Image retargeting is a technique to adapt an original image to diverse screen sizes and aspect ratios on computing devices. This paper focuses on adapting large images for a small display. Some existing methods, such as scaling, cropping, and fisheye warping, are often flawed because they may lose the necessary details of the image, scarify content entirety, or fail to address the content continuity of images with multiple focuses. To address this issue, this paper presents a new fisheye-based approach to retarget images with multiple focuses.With fisheye-based image transformation, this approach can emphasize the focused areas of the image without completely discarding unfocused contents. With a multifocus conflict solution scheme, this approach offers a continuous content transition among multiple focused areas. With image retargeting algorithms implemented with this approach, we conducted experiments to study user preferences of retargeted images under different algorithms. The results of the experiments show that our approach is appropriate for image retargeting.
All Science Journal Classification (ASJC) codes
- Computer Vision and Pattern Recognition
- Computer Graphics and Computer-Aided Design