Perceptual Content-Aware Image Reducing Based on Visual Perception

In this paper, we propose a perceptual content-aware image reducing method based on visual perception. First the Regions Of Interest (ROI) are extracted by visual perception. Then we construct the self-adaptive energy function to denote the importance of pixels. At last, the pixel seams which have lowest energy are deleted to protect image content and provide excellent visual effect. Selecting seams may deduce to an optimal problem, in our algorithm, which is transferred to a Shortest-Path problem and solved by dynamic programming. Experiment results show that the proposed algorithm can preserve the important areas and produce excellent vision effects. The reducing results are more desirable compared to all the state-of-the-art algorithms since we pay especially attentions to the biological structure of human beings.