e propose an algorithm to adaptively segment and fuse images by alternating wavelet package and local cosine transforms each containing best basis selection and thresholding. Within segmented regions fusion is informed by multiple hypotheses testing based on a log-linear factorial model. This fusion identifies homogenous regions from which to select wavelet or local cosine packets, possibly from original images. The successful performance of the fusion algorithm and segmentation is demonstrated on some multispectral thematic mapper imagery.
|Translated title of the contribution||A statistical multiscale approach to image segmentation and fusion|
|Title of host publication||Information Fusion 2005|
|Pages||475 - 482|
|Publication status||Published - 2006|
Bibliographical noteName and Venue of Event: Philadelphia
Conference Proceedings/Title of Journal: Proceedings of the Eighth International Conference on Information Fusion
Conference Organiser: International Society for Information Fusion