Video çerçeve hız artırımında kapatma yapaylıklarının giderilmesi
Yükleniyor...
Dosyalar
Tarih
2009-06-26
Yazarlar
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
IEEE
Erişim Hakkı
info:eu-repo/semantics/closedAccess
Özet
Çerçeve hız arttırımı (ÇHA) yöntemleri, önceleri video standartları arası dönüşümlerde kullanılmış günümüzde ise büyük ekran LCD ve plazma Tv setlerinde daha kaliteli görüntü için sayısal video yayınının zamansal çözünürlüğünün arttırılmasında uygulanmaya başlanmıştır. DSP, FPGA ve ASIC gibi donanımların işlem gücünün artması ÇHA probleminin çözümünde yüksek karmaşıklıkta algoritmaların tasarlanmasına olanak sağlamıştır.Bu bildiride çok aşamalı devinim vektör (DV) iyileştirilmesi art işlemleri ve kapanma/açılma bölgelerine uyarlamalı örtüşmeli blok devinim denkleştirmesi yöntemleri ile uzamsal ve zamansal düzlemde tutarlı bir ÇHA algoritması önermekteyiz. Hekzagonal devinim kestirim (DK) algoritmasının verdiği düşük güvenirlikli DV'leri çok aşamalı iyileştirme art-işlemlerine tabi tutulur. Uzamsal olarak tutarlı bir DV alanı elde edildikten sonra komşu vektörler arasındaki ilişkiler ve ardışık 4 çerçevenin DV bilgisi kullanılarak kapanma/açılma bölgeleri tespit edilir.Örtüşmeli blok devinim denkleştirme filtresi (ÖBDD) tespit edilen bu bölgelelere uyarlanarak kapanma/açılmadan kaynaklanan yapaylıkların azaltılması sağlanır. Böylece var olan ÇHA yöntemlerine göre uzamsal çözünürlükten daha az ödün verilerek ve daha az yapaylıkla ÇHA gerçekleştirilmektedir.
In the past, frame rate up-conversion (FRC) methods were commonly used in standards conversion. Today, FRC methods are used for digital video temporal resolution enhancement to improve the visual quality in large screen LCD and plasma TV sets. The increase in processing power of DSPs, FPGAs and ASICs enables high complexity algorithms in the design of FRC systems. In this paper, we propose a multi-stage motion vector (MV) post-processing algorithm with occlusion adaptive overlapped block motion compensation, which increases the temporal resolution without loss of spatial resolution and consistency. Hexagonal search block matching yields us unreliable MVs which are post-processed in the refinement stages in order to get a consistent MV field in spatial domain. Relations between neighboring MVs and the MV fields of 4 consecutive frames are utilized to detect occlusion regions. Overlapped Block Motion Compensation (OBMC) is adapted to these problematic regions in order to reduce artifacts due to occlusion. Compared to existing methods, the developed algorithm achieves FRC with reduced loss of spatial resolution and reduced amount of artifacts.
In the past, frame rate up-conversion (FRC) methods were commonly used in standards conversion. Today, FRC methods are used for digital video temporal resolution enhancement to improve the visual quality in large screen LCD and plasma TV sets. The increase in processing power of DSPs, FPGAs and ASICs enables high complexity algorithms in the design of FRC systems. In this paper, we propose a multi-stage motion vector (MV) post-processing algorithm with occlusion adaptive overlapped block motion compensation, which increases the temporal resolution without loss of spatial resolution and consistency. Hexagonal search block matching yields us unreliable MVs which are post-processed in the refinement stages in order to get a consistent MV field in spatial domain. Relations between neighboring MVs and the MV fields of 4 consecutive frames are utilized to detect occlusion regions. Overlapped Block Motion Compensation (OBMC) is adapted to these problematic regions in order to reduce artifacts due to occlusion. Compared to existing methods, the developed algorithm achieves FRC with reduced loss of spatial resolution and reduced amount of artifacts.
Açıklama
Anahtar Kelimeler
Adaptive algorithms, Application specific integrated circuits, Artifact removal, ASIC, Block matching, Complexity algorithms, Computer graphics, Deformable models, Digital video temporal resolution enhancement, Digital videos, DSP, Existing method, FPGA, FPGAs and ASICs, Frame rate up-conversion, Hexagonal search block matching, Image enhancement, Image matching, Image resolution, Interpolated frames, Large screen, Large screen LCD, Liquid crystal displays, Motion compensation, Motion estimation, Motion vector, Multimedia systems, Multi-stage, Multistage motion vector post-processing algorithm, Occlusion adaptive overlapped block motion compensation, Occlusion artifact removal, Occlusion region, Overlapped block motion compensations, Plasma TV, Plasma TV sets, Postprocessing algorithms, Processing power, Projection systems, Reduced loss, Signal processing, Spatial domains, Spatial resolution, Temporal resolution, Video frame rate up-conversion, Video signal processing, Video frame, Visual qualities
Kaynak
2009 IEEE 17th Signal Processing and Communications Applications Conference
WoS Q Değeri
N/A
Scopus Q Değeri
N/A
Cilt
Sayı
Künye
Çizmeci, B. & Ateş, H. F. (2009). Occlusion artifact removal in video frame rate up-conversion. Paper presented at the 2009 IEEE 17th Signal Processing and Communications Applications Conference, 277-280. doi:10.1109/SIU.2009.5136386