| | |
Summary: UNSUPERVISED SIMULTANEOUS REGISTRATION AND EXPOSURE CORRECTION
Pedro M. Q. Aguiar
Institute for Systems and Robotics / IST
Lisboa, Portugal
E-mail: aguiar@isr.ist.utl.pt
ABSTRACT
Early approaches to building mosaics by composing photo-
graphic images, assume the input images have similar expo-
sures. Since this is unlikely to happen in practice, it became
common to compensate for different exposures in the blend-
ing step, after the images have been registered, or aligned [1].
However, registration methods usually assume brightness con-
stancy and fail to align images with different exposures. Re-
cent approaches to this problem lead to computationally com-
plex solutions that require either robust statistics or nonlin-
ear optimization. In this paper we propose a computation-
ally simple method to jointly estimate the registration param-
eters and the parameters describing the exposure correction,
directly from the image intensity values. We obtain closed-
form solutions for the estimates of the exposure parameters.
|
