Sergei Sadkov
Портфолио
Pixel AI
Further development of the “DigitalEarth” project aimed to deliver remote sensing analysis to business and society. Was on beta-testing till November 2022. For that project I’ve improved my code for creating deep learning training sets, and also for the remote sensing data visualization, to provide optimal space image looks and readability.
Atlas VR
Virtual reality platform using satellite imagery and thematic maps to create digital twins of landscapes and infrastructure for education, tourism, environmental, and business issues. In this project I was preparing data on space imagery, topography and thematic maps, like forestry of forest canopy height maps. Also I provided the source data reprojection to minimize errors between different map layers.
Digital Earth
The project on cloud software solutions for automated satellite data analysis using AI-technology. My responsibility was to prepare training sets for deep learning to get landcover contours from the remote sensing data. It included remote sensing data refinement and reprojection, correcting the manual interpretation results (geometry and attributes) and creating set of training data with hundreds of images.