How combining AI & Video Encoding is changing the UAV game
“AI and video encoding technologies are revolutionizing the UAV/drone market by enhancing their capabilities and expanding their applications.
Artificial Intelligence (AI) is being integrated into drones to enable autonomous flight, obstacle avoidance, and intelligent navigation. With AI algorithms, drones can analyze their surroundings, detect and avoid obstacles, and make real-time decisions to ensure safe and efficient flight. This has opened up new possibilities for commercial applications such as aerial inspections, delivery services, and surveillance.
Moreover, AI-powered drones are being used for advanced data collection and analysis. They can capture high-resolution images and videos, and AI algorithms can process this data to extract valuable insights. For example, drones equipped with AI can identify objects, monitor crop health, detect anomalies in infrastructure, or assist in search and rescue operations. This combination of AI and drones is transforming industries like agriculture, construction, and public safety.
Video encoding is another crucial aspect that is transforming the UAV/drone market. Drones are equipped with high-quality cameras that capture large amounts of video data. However, transmitting and storing this data in real-time can be challenging due to bandwidth limitations and storage constraints. Video encoding technologies address this issue by compressing the video data without significant loss in quality.
Advanced video encoding algorithms, such as H.264 and H.265, enable efficient compression and transmission of video data from drones. This allows for real-time streaming of high-definition video footage to ground stations or remote operators. Additionally, video encoding reduces the storage requirements, enabling longer flight times and more efficient data management.
The combination of AI and video encoding technologies is also enhancing the capabilities of drones in terms of video analytics. AI algorithms can be applied to the compressed video data, enabling real-time object detection, tracking, and analysis. This enables drones to perform tasks like automated surveillance, crowd monitoring, or traffic analysis.
Overall, AI and video encoding technologies are transforming the UAV/drone market by enabling autonomous flight, intelligent navigation, advanced data collection, and real-time video analytics. These advancements are expanding the applications of drones across various industries and revolutionizing the way tasks are performed, making drones more efficient, versatile, and valuable tools.
The combination of an ultra small video encoder engine and AI accelerator chip allows for the video to be process in parallel and data on tracking identification and classification to be overlayed near real time over the video.
Decisions can be made based on this data /coordinates which can be output locally or sent to the ground over IP.
AI data locally can be used to steer the autopilot or control a gimbal. AI data sent to the ground ,either visually or as raw data, can be used by the flight controller to navigate steer or take decisions on the next course of action.
Finally autonomous decision making is also possible, for example weed detection and classification can allow crop spraying drones to target weeds rather than spraying whole fields. In a military environment classification of targets allows autonomous decisions to be made as to which targets to hit.
A small low power 50mm x25mm x 15mm module assembly is now capable of encoding multiple cameras as well as processing raw video through an AI accelerator chip. In addition streaming real time and recording locally. A further innovation is the ability to record RAW video which can be used on the ground to develop new AI algorithms based on real video recordings with no compression. This is a better way of developing AI based on real life drone recordings.
Manufacturers of UAV and other autonomous vehicles can now add multiple camera technologies for differing situations. For example a zoom block from Sony which can be used for detailed HD and 4K images but with analogue zoom of 20-40x in addition to Thermal cameras and situational awareness low resolution cameras. These situational awareness cameras can be used for collision avoidance whilst the Thermal for night vision or heat detection (disaster recovery scenarios etc) The ultra-small module mentioned can stream all these cameras simultaneously , record locally and allow the AI chip to track detect and classify any objects. An ultra-powerful ready to go solution
AI code can be sourced as open source (examples YOLOv5 YOLOv7 etc ) or purchased from specialist developers or developed from scratch. The module allows direct upload of any of the AI applications to the on board memory.
The central processor has 4 cores and near 90% of the capability is available to developers to host their own Linux applications. This could be flight control algorithms or application specific code to take AI data and translate to Gimble control or autopilot input.
In summary AI capability combined with multi camera streaming will allow the UAV developers to refine the way in which the drone is used without the traditional tether to a ground controller.”
Written by Les Litwin – Technical Director at Antrica.
