Parallel Implementation of Discrete Cosine Transform (DCT) Methods on GPU for HEVC
Küçük Resim Yok
Tarih
2024
Yazarlar
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
Springer Science and Business Media Deutschland GmbH
Erişim Hakkı
info:eu-repo/semantics/restrictedAccess
Özet
The High-Efficiency Video Coding (HEVC) standard represents a significant advancement in video compression technology, offering superior compression efficiency compared to its predecessors. This study investigates into the parallel implementation and efficacy of the Discrete Cosine Transform (DCT) process, a pivotal component of the HEVC standard’s transform step, when executed on GPU hardware. To enhance the performance of DCT, two distinct methods were introduced and implemented, making efficient use of shared memory resources within the GPU. A comprehensive analysis was conducted by measuring the execution times of these methods using a standardized test image. The study rigorously compared the performance of these methods across various scenarios and meticulously assessed their impact. In summary, this research contributes valuable insights into optimizing the DCT process within the HEVC standard through the parallel implementation on GPU hardware. The proposed methods exhibit promising results, demonstrating the potential to enhance the efficiency of HEVC’s transform step.
Açıklama
Anahtar Kelimeler
DCT; GPU; HEVC
Kaynak
EAI/Springer Innovations in Communication and Computing - 2nd International Congress of Electrical and Computer Engineering, ICECENG 2023 Bandirma 22 November 2023 through 25 November 2023
WoS Q Değeri
Scopus Q Değeri
N/A
