Reduction of memory accesses in motion estimation by block-data reuse

Moshnyaga, Vasily G.

doi:10.1109/icassp.2002.5745312

Search citation statements

Order By: Relevance

Paper Sections

Select...

Citation Types

Supporting

Mentioning

Contrasting

Year Published

2006

2011

Publication Types

Select...

Article2

Other2

Relationship

Self Cite0

Independent4

Authors

Journals

Cited by 4 publications

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

Low Power Complexity-Reduced ME and Interpolation Algorithms for H.264/AVC

Hyun¹,

Sunwoo

2008

J Sign Process Syst Sign Image Video Technol

View full text Add to dashboard Cite

This paper presents efficient memory reuse and the modified normalized partial distortion search algorithms for motion estimation of H.264/Advanced Video Coding. In addition, this paper presents multiplication-free sub-pixel interpolation and adaptive sub-pixel interpolation algorithms. The proposed memory reuse algorithm utilizes the position similarity of the predicted motion vectors among neighboring sub-macro blocks and the modified normalized partial distortion search algorithm uses an adaptive search range and the dynamic threshold methods. It can reduce a large number of memory access and can save power consumption by sharing the search range of the current block. In addition, simplifying weights of a sub-pixel interpolation filter can eliminate multiplications and the adaptive sub-pixel interpolation can reduce the number of sub-pixel search points, which can reduce memory access and hardware complexity. Therefore, the proposed algorithms are quite suitable for low power video compression.

show abstract