Search

Deep convolutional networks have been the most competitive method in remote sensing scene classification. Due to the diversity and complexity of scene content, remote sensing scene classification still remains a challenging task. Recently, the second-order pooling method has attracted more interest because it can learn higher-order information and enhance the nonlinear modeling ability of the networks. However, how to effectively learn second-order features and establish the discriminative feature representation of holistic images is still an open question. In this letter, we propose a first and second-order information fusion network (FSoI-Net) that can learn the first-order and second-order features at the same time, and construct the final feature representation by fusing the two types of features. Specifically, a self-attention-based second-order pooling (SaSoP) method based on covariance matrix is proposed to extract second-order features, and a fusion loss function is developed to jointly train the model and construct the final feature representation for the classification decision. The proposed network has been thoroughly evaluated on three real remote sensing scene datasets and achieved better performance than the counterparts.

Impervious surfaces have important effects on the natural environment, including promoting hydrological run-off and impeding evapotranspiration, as well as increasing the urban heat island effect. Obtaining accurate and timely information on the spatial distribution and dynamics of urban surfaces is, thus, of paramount importance for socio-economic analysis, urban planning, and environmental modeling and management. Previous studies have indicated that the fusion of multi-source remotely sensed imagery can increase the accuracy of prediction for impervious surface information across large areas. However, the majority of them are limited to the use of specific data sources to construct a few features with which it can be challenging to characterize adequately the variation in impervious surfaces over large areas. Thus, impervious surface maps are often presented with high uncertainty. In response to this problem, we proposed the use of multi-temporal MODIS and Visible Infrared Imaging Radiometer Suite (VIIRS) nighttime light data to construct a more general and robust feature set for large-area artificial impervious surface percentage (AISP) prediction. Three fusion methods were proposed for application to multi-temporal MODIS surface reflectance product (MOD09A1) and Visible Infrared Imaging Radiometer Suite (NPP-VIIRS) Day/Night Band (DNB) data to construct three different types of features: spectral features, index features (band calculations), and fusion features. These features were then used as variables in a random-forest-based AISP prediction model. The model was fitted to China and then applied to predict AISP across Asia. Fifteen typical cities from different regions of Asia were selected to assess the accuracy of the prediction model. The use of multi-temporal MODIS and VIIRS DNB data was found to significantly increase the accuracy of prediction for large-area AISP. The feature set constructed in this research was demonstrated to be suitable for large-are

The detection of building edges from very high resolution (VHR) remote sensing imagery is essential to various geo-related applications, including surveying and mapping, urban management, etc. Recently, the rapid development of deep convolutional neural networks (DCNNs) has achieved remarkable progress in edge detection; however, there has always been the problem of edge thickness due to the large receptive field of DCNNs. In this paper, we proposed a multi-scale erosion network (ME-Net) for building edge detection to crisp the building edge through two innovative approaches: (1) embedding an erosion module (EM) in the network to crisp the edge and (2) adding the Dice coefficient and local cross entropy of edge neighbors into the loss function to increase its sensitivity to the receptive field. In addition, a new metric, Ene, to measure the crispness of the predicted building edge was proposed. The experiment results show that ME-Net not only detects the clearest and crispest building edges, but also achieves the best OA of 98.75%, 95.00% and 95.51% on three building edge datasets, and exceeds other edge detection networks 3.17% and 0.44% at least in strict F1-score and Ene. In a word, the proposed ME-Net is an effective and practical approach for detecting crisp building edges from VHR remote sensing imagery.

The detection of building edges from very high resolution (VHR) remote sensing imagery is essential to various geo-related applications, including surveying and mapping, urban management, etc. Recently, the rapid development of deep convolutional neural networks (DCNNs) has achieved remarkable progress in edge detection; however, there has always been the problem of edge thickness due to the large receptive field of DCNNs. In this paper, we proposed a multi-scale erosion network (ME-Net) for building edge detection to crisp the building edge through two innovative approaches: (1) embedding an erosion module (EM) in the network to crisp the edge and (2) adding the Dice coefficient and local cross entropy of edge neighbors into the loss function to increase its sensitivity to the receptive field. In addition, a new metric, Ene, to measure the crispness of the predicted building edge was proposed. The experiment results show that ME-Net not only detects the clearest and crispest building edges, but also achieves the best OA of 98.75%, 95.00% and 95.51% on three building edge datasets, and exceeds other edge detection networks 3.17% and 0.44% at least in strict F1-score and Ene. In a word, the proposed ME-Net is an effective and practical approach for detecting crisp building edges from VHR remote sensing imagery.