Road extraction method and system based on improved DLinkNet

• Main Authors: LIN CONG, WANG CHUHONG, WU YUEYANG, WU ZHENLU
• Format: Patent
• Language: Chinese; English
• Published: 02.08.2024
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; PHYSICS

The invention relates to the technical field of image processing, in particular to a road extraction method and system based on an improved DLinkNet, and the method comprises the steps: building a road extraction model, the road extraction model is a UGD-DLinkNet model constructed based on a DLinkNet model, the UGD-DLinkNet model comprises an encoder, a bridging network and a decoder, and the encoder is constructed based on a ResNet34 network; obtaining a data set, and training a road extraction model through the data set to obtain a trained road extraction model; wherein the data set comprises a plurality of remote sensing images; obtaining a to-be-extracted remote sensing image, predicting the to-be-extracted remote sensing image by using the trained road extraction model, and generating a road extraction result; the precision of road extraction in the remote sensing image can be improved. 本发明涉及图像处理技术领域，具体涉及基于改进DLinkNet的道路提取方法及系统，方法包括：建立道路提取模型，所述道路提取模型为基于DLinkNet模型构建的UGD-DLinkNet模型，所述UGD-DLinkNet模型包括编码器、桥接网