Fusion of hyperspectral images and lidar-based dems for coastal mapping

Coastal mapping is essential for a variety of applications such as coastal resource management, coastal environmental protection, and coastal development and planning. Various mapping techniques, like ground and aerial surveying, have been utilized in mapping coastal areas. Recently, multispectral and hyperspectral satellite images and elevation data from active sensors have also been used in coastal mapping. Integrating these datasets can provide more reliable coastal information. This paper presents a novel technique for coastal mapping from an airborne visible/infrared imaging spectrometer (AVIRIS) hyperspectral image and a light detection and ranging (LIDAR)-based digital elevation model (DEM). The DEM was used to detect and create a vector layer for building polygons. Subsequently, building pixels were removed from the AVIRIS image and the image was classified with a supervised classifier to discriminate road and water pixels. Two vector layers for the road network and the shoreline segments were vectorized from road pixels and water-body border pixels using several image-processing algorithms. The geometric accuracy and completeness of the results were evaluated. The average positional accuracies for the building, road network, and shoreline layers were 2.3, 5.7, and 7.2 m, respectively. The detection rates of the three layers were 93.2%, 91.3%, and 95.2%, respectively. Results confirmed that utilizing laser ranging data to detect and remove buildings from optical images before the classification process enhances the outcomes of this process. Consequently, integrating laser and optical data provides high-quality and more reliable coastal geospatial information.