Improvement of Feature Extraction Network Structure for Remote Sensing Image Change Detection
Remote sensing image change detection technology,as an important applied technology,has a wide range of applications in the field of surface change monitoring.Aiming at the problem that most current feature extraction networks cannot balance spatial and semantic information,a feature extraction structure called CDResNet is proposed for remote sensing image change detection based on residual networks.Using this structure for feature extraction of remote sensing images at different times can balance the spatial information and semantic information in the extracted feature maps,and more accurately identify and locate the changing regions in the image.The robustness and accuracy of the proposed method are validated through experiments on two public datasets.
国家科技期刊平台
NSTL
万方数据
