Challenges and Prospects in Using Very High-Resolution Optical Satellite Imagery for Urban Water Body Extraction—A Study Utilizing GeoEye Remote Sensing Data for Surface Water Management Perspective

The availability of high-resolution images has introduced a valuable method for utilizing remote sensing data to derive detailed insights from these images. However, the application of high-resolution images in urban settings presents certain challenges that have garnered significant attention recently. In urban areas, the presence of shadows cast by elevated structures like buildings, towers, and trees creates a form of interference when attempting to extract information from optical images, especially related to water features. Within urban environments, distinguishing between water and shadowed areas in remote sensing images poses a considerable challenge. Furthermore, the presence of vegetation canopies over bodies of water further complicates the accurate delineation of water boundaries. The accurate extraction of urban water bodies from remote sensing imagery is of paramount importance for effective water resource management and urban planning. Various techniques and methodologies have been proposed to tackle the extraction of surface water areas from remote sensing images. This study primarily addresses the issues encountered when employing commonly used methods such as the normalized difference water indexNormalized difference water index (NDWI), object-based image classificationObject-based image classification (OBIC), and support vector machineSupport vector machine (SVM) supervised classification with GeoEye remote sensingGeoeye remote sensing data to extract urban water areas. As a result, a combination of digitization, decision support, and field observation has been adopted to achieve a highly accurate identification and extraction of water bodies from the GeoEye’s very high-resolution data. The investigation reveals that the NDWI, OBICObject-based image classification, and SVM supervised classification techniques are inadequate for achieving sufficient accuracy in identifying and extracting urban water bodies from GeoEye’s very high-resolution satellite imagery due to the intricate characteristics of urban landscapes. Conversely, the manual digitization approach coupled with decision support and on-site observations proves effective in extracting water bodies, as it helps mitigate the complexities inherent in such environments. Nonetheless, there is a pressing need to develop advanced automated techniques tailored to efficiently extract water bodies from extremely high-resolution images in complex urban areas. Additionally, the accuracy of the extracted water bodies has been validated through extensive field surveys and observations, ensuring that the information generated can be seamlessly integrated into urban water management planning efforts.