Chinese scientists have created a new energy-efficient object detection system for UAVs
[ad_1]
Scientists were able to create a more energy-efficient object detection system for unmanned aerial vehicles (UAVs). It can provide the UAV with the ability to detect objects and targets without significantly increasing energy consumption. This is reported by Tech Xplore. In the world, UAVs are widely used for various purposes: from monitoring the natural environment to searching for survivors of natural disasters. Also, in the conditions of the modern war between Russia and Ukraine, UAVs have become an important tool for achieving superiority on the battlefield. To improve the ability of UAVs to detect what is needed, scientists are constantly developing new computational methods. The flight path of the UAV at different altitudes is represented by the red line, and the uniform square segments represent the detection areas captured by the camera at higher altitude (left) and lower altitude (right). An increase in height corresponds to an expanded field of view, which, as a result, leads to a decrease in the length of the detection path. (Suo et al.) In particular, China, which is the main supplier of critical components for Russian drones, is active in this matter. Recently, researchers from Yunnan University and the Chinese Academy of Sciences presented a new object detection system. It is based on boundary calculations. This know-how can give UAVs the ability to spot objects without significantly increasing energy consumption. The method was described in detail in the IEEE Internet of Things Journal. The object detection system was named E3-UAV. In application, edge computing uses multiple networks or devices nearby to perform calculations faster and consume less power. In the case of a new system, networks are used to determine the parameters: UAV height, flight speed, etc. This will allow devices to detect objects in the field of view faster with the minimum possible use of energy. “We first present an efficient evaluation metric for real-world tasks and build a transparent energy consumption model based on hundreds of actual flight data to formalize the relationship between energy consumption and flight parameters,” Suo Zhang writes in her paper. He notes that they then present a lightweight, energy-efficient priority decision-making algorithm based on a large amount of actual flight data to help the system determine flight parameters. The effectiveness of the new system was tested in a series of simulations on an NVIDIA graphics processor. It was used with the Mavic Air 2 drone, which is often used for aerial photography and video. The team noted that in real-world scenarios, energy consumption could also be significantly reduced. Earlier it became known that the German concern will transfer new generation drones to Ukraine. We also wrote about the most interesting technological developments that surprised humanity. Read also: 11 British universities are accused of helping to develop Iranian kamikaze drones
[ad_2]
Original Source Link
