Title :
A low complexity block matching algorithm for fast motion estimation in High Efficiency Video Coding
Author :
K C Ravi Chandra Varma; Venkata Phani Kumar M;Sudipta Mahapatra
Author_Institution :
Electronics and Electrical Communication Engineering, IIT Kharagpur, West Bengal, India
Abstract :
This paper proposes a search mechanism, namely accelerated internal stop search (AISS), to reduce the complexity of the block matching process used for motion estimation in High Efficiency Video Coding (HEVC). The proposed mechanism reduces the number of operations required to find the best matching reference block. Moreover, it uses the correlation with the previous reference frames to further reduce the complexity of motion estimation for the current reference frame. Results demonstrate that the proposed mechanism is able to reduce the complexity of motion estimation in the fast search mode of the HM-14.0 reference software for HEVC, i.e., the TZ Search motion estimation. The AISS mechanism outperforms the dynamic internal stop search algorithm (DISS) proposed by Ismail et al. in the year 2012 in terms of motion estimation time saving. Moreover, when AISS is integrated with the dynamic model based fast motion estimation proposed by Ismail et al. [1], the complexity of motion estimation is further reduced.
Keywords :
"Complexity theory","Motion estimation","Mathematical model","Heuristic algorithms","Bit rate","Video recording","Quality assessment"
Conference_Titel :
Computer Vision, Pattern Recognition, Image Processing and Graphics (NCVPRIPG), 2015 Fifth National Conference on
DOI :
10.1109/NCVPRIPG.2015.7489951