A1 Vertaisarvioitu alkuperäisartikkeli tieteellisessä lehdessä
Golden Eagle Optimization for accelerating intra-mode estimation in HEVC
Tekijät: Tariq, Junaid; Javed, Imran; Rahman, Hameedur; Hussain, Ashfaq; Saeed, Kamran; Ijaz, Amir
Kustantaja: Springer Science and Business Media LLC
Julkaisuvuosi: 2026
Lehti: Multimedia Tools and Applications
Artikkelin numero: 25
Vuosikerta: 85
Numero: 1
eISSN: 1573-7721
DOI: https://doi.org/10.1007/s11042-026-21160-6
Julkaisun avoimuus kirjaamishetkellä: Ei avoimesti saatavilla
Julkaisukanavan avoimuus : Osittain avoin julkaisukanava
Verkko-osoite: https://doi.org/10.1007/s11042-026-21160-6
Tiivistelmä
The optimization of advanced video coding standards remains an ongoing challenge. This study addresses the computational challenges arising from the increasing number of intra-modes in High Efficiency Video Coding (HEVC) and proposes a solution through the Golden Eagle Optimization (GEO) algorithm. Skipping coding steps may reduce computational cost but often increases the Bjontegaard Delta Bit Rate (BD-BR) overhead. Therefore, GEO is used in this study to balance BD-BR increase and complexity reduction. The study analyzes GEO’s modules to align them with those of HEVC. In the proposed analogy, the eagle represents the Rate-Distortion Optimization (RDO) module, while rabbits correspond to intra-modes. Finally, the parameters of GEO are fine-tuned for two scenarios: (i) saving BD-BR or (ii) saving time. Experimental results demonstrate a significant 35.16% reduction in complexity, with only a slight 1.55% BD-BR overhead. These findings contribute to advancements in video compression technologies and encourage further research.
The optimization of advanced video coding standards remains an ongoing challenge. This study addresses the computational challenges arising from the increasing number of intra-modes in High Efficiency Video Coding (HEVC) and proposes a solution through the Golden Eagle Optimization (GEO) algorithm. Skipping coding steps may reduce computational cost but often increases the Bjontegaard Delta Bit Rate (BD-BR) overhead. Therefore, GEO is used in this study to balance BD-BR increase and complexity reduction. The study analyzes GEO’s modules to align them with those of HEVC. In the proposed analogy, the eagle represents the Rate-Distortion Optimization (RDO) module, while rabbits correspond to intra-modes. Finally, the parameters of GEO are fine-tuned for two scenarios: (i) saving BD-BR or (ii) saving time. Experimental results demonstrate a significant 35.16% reduction in complexity, with only a slight 1.55% BD-BR overhead. These findings contribute to advancements in video compression technologies and encourage further research.
Turun yliopiston Tutkimustietojärjestelmän portaali