Abstract
Urban land cover and land use mapping plays an important role in urban planning and management. In this paper, novel multi-scale deep learning models, namely ASPP-Unet and ResASPP-Unet are proposed for urban land cover classification based on very high resolution (VHR) satellite imagery. The proposed ASPP-Unet model consists of a contracting path which extracts the high-level features, and an expansive path, which up-samples the features to create a high-resolution output. The atrous spatial pyramid pooling (ASPP) technique is utilized in the bottom layer in order to incorporate multi-scale deep features into a discriminative feature. The ResASPP-Unet model further improves the architecture by replacing each layer with residual unit. The models were trained and tested based on WorldView-2 (WV2) and WorldView-3 (WV3) imageries over the city of Beijing. Model parameters including layer depth and the number of initial feature maps (IFMs) as well as the input image bands were evaluated in terms of their impact on the model performances. It is shown that the ResASPP-Unet model with 11 layers and 64 IFMs based on 8-band WV2 imagery produced the highest classification accuracy (87.1% for WV2 imagery and 84.0% for WV3 imagery). The ASPP-Unet model with the same parameter setting produced slightly lower accuracy, with overall accuracy of 85.2% for WV2 imagery and 83.2% for WV3 imagery. Overall, the proposed models outperformed the state-of-the-art models, e.g., U-Net, convolutional neural network (CNN) and Support Vector Machine (SVM) model over both WV2 and WV3 images, and yielded robust and efficient urban land cover classification results.
Highlights
Urban land use and land cover mapping is a fundamental task in urban planning and management.Very High Resolution (VHR) remote sensing satellite imagery (Ground Sampling Distance (GSD) < 5 m) such as those acquired by QuickBird, IKONOS, GeoEye, WorldView-2/3/4, and GaoFen-2 have shown great advantage in urban land monitoring due to the spatial details they provide
Compared to atrous spatial pyramid pooling (ASPP)-Unet and U-Net models, we found that the ResASPP-Unet model was less sensitive to layer depth, as the information copied from shallower layers ensured that training errors did not change substantially
We proposed the ASPP-Unet and ResASPP-Unet models for urban land cover classification from high spatial resolution satellite images
Summary
Urban land use and land cover mapping is a fundamental task in urban planning and management.Very High Resolution (VHR) remote sensing satellite imagery (Ground Sampling Distance (GSD) < 5 m) such as those acquired by QuickBird, IKONOS, GeoEye, WorldView-2/3/4, and GaoFen-2 have shown great advantage in urban land monitoring due to the spatial details they provide. Many research efforts have been made on urban land use and land cover classification based on VHR. The classification methods can be generally categorized into two classes, i.e., pixel-based methods and object-based methods [2,5,6] The former defines classes for individual pixels mainly based on the spectral information. In VHR imagery, pixel-based methods may cause salt-and-pepper problems because the spectral responses of individual pixels do not represent the characteristics of the surface object. In contrast to the pixel-based methods, object-based methods merge neighboring pixels into objects using image segmentation techniques such as Multi-Resolution [7], Mean-Shift [8], or Quadtree-Seg [9] approaches, and the objects are treated as classification units. Analysts need to first arbitrarily select a specific set of object features, or extract representative features using feature engineering techniques as classification inputs
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
