Research >> 2D images
Our goal is to develop a registration algorithm capable of
aligning a wide variety of images, including image pairs taken under
substantially-different illumination conditions, image pairs
taken before and after physical changes have occurred in the scene,
image pairs that overlap by extremely small amounts, and image pairs
taken with different modalities. Our Generalized Dual-Bootstrap
algorithm has shown success in all of these problem areas.
It will successfully register two images when it can (a) find a
common location in the two images to start, and (b) there is
sufficient common structure between the images, even when there are
many differences as well.
Faculty: Chuck Stewart
Former Students: Charlene Tsai
Gehua Yang, Charles V. Stewart, Michal Sofka, and Chia-Ling Tsai, "The Generalized Dual-Bootstrap
ICP algorithm with application to registering challenging image pairs"
IEEE Trans. on Pattern Analysis and Machine Intelligence, accepted for publication.
[ pdf ]
C. Stewart, C.-L. Tsai, and B. Roysam. The dual-bootstrap iterative closest point
with application to retinal image registration. IEEE Trans. Med. Imaging.,
22(11):1379– 1394, 2003.
[ pdf ] [ bib ]