TY - GEN
T1 - Selective Bayesian estimation for efficient super-resolution
AU - Ivanovski, Zoran A.
AU - Karam, Lina
AU - Abousleman, Glen P.
PY - 2004/12/1
Y1 - 2004/12/1
N2 - In this paper, a new approach to efficient and robust super-resolution is presented. Our method is based on selectively applying a Bayesian MAP estimator to image regions with high spatial activity. The degree of spatial activity is measured using the gradient of the estimated high-resolution image at each iteration. In addition, selective filtering is applied to enhance the visual quality of the estimated high-resolution image. The results obtained via simulation and with real video sequences demonstrate up to a 50% reduction in computational complexity, with improved visual quality, and higher SNR gains for magnification factors of four or more.
AB - In this paper, a new approach to efficient and robust super-resolution is presented. Our method is based on selectively applying a Bayesian MAP estimator to image regions with high spatial activity. The degree of spatial activity is measured using the gradient of the estimated high-resolution image at each iteration. In addition, selective filtering is applied to enhance the visual quality of the estimated high-resolution image. The results obtained via simulation and with real video sequences demonstrate up to a 50% reduction in computational complexity, with improved visual quality, and higher SNR gains for magnification factors of four or more.
UR - http://www.scopus.com/inward/record.url?scp=21544468599&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=21544468599&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:21544468599
SN - 0780386892
SN - 9780780386891
T3 - Proceedings of the Fourth IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2004
SP - 433
EP - 436
BT - Proceedings of the Fourth IEEE International Symposium on Signal Processing and Information Technology
T2 - Fourth IEEE International Symposium on Signal processing and Information Technology, ISSPIT 2004
Y2 - 18 December 2004 through 21 December 2004
ER -