全尺度特征聚合的高分辨率遥感影像变化检测网络  

Full-scale feature aggregation network for high-resolution remote sensing image change detection

作　　者：姜明 张新长 孙颖[2] 冯炜明 阮永俭 JIANG Ming;ZHANG Xinchang;SUN Ying;FENG Weiming;RUAN Yongjian(School of Geography and Remote Sensing,Guangzhou University,Guangzhou 510006,China;School of Geography and Planning,Sun Yat-Sen University,Guangzhou 510275,China;Guangdong Provincial Key Laboratory of Urban Security Intelligent Monitoring and Smart City Planning,Guangzhou 510290,China)

机构地区：[1]广州大学地理科学与遥感学院,广东广州510006 [2]中山大学地理科学与规划学院,广东广州510275 [3]广东省城市安全智能监测与智慧城市规划企业重点实验室,广东广州510290

出　　处：《测绘学报》2023年第10期1738-1748,共11页Acta Geodaetica et Cartographica Sinica

基　　金：国家自然科学基金面上项目(42371406;42071441);广东省城市安全智能监测与智慧城市规划企业重点实验室资助项目(GPKLIUSMSCP-2023-KF-05)。

摘　　要：利用遥感影像检测地表变化对了解地表动态至关重要。近年来,基于深度学习的变化检测方法因其优异的特征提取和表达能力而成为研究的热点。在全卷积网络结构方法中,融合多尺度特征信息是提高变化检测性能的关键,以往方法大多采用跳跃连接或密集连接结构,一定程度上提高了变化检测方法的精度。然而,此类方法只对相同尺度上的特征进行融合,无法从多尺度上融合足够的信息而导致达不到令人满意结果。本文提出了一种全尺度特征聚合网络(FSANet),用于解决遥感影像变化检测问题。首先,使用孪生网络提取双时相影像的特征;然后,利用全尺度特征连接结构将提取的特征有效地连接起来,为了防止特征冗余,使用特征细化模块将特征细化;最后,为了优化模型训练,采用多尺度监督策略,在解码器中额外输出多个检测结果,共同计算最终的损失值。为了验证方法的可行性,本文使用LEVIR-CD数据集和SVCD数据集来评估模型。试验结果表明,本文方法优于其他主流的变化检测方法,同时在精度和复杂度之间有着较好的平衡。Using remote sensing imagery to detect changes is crucial for understanding land surface dynamics.In recent years,deep learning-based methods have become a focus area owing to their excellent feature extraction and representation ability.The fusion of multi-scale feature information is the key to improving change detection performance in fully convolutional network-based structural methods.Most of the previous methods use skip connection or dense connection structure,which improves the accuracy of change detection methods to a certain extent.However,such methods only fuse features at the same scale and lack sufficient information from multiple scales to achieve satisfactory results.In this paper,a full-scale feature aggregation network(FSANet)is proposed to solve the problem of remote sensing image change detection.Firstly,the features of the bi-temporal images are extracted using a siamese network,then the features are efficiently concatenated using a full-scale feature concatenation structure,and to prevent feature redundancy,the features are refined using a feature refinement module.Finally,to optimize model training,a multiscale supervision strategy is used.Multiple additional detections are output in the decoder,which calculate the final loss value together.To check the reliability of FSANet,we tested it on two public datasets,the LEVIR-CD and SVCD datasets.The experimental results show that the method outperforms other mainstream change detection methods,while having a good balance between accuracy and complexity.

关 键 词：高分辨率遥感影像 变化检测 全尺度跳跃连接 注意力机制 多尺度监督 

分 类 号：P237[天文地球—摄影测量与遥感]