Robust image registration with global intensity transformation

Author

Ruano, Susana ; Gallego, Guillermo ; Yezzi, Anthony ; Cuevas, Carlos ; Garcia, Narciso

Author_Institution

Grupo de Tratamiento de Imagenes (GTI), Univ. Politec. de Madrid (UPM), Madrid, Spain

fYear

2015

Firstpage

1

Lastpage

2

Abstract

This paper presents a registration method for images with global illumination variations. The method is based on a joint iterative optimization (geometric and photometric) of the L₁ norm of the intensity error. Two strategies are compared to directly find the appropriate intensity transformation within each iteration: histogram specification and the solution obtained by analyzing the necessary optimality conditions. Such strategies reduce the search space of the joint optimization to that of the geometric transformation between the images.

Keywords

geometry; image registration; iterative methods; optimisation; L1 norm; geometric transformation; global illumination variations; global intensity transformation; histogram specification; intensity error; joint iterative optimization; photometric; robust image registration; search space; Histograms; Image registration; Iterative methods; Joints; Lighting; Optimization; Sensors; aerial imaging; first variation; histogram specification; illumination transformations; image registration;

fLanguage

English

Publisher

ieee

Conference_Titel

Consumer Electronics (ISCE), 2015 IEEE International Symposium on

Type

conf

DOI

10.1109/ISCE.2015.7177768

Filename

7177768