CRSNet: Cloud and Cloud Shadow Refinement Segmentation Networks for Remote Sensing Imagery

Cloud detection is a critical task in remote sensing image tasks. Due to the influence of ground objects and other noises, the traditional detection methods are prone to miss or false detection and rough edge segmentation in the detection process. To avoid the defects of traditional methods, Cloud and Cloud Shadow Refinement Segmentation Networks are proposed in this paper. The network can correctly and efficiently detect smaller clouds and obtain finer edges. The model takes ResNet-18 as the backbone to extract features at different levels, and the Multi-scale Global Attention Module is used to strengthen the channel and spatial information to improve the accuracy of detection. The Strip Pyramid Channel Attention Module is used to learn spatial information at multiple scales to detect small clouds better. Finally, the high-dimensional feature and low-dimensional feature are fused by the Hierarchical Feature Aggregation Module, and the final segmentation effect is obtained by up-sampling layer by layer. The proposed model attains excellent results compared to methods with classic or special cloud segmentation tasks on Cloud and Cloud Shadow Dataset and the public dataset CSWV.

Automated summary

A new cloud detection method is proposed in this paper, which can accurately and efficiently detect smaller clouds and obtain finer edge segmentation. By using ResNet-18 as the backbone, and combining the Multi-scale Global Attention Module and Strip Pyramid Channel Attention Module, the detection accuracy of clouds is improved. The Hierarchical Feature Aggregation Module fuses high-dimensional and low-dimensional features, and the final segmentation effect is obtained by layer-by-layer upsampling. The proposed model achieves excellent results on the Cloud and Cloud Shadow Dataset and the public dataset CSWV.