Scientists in Samara have developed an artificial intelligence system with stereoscopic vision for drones.
Samara has created a navigation system with artificial intelligence for drones.
Author: Maria Ivanova
According to RIA Novosti, Samara has completed the development of an artificial intelligence stereoscopic vision navigation system for drones. This system enables drones to safely avoid obstacles and complete landings without operator involvement.
Scholars from Samara University have developed an artificial intelligence system with stereoscopic vision for drones, which will enhance the autonomy and safety of drone operations.
The software system called "Navigator" allows drones to fly autonomously along preset routes, avoid restricted flight areas, and respond to sudden obstacles such as birds or oncoming drones.
One of the key features of "Navigator" is stereoscopic vision - the system uses dual video stream technology to accurately measure distances to objects and analyze terrain from tens of meters away. Typically, even drones equipped with multiple cameras mostly use monocular vision; this new development can comprehensively analyze space.
The built-in artificial intelligence analyzes images, determining whether selected landing sites are safe based on the presence of personnel, equipment, or other obstacles. This is particularly important during emergency landings. Additionally, "Navigator" can determine location even when GPS signals are lost - the artificial intelligence compares real-time video images with pre-segmented terrain samples.
Currently, the visual module of this technology is undergoing testing. This system was developed according to an order from "Future Transport," a company focused on drone research. In the future, plans are in place to expand the functionality of "Navigator" to build a secure domestic drone aviation ecosystem.
As previously reported by Vzglyad, Rostec has already started production of the "Identification Friend or Foe" system for drones.
Kazan has developed the "Avatar" fixed-wing drone, equipped with three cameras including a thermal imager and a 30x zoom lens, with a reconnaissance depth of up to 40 kilometers.
Moscow Aviation Institute has developed an autonomous flight controller that can control different types of drones and process onboard video signals.
Original article: https://www.toutiao.com/article/7507196721152164361/
Disclaimer: The article solely represents the author's views. Please express your opinions by clicking the "thumbs up/thumbs down" buttons below.