Saliency detection with relative location measure in light field image
Chao Li, Bin Zhan, Shuo Zhang, Hao Sheng · 2017
Saliency detection becomes a crucial requirement for numerous computer vison application. Conventional manifold ranking models have been widely used for saliency detection because it can measure similarity efficiently between the regions, but most of them make use of color and texture information and location information of objects didn't be well exploited, so it cannot work properly when objects is low-contrast. Based on manifold ranking model, this paper proposes a relative location measure on object in light field image. Then an adaptive foreground selection and background selection method using relative location is adopted to get preliminary saliency maps. Finally, an optimization model is presented to combine the saliency maps to get saliency detection results. Quantitative evaluations are carried out on public saliency detection dataset. The results demonstrate that the proposed method can outperform the other state-of-the-art method by using relative location measure.