was 5.95. On the other hand, the RMSE of the estimated elevations by the IPWI approach has been improved to 1.98 compared with that of the MLR method (3.01). The study shows a series of significant improvements in the SRTM when assessed with the reference DEM, with an RMSE reduction of (66.72%) when compared to the widely utilized multiple linear regression (MLR) method. It can be concluded that the elevation error of the original SRTM DEM is clearly reduced by the suggested approach.

Road pavement deformation monitoring is considered the main task for maintenance purposes, especially potholes and cracks, which are the most common types of road deformation surfaces. In order to make pavement inspections more effective, new types of remote sensing data that do not damage the pavement are being used more and more to find pavement distress. This article presents a proposed approach for extracting surface cracks from unmanned aerial vehicle (UAV) images using machine learning, focusing on the data pre-treatment processes. The objective of this study is to evaluate the effectiveness of decision tree classification (DT) in detecting cracks. The performance of the models is also evaluated. The performance evaluation approach is predicated on two primary criteria: model validation and testing. Also, the extent of the impact of post-classification operations, edge detection technology, and morphological processes on crack identification as well as classification accuracy, the digital orthomosaic was generated by the use of a technique commonly referred to as backward projection. To achieve this, the study uses a fusion of gray-level co-occurrence matrix (GLCM) attribute data and RGB images. Cracks are discovered using a classification tree (CT)-based classification approach with an overall classification rate of 86%. Ultimately, morphological processes using the closed image that was formed had a commendable level of accuracy, with an overall classification rate of 96%. The Canny edge detection algorithm has demonstrated its efficacy as a preferred method for detecting cracks from UAV images, providing invaluable decision support for actual road maintenance.