Self-learning-based post-processing for image/video deblocking via sparse representation

•First MCA self-learning-based image decomposition framework.•Fully automatic and self-contained learning, no extra training samples required.•Highly adaptable to any block transform-based compressed image/video.•Adaptable with no info on data source, coding bitrates, and compression algorithms.•Ach...

Celý popis

Uloženo v:

Podrobná bibliografie
Vydáno v:	Journal of visual communication and image representation Ročník 25; číslo 5; s. 891 - 903
Hlavní autoři:	Yeh, Chia-Hung, Kang, Li-Wei, Chiou, Yi-Wen, Lin, Chia-Wen, Fan Jiang, Shu-Jhen
Médium:	Journal Article
Jazyk:	angličtina
Vydáno:	Amsterdam Elsevier Inc 01.07.2014 Elsevier
Témata:	Algorithms Applied sciences Artificial intelligence Blocking Blocking artifact Coding Coding, codes Computer science; control theory; systems Deblocking Decomposition Detection, estimation, filtering, equalization, prediction Dictionary learning Exact sciences and technology Image/video enhancement Image/video restoration Information, signal and communications theory Morphological component analysis Pattern recognition. Digital image processing. Computational geometry Post-processing Representations Signal and communications theory Signal, noise Sparse representation Telecommunications and information theory Image/video restoration Blocking artifact Deblocking Post-processing Image/video enhancement Sparse representation Morphological component analysis Dictionary learning Block code Dictionaries Image processing Noise reduction Data compression Video signal Multivariate analysis Low frequency Block matching Correspondence analysis Morphological analysis Self learning Computer vision Filtering blurring image Blocking Image compression Software component Sparse coding Noisy image Image analysis Experimental result Boundary value problem Scene analysis High frequency Image enhancement
ISSN:	1047-3203, 1095-9076
On-line přístup:	Získat plný text
Tagy:	Přidat tag Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!

Buďte první, kdo okomentuje tento záznam!