| Popis: |
For the last few years, the ease of purchasing and using unmanned aerial vehicles (UAVs), their affordable cost has increased the demand for them both by companies and individuals. However, these devices might carry out illegal actions, starting with smuggling of illegal goods, unauthorized intelligence and computer attacks. As a result, this led to the urgency of developing effective and available countermeasures to detect and neutralize drones that perform reconnaissance of objects with confidential information. The most successful are autonomous systems for detecting and suppressing drones, which include optoelectronic, acoustic radar and radio frequency sensors, information from which is combined on the main computer to identify the threat and make further decisions. However, real-time monitoring is a rather difficult process that requires timely detection of adverse events or conditions. This creates many complex tasks such as object detection, classification, tracking multiple objects, and combining information from multiple sensors. In recent years, researchers have used various techniques to solve these problems and made notable progress. Applying deep learning to detect and classify UAVs is considered a new concept. In this regard, it became necessary to provide a generalized overview of UAV control technologies used for reconnaissance. |
