Saliency Detection combining Multi-layer Integration algorithm with background prior and energy function

摘要

In this paper, we propose an improved mechanism for saliency detection.Firstly,based on a neoteric background prior selecting four corners of an imageas background,we use color and spatial contrast with each superpixel to obtaina salinecy map(CBP). Inspired by reverse-measurement methods to improve theaccuracy of measurement in Engineering,we employ the Objectness labels asforeground prior based on part of information of CBP to construct amap(OFP).Further,an original energy function is applied to optimize both ofthem respectively and a single-layer saliency map(SLP)is formed by merging theabove twos.Finally,to deal with the scale problem,we obtain our multi-layermap(MLP) by presenting an integration algorithm to take advantage of multiplesaliency maps. Quantitative and qualitative experiments on three datasetsdemonstrate that our method performs favorably against the state-of-the-artalgorithm.