Edge detectionWe propose a edge detection framework that takes as its input a xation (a pointlocation) in the scene and outputs the region containing that xation. The xatedregion is segmented in terms of the area enclosed by the \optimal" closed boundaryaround the xation using the probabilistic boundary edge map of the scene (or image).The probabilistic boundary edge map, which is generated by using all available visualcues, contains the probability of an edge pixel being at an object (or depth) boundary




File list:
edge detection
.............\edge detection_files
.............\....................\kfig1.jpg

.............\....................\kfig10.jpg

.............\....................\kfig2.jpg

.............\....................\kfig3.jpg

.............\....................\kfig4.jpg

.............\....................\kfig4a.jpg

.............\....................\kfig8a.jpg

.............\....................\Thumbs.db
.............\disorder_ext.m
.............\image.fig
.............\img_seg.asv
.............\img_seg.m
.............\img_seg3.asv
.............\input.bmp

.............\karadi.JPG

.............\king.bmp

.............\kn
.............\kni.asv
.............\kni.m
.............\main.m
.............\main1.m
.............\main2.asv
.............\main2.m
.............\melanoma-1.jpg

.............\melanoma-11.jpg

.............\melanoma-2.jpg

.............\melanoma.jpg

.............\seg.m
.............\thirdrev.m
.............\untitled.fig