KB Panorama has released version 2.0.1 of its Panorama Vision automated recognition and vectorization system. The update introduces support for processing large-scale territories, including cities, regions, and republics, in a single request. The system automatically divides large areas into smaller segments for processing, ensuring that vector contours are aligned at the boundaries to prevent gaps in the final map.
Users can access intermediate results before the entire processing task is complete, which is useful for large-scale projects that may take several days. The system supports the vectorization of buildings, roads, forests, agricultural fields, and hydrographic objects, as well as the identification of vehicles on images. Access to the service is provided through the GIS Panorama application suite.
The new functionality was tested on a forest map of the Republic of Bashkortostan, covering an area of approximately 247,000 square kilometers. The processing took seven days on a single server and resulted in the creation of about 400,000 vector objects. The generated data was subsequently used to update the OpenStreetMap database.
