Adversarial Shape Learning for Building Extraction in VHR Remote Sensing Images
This work addresses the problem of accurate building segmentation for remote sensing applications, representing an incremental advance by integrating shape constraints into existing CNN methods.
The paper tackles building extraction in very high-resolution remote sensing images by addressing occlusion and boundary ambiguity through adversarial shape learning, resulting in significant improvements in both pixel-based accuracy and object-based quality metrics on two benchmark datasets.
Building extraction in VHR RSIs remains a challenging task due to occlusion and boundary ambiguity problems. Although conventional convolutional neural networks (CNNs) based methods are capable of exploiting local texture and context information, they fail to capture the shape patterns of buildings, which is a necessary constraint in the human recognition. To address this issue, we propose an adversarial shape learning network (ASLNet) to model the building shape patterns that improve the accuracy of building segmentation. In the proposed ASLNet, we introduce the adversarial learning strategy to explicitly model the shape constraints, as well as a CNN shape regularizer to strengthen the embedding of shape features. To assess the geometric accuracy of building segmentation results, we introduced several object-based quality assessment metrics. Experiments on two open benchmark datasets show that the proposed ASLNet improves both the pixel-based accuracy and the object-based quality measurements by a large margin. The code is available at: https://github.com/ggsDing/ASLNet