Fine Classification of Industrial and Mining Areas in XuZhou City Using Sentinel-2 Imagery

Restoring the ecological function and land value of mining areas is crucial for achieving sustainable development. The spatial distribution of fine classification in mining areas serves as a scientific basis for related policy planning and implementation, providing theoretical support for ecological restoration in these regions. This study focuses on the Chengbei mining area of Xuzhou City as the research area, utilizing Sentinel-2 imagery to establish a classification system for fine classification specific to mining areas. By coupling remote sensing image texture features, remote sensing indices, and topographic parameters, a Random Forest model was constructed to extract fine classification from the mining area and analyze the evolution of their spatial distribution patterns. The experimental results demonstrate that this model significantly enhances the efficiency and accuracy of extracting fine classification in mining areas. Following the restoration of the mining area, there is a noticeable increase in types such as arbor forests and other woodland, resulting in a more stable ecosystem conducive to the sustainable development of fine classification in the mining area.