Semantic Segmentation of Satellite Images Using a U-Shaped Fully Connected Network with Dense Residual Blocks

摘要

Semantic segmentation is the task of clustering pixels into an object class. In the field of remote sensing semantic segmentation has wide applications ranging from scene cover classification to change detection for scene understanding. With the success of deep learning algorithms for classification tasks, there has been much work to apply convolutional neural networks in remote sensing with much success. However, feature extraction of high resolution remote sensing imagery poses a challenge when applying such networks. In particular, there is a need to extract high level features while maintaining an objects resolution in the networks feature space. This work proposes an efficient deep fully convolution architecture that obtains high level features without loss of spatial resolution by replacing the standard convolutional layers in U-Net with dense residual blocks. By stacking identity blocks, we allow the input to flow through the network at every proceeding layer. Our network is termed DRU-Net, and is shown to outperform standard U-Net.