Improved Graph-based Visual Saliency High Dimensional Image Retrieval System

Salient features are the most important part of high dynamic range (HDR) images. High dynamic range is the technique which has been widely used around the world. People are intended to get brighter and clear images. In this paper, the new method of Graph-Based Visual Saliency (GBVS) has been used to extract the images promptly and accurately. Along with it, the comparison has been done with the help of different statistical parameters. High dynamic range images have converted into the Joint Photographic Experts Group (JPEG). The main objectives of this research is to check the performance of improved Graph-Based Visual Saliency using various technical and engineering parameters such as Mean square error (MSE), peak signal-to-noise ratio (PSNR), Kullback Leibler divergence (KLD), The number of changing pixel rate (NPCR) and unified averaged changed intensity (UACI) etc. These parameters have been used to show the exact result of the experiment. This research also shows the salient features of the HDR images in an accurate manner.