Remote sensing image field building segmentation method based on receptive field lifting network

• Main Authors: YU WEIXUE, LI GUANQUN
• Format: Patent
• Language: Chinese; English
• Published: 24.01.2023
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; PHYSICS

The invention relates to the technical field of image processing, and discloses a remote sensing image field building segmentation method based on a receptive field lifting network, and the method comprises the steps: obtaining a remote sensing field building image, obtaining a feature image according to the receptive field lifting network, inputting the feature image to a field building segmentation network, and obtaining a segmentation result. According to the invention, a down-sampling-up-sampling overall architecture is not used, but a tiled network overall architecture is adopted, so that the problems of information loss and feature loss are avoided; a receptive field promoting module fusing serial and parallel double links is constructed and serves as a main body to form the network, the receptive field of the network is greatly increased, the network can refer to more buildings and consider more buildings when facing the conditions of a large ground coverage area, a small number of buildings and sparse