CS766 Project: Image Cropping Study

Team Member: Lin Liu (lliu295), Liang Zhang (lzhang432)

Project Proposal:
Project Proposal, Created at Feb 13, 2017

High resolution digital images capcatured by modern devices usually contain millions of pixels. To reduce the difficulties of image transmission casued by large size of high resolution images, down sampling or compression are usually applied. As a result, the image quality at the receiver side is inevitable degradated. In real life images, however, the visual importance of different regions varies. In high resolution images, unnecessary background and unrelated scenery are usually included as well. These relatively less important pixels not only take transmission bandwith but also harm the visual effectitiveness of the important part of pixels. In order to slove the problem, image cropping/resizing moethods are researched.

Sample Image 1

Sample Image 2

Sample Code:
Sample Code, Coming Soon
Related Publications:
Automatic Image Cropping : A Computional Complexity Study, IEEE Conference on Computer Vison and Pattern Recognition (CVPR), 2016
Last Changed: Feb 14, 2017