Visual saliency detection via hypergraph based re-ranking using background priors
Kyung-Wook Park, Dong‐Ho Lee · 2015
Salient object detection is a powerful tool to be applied to many computer vision tasks such as object recognition, image segmentation and scene understanding. We formulate salient object detection as a hypergraph based ranking problem which ranks the similarity of the image elements with foreground or background cues. In addition, we introduce an adaptive background prior to prevent suppression of salient objects touching image boundary. We can improve the results of saliency detection by using the adaptive background priors. Experimental results on three public image dataset demonstrate that our method performs better than the state-of-the-art saliency detection methods.