Multiple structure based saliency detection and its application in image retrieval

Saliency Detection is a hot research topic in both biological and computer vision. Salient structures, edges, regions would greatly contribute to high-level seman- tics understanding of people's attention and improve retrieval precision, object detection, edge detection and etc. In this paper, based on the biological principle in visual system, we present a saliency detection system which combines global saliency structure, global saliency in color, local saliency, people/cars detection to give a more comprehensive un- derstanding of people attention system. Besides this, ASSOM is imported to provide feature descriptor of each saliency map for further analysis. Experiments show that the proposed method can provide clearer, accurate saliency detection and improve the preci- sion in content based image retrieval. © 2015.