Video encoders for unmanned systems such as UAVs and drones, take the raw output from video cameras and imaging payloads and compress it, so that it requires less storage space and uses less bandwidth to transmit over radio, Ethernet, SATCOM and other communication methods. The algorithms used by encoder systems may reduce the amount of data by hundreds of times.
Video encoding, as well as the decoding required after transmission, is performed by hardware or software known as codecs. Codecs can use a variety of data formats and resolutions depending on the requirements for storage, transmission, processing, and image detail. High-quality video encoding standards such as H.264 and H.265 are becoming more widespread as the SWaP (size, weight and power) requirements of high performance embedded computing devices continue to shrink. These devices may enable HD (high-definition) and 4K (ultra-high definition) video encoding.
Video encoding can be performed by dedicated hardware devices or in software. Hardware video encoding solutions can devote all their processing power to encoding, whereas software running on a UAV’s main computing device will have to share CPU resources with other system functions. Hardware solutions will therefore provide lower-latency video encoding, meaning that the time taken for the video data to be processed and sent by the drone and received by the operators is closer to real time. This minimal lag is important for operators to maintain effective control of the aircraft as well as for real-time decision making based on information gathered by the aircraft’s imaging payload.
Successful low-latency video encoding and streaming requires very high data rates when broadcasting. This becomes harder to maintain the further a drone flies from its base station, and the video data will have to be compressed more. Eventually the video stream will become too low-resolution to make out useful details. The application of video compression is thus a tradeoff between image quality, bandwidth requirements and operational distance.