We provide lightweight AI-based solutions for object detection, tracking, and image classification, as well as efficient implementation of classic computer vision methods. These solutions can run on a server, or we can embed them on edge devices for real-time drone operation. This enables our solutions to be used for drone autonomy or to guide drone operators, as we also provide visualization of results in a custom-tailored user interface.
For example, a search and rescue-oriented interface can highlight regions of interest for the operator, zooming in on those areas and applying image processing to emphasize relevant features for faster decision-making. The specific interface and processing capabilities are defined by the client and can be adapted to various use cases.
In addition, our engineering team is experienced (10+ years) with conventional computer vision stack: camera video processing, image transofmations, visual annotations and informative overlays, camera mono and stereo calibrations, depth extraction.