Parallel genetic search algorithm for motion estimation

Savio Lai Yin Lam; Ishfaq Ahmad

doi:10.1117/12.230425

19 January 1996 Parallel genetic search algorithm for motion estimation

Savio Lai Yin Lam, Ishfaq Ahmad

Proceedings Volume 2617, Multimedia: Full-Service Impact on Business, Education, and the Home; (1996) https://doi.org/10.1117/12.230425
Event: Photonics East '95, 1995, Philadelphia, PA, United States

Abstract

Motion estimation is perhaps the most computationally intensive aspect of video compression. There are two major approaches to motion estimation: pel recursive and block matching. In the former approach, estimation of motion vectors is done recursively such that the motion compensated prediction error at each pel instant is minimized. In the latter approach, motion estimation is carried out on a block-by-block basis. Comparing with the pel recursive algorithms, the block matching algorithms are more realizable due to their computational simplicity. In this paper, we present a parallel algorithm using genetic search for block-based motion. The first objective of the proposed approach is to remove the need for exhaustive search by making use of genetic algorithms (GA). The second objective is to run this algorithm in parallel so that the computing time is further reduced. The algorithm is implemented using the Express library for a network of workstations and NX for the Intel Paragon.

Citation Download Citation

Savio Lai Yin Lam and Ishfaq Ahmad "Parallel genetic search algorithm for motion estimation", Proc. SPIE 2617, Multimedia: Full-Service Impact on Business, Education, and the Home, (19 January 1996); https://doi.org/10.1117/12.230425

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