Saliency Object Detection Based on Domain Transform and Contour Detection

In order to solve the problem of edge blur and brightness non-uniformity of salient object in theprocess of multi-level saliency maps integration, this paper proposes a saliency detection method based ondomain transform and contour detection. Firstly, we obtain initial saliency map by integrate three saliencymaps using DRFI method. Then, the salient object contour of image are computed by convolutional neuralnetwork. Finally, we use domain transform to blend the initial saliency map and the salient object contour.Under the constraints of the salient object contour, we can filter out the errors in the initial saliency map andcompensate the missed region. The experimental results on three public datasets demonstrates that ourmethod can produce a pixel-wised clearly saliency map with brightness uniformity and outperform othereight state-of-the-art saliency detection methods in terms of precision-recall curves, F-measure, ROC andAUC.