Share Email Print
cover

Proceedings Paper

Toward image registration process: Using different interpolation methods in case of subpixel displacement
Author(s): Deyanira Flores Padilla; Hugo Jimenez-Hernández; Jaqueline Reynosa Canseco
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Interpolating for data sample is required in many image processing methods. For example, in the estimation of displacement which are smaller than one pixel. In the case of displacement calculation in an image sequence using the numerical method, Newton-Rapson is very common to use a linear interpolator due to its simplicity and speed. However this method generates discontinuous functions. Therefore, theoretically it should not be used in combination with Newton-Rapson due to the use of the derivative in the numerical method. This work shows a comparative analysis of different interpolators, along with a comparison between "real world" and image displacements and their relationship. All of this with the purpose of identifying which interpolator offers the most exact approximation in the estimation of displacement calculation.

Paper Details

Date Published: 27 September 2016
PDF: 8 pages
Proc. SPIE 9951, Optical System Alignment, Tolerancing, and Verification X, 99510O (27 September 2016); doi: 10.1117/12.2237961
Show Author Affiliations
Deyanira Flores Padilla, Ctr. de Ingenieria y Desarrollo Industrial (Mexico)
Hugo Jimenez-Hernández, Ctr. de Ingenieria y Desarrollo Industrial (Mexico)
Jaqueline Reynosa Canseco, Ctr. de Ingenieria y Desarrollo Industrial (Mexico)


Published in SPIE Proceedings Vol. 9951:
Optical System Alignment, Tolerancing, and Verification X
José Sasián; Richard N. Youngworth, Editor(s)

© SPIE. Terms of Use
Back to Top