Abstract
Image retargeting has drawn much attention in image and vision research in recent years. Image similarity measure is very important for image retargeting applications. In this paper, we propose a novel similarity metric for image retargeting. We nd some optimal vertical and horizontal seams in the original image and the retargeted image, respectively. Then two images are divided into several regions by virtue of the seams, respectively. The region matching is performed between the two images with SIFT ow and the image similarity is calculated according to the results of region matching. Real data have been used to test the proposed metric and very good results have been obtained, validating it.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
