Efficient DSP implementation of fractional-pixel interpolation for AVS

Authors:
Zhigang Yang;Shuhong Jiao;Lutao Liu
Affiliations:
College of Information and Communication, Harbin Engineering University, Harbin, China;College of Information and Communication, Harbin Engineering University, Harbin, China;College of Information and Communication, Harbin Engineering University, Harbin, China
Venue:
PCM'12 Proceedings of the 13th Pacific-Rim conference on Advances in Multimedia Information Processing
Year:
2012

Citing 6
Cited 0

Efficient Implementation of Interpolation for AVS

CISP '08 Proceedings of the 2008 Congress on Image and Signal Processing, Vol. 3 - Volume 03
Adaptive interpolation filter for H.264/AVC

IEEE Transactions on Circuits and Systems for Video Technology
Subpixel interpolation architecture for multistandard video motion estimation

IEEE Transactions on Circuits and Systems for Video Technology
High throughput and low memory access sub-pixel interpolation architecture for H.264/AVC HDTV decoder

IEEE Transactions on Consumer Electronics
Adaptive interpolation filters and high-resolution displacements for video coding

IEEE Transactions on Circuits and Systems for Video Technology
Interpolation Free Subpixel Accuracy Motion Estimation

IEEE Transactions on Circuits and Systems for Video Technology

Quantified Score

Hi-index	0.00

Visualization

Abstract

Fractional-pixel motion compensation can greatly improve the compressing efficiency in video coding, while quarter-pixel interpolation also leads to a significant increment in computational complexity. This paper presents some techniques for efficient implementation of quarter-pixel interpolation in AVS-P2 on a fix-point digital signal processor (DSP). Firstly, the whole interpolation process is divided into five sub-processes from the DSP-oriented viewpoint. Then highly parallel software pipelines are designed for each sub-process with elaborately balancing the resources on each side of the CPU data path. A task-level optimization strategy is also applied to arrange the software pipelines. Finally, the simulated results demonstrate that the execution time of interpolation can be greatly reduced by using this specific design.