Study on building extraction from high resolution remote sensing images in mountain area based on U-Net network model
In view of the complex topography and varied natural environment in mountainous areas,the layout of buildings will not be distributed in a network like that in plain areas,which leads to the prob-lems of debris and incorrect range in the extraction of buildings.This paper conducts an experimental investigation into the extraction of typical buildings in mountainous areas using the U-Net network model based on BJ-2 and GF-7 remote sensing images.The experimental data is compared and ana-lyzed with the Third National Land Resources Survey data,the results indicate that the buildings ex-tracted using the method proposed in this paper have high accuracy,and the U-Net network model is feasible for extracting buildings in mountainous areas.
high resolution remote sensing imagesU-Net neural network modelbuilding extraction
万方数据
维普
