Image and video retargeting by seam carving

• Main Author: Ong, Wee Jin.
• Other Authors: Deepu Rajan
• Format: Final Year Project
• Language: English
• Published: 2009
• Subjects: DRNTU::Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision
• Online Access: http://hdl.handle.net/10356/20400
• Institution: Nanyang Technological University

Resizing techniques such as scaling and cropping shrinks, enlarges or crop away contents in the image and video during resizing operation in order to meet the geometric constraints. The result is often an image or video with distorted contents or even suffers lost in information. An effective resizing operation should not only consider the geometric constraints but also the contents in the image and video. In this project, a technique called seam carving will be presented. It not only resizes image and video but also preserves the contents. This technique identifies low energy seam which is made up of a path of low energy pixels which are monotonic and connected. By successively removing or adding low energy seams, the image is retargeted with its contents preserved. Other than retargeting images, this technique is able to seamlessly remove objects from images and amplify contents in the image.