Mapping-singularities-based motion estimation

Igor V. Ternovskiy; Tomasz P. Jannson

doi:10.1117/12.294525

12 December 1997 Mapping-singularities-based motion estimation

Proceedings Volume 3173, Ultrahigh- and High-Speed Photography and Image-based Motion Measurement; (1997) https://doi.org/10.1117/12.294525
Event: Optical Science, Engineering and Instrumentation '97, 1997, San Diego, CA, United States

Abstract

Existing methods of image-based motion measurement (except point source cases) idealize borders (edges) of objects. Small changes in movement, direction, or projection can introduce errors into movement measurement. We propose a method that applies differential mapping singularities theory (catastrophe theory) in the context of 3-D object projection into a 2-D image plane, and takes advantage of the fact that the edges of an object can be interpreted in mapping singularities (catastrophes). Several theorems show that mapping of an arbitrary smooth surface can create only 14 singularities. Small changes in object position do not change the type of singularity, but simply shift its critical point. A trajectory of moving critical points, extracted from edge and edge vicinity pixels, can be divide into areas that correspond to different singularity types (a so-called phase diagram). Based on a phase diagram, it is possible to select a corresponding singularity relationship graph (SRG). Knowledge of a SRG allows correct prediction of changes during object movement. This approach provides a significant reduction in calculations for motion prediction and permits correction of predicted motion in all sets of predicted frames. In addition, SRG knowledge allows object tracking, even with sudden changes of direction and use of camouflage.

Citation Download Citation

Igor V. Ternovskiy and Tomasz P. Jannson "Mapping-singularities-based motion estimation", Proc. SPIE 3173, Ultrahigh- and High-Speed Photography and Image-based Motion Measurement, (12 December 1997); https://doi.org/10.1117/12.294525

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