RoadDetections
This resource provides a comprehensive dataset of 54.2 million kilometers of roads detected worldwide using Microsoft Maps aerial imagery, including Maxar and Airbus sources. The data is openly accessible under the Open Data Commons Open Database License (ODbL), with plans to open-source the underlying AI model and geometry generation code in 2025.
Microsoft's RoadDetections project provides a vast dataset of 54.2M km of roads globally, meticulously mined from Bing Maps aerial imagery, which includes contributions from Maxar and Airbus. This extensive data is freely available for download and use under the Open Data Commons Open Database License (ODbL), promoting accessibility for various applications.
The project aims for further transparency by planning to open-source both the Neural Network (NN) model and the geometry generation code during the first half of 2025. The data is organized regionally, offering segmented downloads for areas such as Northern America, South America, Eastern Europe, Southern Asia, and Australia, among others. Each regional package includes the detected road lengths in kilometers and the corresponding file size for easy access and management.
Tags:
Disclaimer: We do not guarantee the accuracy of this information. Our documentation of this website on Geospatial Catalog does not represent any association between Geospatial Catalog and this listing. This summary may contain errors or inaccuracies.
Comments (0)
Sign in to leave a comment