基于双注意力机制与UNet的遥感影像变化检测方法

Remote sensing image change detection method based on double attention mechanism and Unet

遥感图像变化检测长期以来都是遥感领域的重要的研究方向。传统的深度学习语义分割模型难以充分地提取两期遥感影像中的变化信息,为了解决此问题,文章提出了一种基于双注意力机制的UNet遥感影像变化检测模型。该方法首先通过图像融合将两期影像送入UNet之中。而后通过在跳跃级联与特征提取的高层次特征后使用双注意力机制模块来建立起丰富的上下文信息与凸显特征中的变化信息。最后通过反卷积来恢复尺寸获取变化二值图。实验结果表明,所提出的方法提高了遥感影像变化检测的精度。

Remote sensing image change detection has long been an important research direction in the field of remote sensing.The traditional deep learning semantic segmentation model is difficult to fully extract the change information in two periods of remote sensing images. In order to solve this problem, this paper proposes a remote sensing change detection model of UNet images based on double attention mechanism. This method first sends the two images into UNet through image fusion. Then,after the high-level features of jump cascade and feature extraction, the dual attention mechanism module is used to establish rich context information and change information in salient features. Finally, the size is restored by deconvolution to obtain the changing binary image. Experimental results show that the proposed method improves the accuracy of remote sensing image change detection.