Nowadays, Salient region detection is a critical problem in computer vision. It has a huge amount of applications, like object recognition, adaptive compression and segmentation. To detect salient regions many different approaches have been proposed. Visual saliency is which capture our attention is an important quality which makes an object, person, or pixel that relative to its neighbours. Bottom-up and top-down methods are used for visual saliency. Bottom-up salient region detection methods

more »

... hich is classify into uniqueness, compactness and background based.The uniqueness based method is addition of local and global contrast based techniques.The bottom-up salient region detection method is introduced that integrates compactness and local contrast cues. Furthermore, in order to produce a pixel-accurate saliency map which covers the salient objects output is propagated through the diffusion process.