Fast subpel motion estimation for H.264/advanced video coding with an adaptive motion vector accuracy decision

Hoyoung Lee; Bong-Soo Jung; Jooyoung Jung; Byeungwoo Jeon

doi:10.1117/1.OE.51.11.117001

1 November 2012 Fast subpel motion estimation for H.264/advanced video coding with an adaptive motion vector accuracy decision

Hoyoung Lee, Bong-Soo Jung, Jooyoung Jung, Byeungwoo Jeon

Author Affiliations +

Optical Engineering, Vol. 51, Issue 11, 117001 (November 2012). https://doi.org/10.1117/1.OE.51.11.117001

Abstract

The quarter-pel motion vector accuracy supported by H.264/advanced video coding (AVC) in motion estimation (ME) and compensation (MC) provides high compression efficiency. However, it also increases the computational complexity. While various well-known fast integer-pel ME methods are already available, lack of a good, fast subpel ME method results in problems associated with relatively high computational complexity. This paper presents one way of solving the complexity problem of subpel ME by making adaptive motion vector (MV) accuracy decisions in inter-mode selection. The proposed MV accuracy decision is made using inter-mode selection of a macroblock with two decision criteria. Pixels are classified as stationary (and/or homogeneous) or nonstationary (and/or nonhomogeneous). In order to avoid unnecessary interpolation and processing, a proper subpel ME level is chosen among four different combinations, each of which has a different MV accuracy and number of subpel ME iterations based on the classification. Simulation results using an open source x264 software encoder show that without any noticeable degradation (by - 0.07 dB on average), the proposed method reduces total encoding time and subpel ME time, respectively, by 51.78% and by 76.49% on average, as compared to the conventional full-pel pixel search.

Citation Download Citation

Hoyoung Lee, Bong-Soo Jung, Jooyoung Jung, and Byeungwoo Jeon "Fast subpel motion estimation for H.264/advanced video coding with an adaptive motion vector accuracy decision," Optical Engineering 51(11), 117001 (1 November 2012). https://doi.org/10.1117/1.OE.51.11.117001

Published: 1 November 2012

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available