A Visual Attention Model for Image Adaptation
He Zhou · Jisuanji fangzhen · 2005
Images have become more and more common in mobile communications. However, a crucial challenge is to provide adaptive displays on limited and heterogeneous screen sizes of mobile devices. To hurdle this problem, we proposed the solution of attention - based image adaptation, which is based on the visual attention modeling. In this paper, an effective attention model is proposed. Instead of treating an image as a whole, we decompose it into a series of Attention Objects ( AOs) and determine the attributions of each AO by the approaches of size estimation, constraint segmentation and region - based saliency extraction. The experiment results demonstrate its effectiveness.