Using Convolutional Networks and Satellite Imagery to Identify Patterns in Urban Environments at a Large Scale

Abstract

Urban planning applications (energy audits, investment, etc.) require an understanding of built infrastructure and its environment, i.e., both low-level, physical features (amount of vegetation, building area and geometry etc.), as well as higher-level concepts such as land use classes (which encode expert understanding of socioeconomic end uses).is kind of data is expensive and laborintensive to obtain, which limits its availability (particularly in developing countries). We analyze pa erns in land use in urban neighborhoods using large-scale satellite imagery data (which is available worldwide from third-party providers) and state-of-theart computer vision techniques based on deep convolutional neural networks. For supervision, given the limited availability of standard benchmarks for remote-sensing data, we obtain ground truth land use class labels carefully sampled from open-source surveys, in particular the Urban Atlas land classi cation dataset of 20 land use classes across 300 European cities. We use this data to train and compare deep architectures which have recently shown good performance on standard computer vision tasks (image classi cation and segmentation), including on geospatial data. Furthermore, we show that the deep representations extracted from satellite imagery of urban environments can be used to compare neighborhoods across several cities. We make our dataset available for other machine learning researchers to use for remote-sensing applications.

Automatic vegetation identification in Google Earth images using a convolutional neural network: A case study for Japanese bamboo forests

“…Additionally, even when we applied a model that learned all traning images (all), the performance of model was not as good as when traning data was obtained within the same city. The same tendencies are reported in studies that classified land use using deep learning (Albert et al, 2017). This may suggest that increasing the number of training data may also lead to a decrease in identification accuracy and it is difficult to construct an identification model applicable to a broad area.…”

Apache Spark Accelerated Deep Learning Inference for Large Scale Satellite Image Analytics

“…As noted above, computer vision and machine learning algorithms are proving superior in providing automated means to describe the distinctive nature of objects in remotely-sensed image data [1], [5], [6], [29]. However, the deployment of such algorithms remains a significant challenge when considered on large geographic areas covered by hundreds of thousands of images [29].…”

Infrastructure Quality Assessment in Africa using Satellite Imagery and Deep Learning

“…Such rich and high quality image data enable advanced machine learning techniques to perform sophisticated tasks like object detection and classification, and deep learning in particular has shown great promise [2,12,13,16]. While a number of recent papers discuss the use of deep learning on satellite imagery for applications in land use cover [2], urban planning [3], environmental science [6], etc. [25,34,42], many unanswered questions remain in the field, particularly in the application of deep learning to social and economic development.…”

“…Manuscript submitted to ACM study [2] uses state-of-the-art deep CNNs VGG-16 [38] and Residual Neural Networks [16] to analyze land use in urban neighborhoods with large scale satellite data.…”