The AI-powered platform ATLAS enables to automate object detection on drone and satellite imagery to detect and count or measure the identified objects of interest.
“Companies performing large scale drone inspections have to manually filter tens of thousands of images or analyze many hectares of orthophoto maps to identify data for decision-making. This applies to such industries as solar farms, wind turbines, utility inspections, road maintenance, pipeline monitoring, agriculture, and many others. ATLAS provides enhanced analytical capability for drone inspections to save days of manual processing” says Alexei Yankelevich, R&D Director of SPH Engineering.
Manual image annotation and report preparation consists of many repeating tasks and tends to be error-prone and time-consuming.
ATLAS introduces a straightforward visual workflow of 3 easy steps:
- uploading the images to ATLAS;
- annotating a small amount of data manually to training the AI-detector;
- processing the rest of your data automatically.
ATLAS can detect a wide range of objects of very different nature and shapes on raw imagery and orthophoto maps. ATLAS can be trained to identify uncomplicated objects as cars, boats, people, road signs, and also more complex shapes like scratches and cracks on solar panels, vegetation, water resources, and others.
ATLAS is available as a cloud-based as well as on-premises deployed solution. It is developed by SPH Engineering - the creator of mission planning software UgCS.