Buildings' segmentation is a fundamental task in the field of earth observation and aerial imagery analysis. Most existing deep learning based algorithms in the literature can be applied to fixed or narrow-ranged spatial resolution imagery. In practical scenarios, users deal with a wide spectrum of images resolution and thus, often need to resample a given aerial image to match the spatial resolution of the dataset used to train the deep learning model. This however, would result in a severe degradation in the quality of the output segmentation masks. To deal with this issue, we propose in this research a Scale-invariant neural network (Sci-Net) that is able to segment buildings present in aerial images at different spatial resolutions. Specifically, we modified the U-Net architecture and fused it with dense Atrous Spatial Pyramid Pooling (ASPP) to extract fine-grained multi-scale representations. We compared the performance of our proposed model against several state-of-the-art models on the Open Cities AI dataset, and showed that Sci-Net provides a steady improvement margin in performance across all resolutions available in the dataset.

Paper preprint can be fetched here.

Dr Ali J. Ghandour
Dr Ali J. Ghandour
Associate Researcher

My research interests include earth observation, smart city transportation, urban features detection from high resolution aerial imagery and geospatial deep learning.