基于CAR-Siamese网络的高分辨率遥感图像建筑物变化检测  

Building change detection from remote sensing images using CAR-Siamese net

作　　者：姚沐风 昝露洋 李柏鹏 李庆亭 陈正超 YAO Mufeng;ZAN Luyang;LI Baipeng;LI Qingting;CHEN Zhengchao(Airborne Remote Sensing Center,Aerospace Information Research Institute,Chinese Academy of Sciences,Beijing 100094,China;College of Resources and Environment,University of Chinese Academy of Sciences,Beijing 100049,China)

机构地区：[1]中国科学院空天信息创新研究院航空遥感中心,北京100094 [2]中国科学院大学资源与环境学院,北京100049

出　　处：《中国科学院大学学报（中英文）》2023年第3期380-387,共8页Journal of University of Chinese Academy of Sciences

基　　金：国家重点研发计划项目(2016YFB0500304);北京市自然科学基金(9182004)资助。

摘　　要：准确提取建筑物变化区域对城乡规划、地理国情监测、城市扩张分析有着重要意义。传统遥感变化检测方法难以适应遥感图像复杂场景下的变化检测任务的要求。近年来广泛应用于计算机视觉领域的深度学习变化检测算法相对于传统方法在效率和精度上有明显提升。然而遥感图像上建筑物特征丰富、变化多样,且建筑物变化样本获取难度大,导致现有深度学习模型在建筑物变化检测任务上精度受限。针对这一问题,提出变化注意力残差孪生网络(CAR-siamese net),增强不同尺度下图像信息的共享交流,充分学习建筑物的变化特征,同时,提出建筑物语义分割样本预训练策略,有效利用现有建筑物分割样本,最终提升了变化检测网络对建筑物变化的解译能力。以北京昌平区影像为底图制作建筑物变化检测数据集,在该数据集和Levir-CD公开数据集上的实验结果表明,该方法能有效提高建筑物变化检测精度。Accurately extracting building change regions is of great significance to urban and rural planning,geographic national conditions monitoring,and urban expansion analysis.Traditional remote sensing change detection methods are difficult to adapt to the change detection tasks in complex scenes of remote sensing images.In recent years,deep learning change detection algorithms,which have been widely used in the field of computer vision,have significantly improved efficiency and accuracy compared to traditional methods.However,the features of buildings on remote sensing images are rich and varied,and it is difficult to obtain samples of building changes,which leads to the limited accuracy of existing deep learning models in building change detection tasks.This paper proposes a change attention residual siamese network(CAR-siamese net),which enhances the interaction of image information at different scales,and fully learns the change features of buildings.In addition,a pre-training strategy is proposed in this paper to effectively use building segmentation samples,and the ability of the change detection network to interpret building changes is improved.In this paper,a building change detection data set is made based on images of Changping District,Beijing.Experimental results on this data set and Levir-CD public data set show that the method in this paper can effectively improve the accuracy of building change detection.

关 键 词：变化检测 建筑物 深度学习 卷积神经网络 孪生网络 变化注意力残差 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]