MOSCOW, April 29. /TASS/. Rostec has developed a line of optoelectronic systems, some of which use a neural network for automatic object recognition, the Russian state tech corporation told TASS.
"Rosel Holding, a subsidiary of Rostec, has developed a line of optoelectronic systems for unmanned aerial vehicles (UAVs), ground-based robotic systems and surface robotic vessels. The equipment can detect moving and stationary targets at any time of day. Two modifications of the new line of products additionally implement neural network data processing algorithms running on the device's built-in computer. This enables automatic object recognition and reduces operator workload," the statement reads.
The line includes seven models, varying in weight, dimensions, and detection range. All models are energy-efficient. They are built on a gyrostabilized platform with two-axis rotation, ensuring stable imaging while the carrier is moving.
The devices operate in the visible spectrum, and some models are additionally equipped with infrared channels for low-light monitoring, with image output in a standard video format.
The most compact systems are designed for installation on UAVs. The medium class solutions are for robotic ground platforms. Larger ones can be used on unmanned boats and warships. The equipment can be used, among other things, for traffic monitoring by traffic police officers, as well as for forest surveillance and fire detection.