With the development of Web2.0 and mobile Internet, urban residents, a new type of “sensor”, provide us with massive amounts of volunteered geographic information (VGI). Quantifying the spatial patterns of VGI plays an increasingly important role in the understanding and development of urban spatial functions. Using VGI and social media activity data, this article developed a method to automatically extract and identify urban spatial patterns and functional zones. The method is put forward based on the case of Beijing, China, and includes the following three steps: (1) Obtain multi-source urban spatial data, such as Weibo data (equivalent to Twitter in Chinese), OpenStreetMap, population data, etc.; (2) Use the hierarchical clustering algorithm, term frequency-inverse document frequency (TF-IDF) method, and improved k-means clustering algorithms to identify functional zones; (3) Compare the identified results with the actual urban land uses and verify its accuracy. The experiment results proved that our method can effectively identify urban functional zones, and the results provide new ideas for the study of urban spatial patterns and have great significance in optimizing urban spatial planning.
Studying urban expansion from a longer-term perspective is of great significance to obtain an in-depth understanding of the process of urbanization. Remote sensing data are mostly selected to investigate the long-term expansion of cities. In this study, we selected the world-class urban agglomeration of Beijing-Tianjin-Hebei (BTH) as the study area, and then discussed how to make full use of multi-source, multi-category, and multi-temporal spatial data (old maps and remote sensing images) to study long-term urbanization. Through this study, we addressed three questions: (1) How much has the urban area in BTH expanded in the past 100 years? (2) How did the urban area expand in the past century? (3) What factors or important historical events have changed the development of cities with different functions? By comprehensively using urban spatial data, such as old maps and remote sensing images, geo-referencing them, and extracting built-up area information, a long-term series of urban built-up areas in the BTH region can be obtained. Results show the following: (1) There was clear evidence of dramatic urban expansion in this area, and the total built-up area had increased by 55.585 times, from 126.181 km2 to 7013.832 km2. (2) Continuous outward expansion has always been the main trend, while the compactness of the built-up land within the city is constantly decreasing and the complexity of the city boundary is increasing. (3) Cities in BTH were mostly formed through the construction of city walls during the Ming and Qing dynasties, and the expansion process was mostly highly related to important political events, traffic development, and other factors. In summary, the BTH area, similarly to China and most regions of the world, has experienced rapid urbanization and the history of such ancient cities should be further preserved with the combined use of old maps.
The growing volume of Earth science data available from climate simulations and satellite remote sensing offers unprecedented opportunity for scientific insight, while also presenting computational challenges. One potential area of impact is atmospheric correction, where physics-based numerical models retrieve surface reflectance information from top of atmosphere observations, and are too computationally intensive to be run in real time. Machine learning methods have demonstrated potential as fast statistical models for expensive simulations and for extracting credible insights from complex datasets. Here, we develop DeepEmSat: a deep learning emulator approach for atmospheric correction, and offer comparison against physics-based models to support the hypothesis that deep learning can make a contribution to the efficient processing of satellite images.
Self-training has greatly facilitated domain adaptive semantic segmentation, which iteratively generates pseudo labels on the target domain and retrains the network. However, since the realistic segmentation datasets are highly imbalanced, target pseudo labels are typically biased to the majority classes and basically noisy, leading to an errorprone and sub-optimal model. To address this issue, we propose a region-based active learning approach for semantic segmentation under a domain shift, aiming to automatically query a small partition of image regions to be labeled while maximizing segmentation performance. Our algorithm, Active Learning via Region Impurity and Prediction Uncertainty (AL-RIPU), introduces a novel acquisition strategy characterizing the spatial adjacency of image regions along with the prediction confidence. We show that the proposed region-based selection strategy makes more efficient use of a limited budget than image-based or pointbased counterparts. Meanwhile, we enforce local prediction consistency between a pixel and its nearest neighbor on a source image. Further, we develop a negative learning loss to enhance the discriminative representation learning on the target domain. Extensive experiments demonstrate that our method only requires very few annotations to almost reach the supervised performance and substantially outperforms state-of-the-art methods.* Corresponding author (a) Target image (b) Image-based selection (100%) (c) Point-based selection (2.2%) (d) Region-based selection (2.2%)
With the exponential increase in the volume of automatic dependent surveillance-broadcast (ADS-B), and other types of air traffic control (ATC) data containing spatiotemporal attributes, it remains uncertain how to respond to immediate ATC data access within a target area. Accordingly, an original multi-level disaggregated framework for airspace, and its corresponding information management is proposed. Further, a multi-scale grid modeling and coding mapping method of airspace information represented by ADS-B is put forth. Finally, tests on the validity of the 4D airspace-temporal grid we named as the GeoSOT-AS framework were conducted across key areas based on the development of an effective data organization method for ADS-B, or an effective algorithm for extracting relevant spatiotemporal data. Experimentally, it was demonstrated that GeoSOT-AS conforms to the existing Chinese specification of civil aeronautical charting and is advantageous for its low deformation and high practicality; furthermore, the airspace grid identification code modeling was less costly, and improved performance by >80% when used for ADS-B data extraction. GeoSOT-AS can thus provide effective reference and practical information for existing airspace data management methods represented by ADS-B and can subsequently be extended to other forms of airspace management scenarios.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.