针对普通跳跃连接缺乏从全尺度角度获取变化信息及编码器特征提取能力不足的问题，本文设计了一种耦合注意力机制(Convolutional Block Attention Module,CBAM)的UNet+++高分辨率遥感影像变化检测网络CBAM UNet+++。CBAM UNet+++基于融合全尺度特征的语义分割结构UNet+++，通过替换基本卷积单元为残差注意力模块(Residual Block_CBAM, ResBlock_CBAM)抑制背景影响，增强编码器对显著特征的学习能力。并在两种不同类型的高分辨率遥感影像变化检测数据集上进行验证。结果表明：该方法在LEBEDEV多地物变化数据集上取得最高精度，F1和OA值分别为88.9%、97.3%；在LEVIR-CD建筑物变化数据集上取得次高精度，F1和OA值分别为86.7%、96.8%；同时，该方法从定性及定量角度分析优于其它基准网络，对变化检测网络结构设计起到一定参考作用。

To address the problem of the lack of change information from the full-scale perspective and the lack of encoder feature extraction capability of the common jump connection. In this paper, we design a UNet+++ high-resolution remote sensing image change detection network with coupled attention mechanism (Convolutional Block Attention Module,CBAM). CBAM UNet+++ is based on the semantic segmentation structure UNet+++ that fuses full-scale features, and enhances the learning ability of the encoder for salient features by replacing the basic convolutional unit with the residual attention module (Residual Block_CBAM, ResBlock_CBAM) to suppress background effects. And it is validated on two different types of high-resolution remote sensing image change detection datasets. The results show that the method achieves the highest accuracy on LEBEDEV multi-object change dataset, with F1 and OA values of 88.9% and 97.3%, respectively, and the second highest accuracy on LEVIR-CD building change dataset, with F1 and OA values of 86.7% and 96.8%, respectively; meanwhile, the method outperforms other benchmark networks in qualitative and quantitative analysis, and serves as a reference for change detection The method also outperforms other benchmark networks in terms of qualitative and quantitative analysis, and plays a reference role in the design of network structure.