Aerial Height Prediction and Refinement Neural Networks with Semantic and Geometric Guidance
This work provides an incremental improvement in aerial height prediction, which is important for applications like urban planning and environmental monitoring.
This paper proposes a two-stage neural network approach for predicting height maps from single RGB aerial images. The method first uses a multi-task network for initial prediction, followed by a denoising autoencoder for refinement, achieving state-of-the-art results on two public datasets.
Deep learning provides a powerful new approach to many computer vision tasks. Height prediction from aerial images is one of those tasks that benefited greatly from the deployment of deep learning which replaced old multi-view geometry techniques. This letter proposes a two-stage approach, where first a multi-task neural network is used to predict the height map resulting from a single RGB aerial input image. We also include a second refinement step, where a denoising autoencoder is used to produce higher quality height maps. Experiments on two publicly available datasets show that our method is capable of producing state-of-the-art results. Code is available at https://github.com/melhousni/DSMNet.