Hdr Image Saliency Estimation by Convex Optimization

In this paper, we propose a convex optimization-based method for the visual saliency prediction of high dynamic range (HDR) images, which allows straightforward reuse of any existing saliency estimation methods for typical images with low dynamic range (LDR). First, the proposed method decomposes a given HDR image into multiple LDR images with different levels of intensity using a tone-mapping-based synthesis of imaginary multiexposure images. For each decomposed image, a standard saliency estimation method is then applied for typical LDR images. Finally, the saliency map of each decomposed image is integrated into a single map by solving convex optimization problems. The proposed method is applied to actual HDR images and its effectiveness is demonstrated.