Key points are not available for this paper at this time.
Satellite images provide a wealth of information that is used in various applications such as urban planning, environmental monitoring and terrain analysis. However, converting raw satellite data into a grid image suitable for these applications remains a challenge. this study proposes a three-stage methodology that integrates advanced image processing techniques to transform satellite image surfaces into a mesh image to enhance the utility in e geometric topology. The first phase involves applying intelligent detectors technology to identifying the edges of objects within the satellite image. By detecting high-density change points which are characteristic points at edge intersections especially at corners or angular features formed by objects. In the second stage, these features are processed using SIFT which ensures that scale-invariant features are extracted across different images. The final stage utilizes Delaunay triangulation to create the mesh, effectively converting the satellite image surfaces into a mesh representation. This mesh representation is a type of graph consisting of nodes (representing extracted features) and edges (connecting these nodes). Such a representation opens up new possibilities for analysis and study and providing an organized and detailed depiction of the Earth's surface. The mesh image produced through this methodology can be applied to numerous scientific and practical applications, bridging the gap between raw satellite data and its practical utilization in various fields. The satellite images used in this study are high-resolution raster optical images, commonly employed in applications such as urban mapping, environmental monitoring, and geographic analysis. These types of images are typically captured by satellites like Landsat. Such raster optical imagery provides detailed visual information that is essential for analyzing surface features and patterns. The results demonstrate the effectiveness of the proposed system, achieving a Ki metric value of (104.46) compared to (78.51) in previous approaches, indicating superior mesh quality
Raed Abd Alreda Shekan (Sat,) studied this question.