Highly efficient predictive zonal algorithms for fast block-matching motion estimation

Tourapis, A.M.; Au, Oscar C.; Liou, M.L.

doi:10.1109/tcsvt.2002.804894

Cited by 158 publications

(89 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In popular video codecs, methods based on "predictors" are used that similarly exploit the principle of a search started around the MVs of neighboring blocks, e.g., see Ref. 20. Although, efficient search schemes have drastically reduced computational and memory bandwidth demands, from a memory perspective there is one remaining issue: large velocity ranges.…”

Section: Methodsmentioning

confidence: 99%

A system approach to high‐quality picture‐rate conversion

Bartels¹,

Cordes²,

Riemens³

et al. 2010

J Soc Info Display

View full text Add to dashboard Cite

-Various techniques have been implemented to improve the motion portrayal of flat-panel displays, of which the widespread introduction of motion-compensated picture-rate conversion systems is an essential part. However, a careful design of such systems is critical as they have the potential to introduce perceptually annoying artifacts. In this paper, an overview of the required "ingredients" of high-quality motion-compensated picture-rate conversion is presented, as implemented in state-ofthe-art systems-on-chip. The visual impact of individual components on picture quality is illustrated.Keywords -Motion blur, motion judder, motion compensation, picture-rate conversion, frame-rate conversion. DOI # 10.1889/JSID18.11.922 IntroductionMotion portrayal is an important aspect of picture quality. Motion blur affects the "dynamic" resolution of a display and in this respect, flat-panel-display technologies such as LCD technology have long remained inferior to CRT displays. 1 The two characteristics that caused this were (1) lowresponse times and (2) the picture sample-and-hold effect. Over time, the slow response has been addressed by panel makers through material innovations, while the remaining effect has been further reduced by display processing such as overdrive. 2 The negative consequences of the picture sampleand-hold have been targeted by increasing the display picture rate from 50/60 Hz to 100/120 Hz, and even to 200/240 Hz. However, since video content is mostly captured at relatively low picture rates, such as 24 Hz for film or 50/60 Hz for video cameras, high-quality picture-rate conversion (PRC) is now essential to profit from fast-responding panels.This poses a significant challenge, as the eye-tracking of the human viewer requires PRC algorithm designers to apply motion estimation (ME) and motion compensation (MC) techniques. 3,4 Simpler PRC methods using repetition or linear interpolation of images cause motion blur and/or judder which render the advances in display technology useless. Motion-compensated PRC (MC-PRC) can enable perfect motion portrayal, but, to this day, there is no design without occasional failure. Limiting the occurrence frequency of failures and avoiding perceptually annoying artifacts is challenging; particularly now, recent high-quality displays make even minor artifacts visible.In this paper, we shall discuss the key requirements and lessons learned in the design of state-of-the-art MC-PRC systems, such as those discussed in Ref. 6.In Sec. 2, we show the components of a high-quality PRC system. In Secs. 3-7, these will be illustrated individually by showing their visual impact from an end-user perspective. We draw our conclusions in Sec. 8. Components of MC-PRCMC-PRC is based on the estimation of the displacement path or "motion trajectory" of all pixels between consecutive frames, i.e., (dense) motion estimation. Early MC-PRC algorithms relied on relatively simple interpolation technologies to profit from the ME. However, since the introduction of MC-PRC in CRT-TV ...

show abstract

Section: Methodsmentioning

confidence: 99%

A system approach to high‐quality picture‐rate conversion

Bartels¹,

Cordes²,

Riemens³

et al. 2010

J Soc Info Display

View full text Add to dashboard Cite

show abstract

“…3(c)] and its later modifications: modified circular zonal search (MCZS) [27] and advanced diamond zonal search (ADZS) [28,29] exploit spatiotemporal correlations among MVs. To deter− mine the search centre, the set of predictors is used includ− ing MVs of left, top and top right neighbours and MV of the co−located MB from the previous frame.…”

Section: Characterization Of Algorithmsmentioning

confidence: 99%

Block-based motion estimation algorithms — a survey

Jakubowski

Pastuszak

2013

Opto-Electronics Review

View full text Add to dashboard Cite

In the multi-view video coding, both temporal and inter-view redundancies can be exploited by using standard block-based motion estimation (BBME) technique. In this paper, an extensive review of BBME algorithms proposed within the last three decades is presented. Algorithms are divided into five categories: 1) based on the search position number reduction; 2) multiresolution; 3) based on the simplification of matching criterion; 4) fast full search; 5) computation-aware. Algorithms are compared in terms of their efficiency and computational complexity.

show abstract

“…7(a) shows the M2SE and SI values of some existing block matching methods, i.e., Full Search, Diamond Search [28], Hexagonal Search [29], and predictive motion vector field adaptive search technique (PMV-FAST) [30] (with early termination disabled). Note that the M2SE metric favors "true" VFs, e.g., PMVFAST outperforms a Full Search in M2SE value, whereas this is never the case for the MSE metric.…”

Section: A M2se Vs Simentioning

confidence: 99%

Smoothness Constraints in Recursive Search Motion Estimation for Picture Rate Conversion

Bartels

Haan

2010

IEEE Trans. Circuits Syst. Video Technol.

View full text Add to dashboard Cite

Abstract-Many motion compensation algorithms are based on block matching. The quality of the block correlation depends on the validity of the brightness constancy assumption and the assumption of fixed translational motion within a block. These assumptions are invalid in areas with texture changes, noise, lighting changes, and rapid deformations. Smoothness priors should enforce stable estimates in these regions by propagating neighboring estimates, while preserving hard object boundaries (piecewise smoothness). Most motion estimation algorithms that successfully implement these constraints are computationally complex. In this paper, we show an intuitive and computationally efficient way to implement them within the framework of (realtime) recursive search, targeting consumer-market embedded devices with limited resources.

show abstract

Highly efficient predictive zonal algorithms for fast block-matching motion estimation

Cited by 158 publications

References 14 publications

A system approach to high‐quality picture‐rate conversion

A system approach to high‐quality picture‐rate conversion

Block-based motion estimation algorithms — a survey

Smoothness Constraints in Recursive Search Motion Estimation for Picture Rate Conversion

Contact Info

Product

Resources

About