Naslov
Offloading Video Encoding Energy Consumption to the Decoder

Autori
Hofman, Daniel ; Benjak, Jakov

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni

Izvornik
International Multidisciplinary Conference on Computer and Energy Science (SpliTech) / - : Institute of Electrical and Electronics Engineers (IEEE), 2022, 1-5

Skup
2022 7th International Conference on Smart and Sustainable Technologies (SpliTech)

Mjesto i datum
Split, Hrvatska, 05.07.2022. - 08.07.2022

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
Energy Consumption , Image Processing , H.264 , H.265 , Video Coding , UAV

Sažetak
As the market for Unmanned Aerial Vehicles (UAVs) expands, there is a greater demand for improved technology for the next generation of UAVs. Newer UAVs feature greater resolution cameras, which adds to their total power consumption. The video encoding process can be altered to offload parts of the encoding process to the end-user side, which is typically less power constrained. In this research, a method for preparing the input camera stream for the encoder, without performing picture demosaicking is presented. With this approach, a considerable portion of the energy consumption may be offloaded to the decoder and executed while the video stream is being decoded. Our findings demonstrate that the processing time required for encoding may be reduced by up to 54% for H.264 and up to 49% for H.265. The resulting video streams produced with standard encoding and using our approach have the same video quality measured by the Video Multimethod Assess-ment Fusion (VMAF). Besides the lower power consumption in the encoding process, there is an additional benefit of lowered bandwidth of up to 42 % for H.264 encoders. This also reduces the power consumption of the UAV by lowering the network adapter's utilisation.

Izvorni jezik
Engleski

Znanstvena područja
Informacijske i komunikacijske znanosti

POVEZANOST RADA