SURVEY ON SOCIAL IMAGE RETRIEVING TECHNIQUES

Abstract

In the era of digital communication, there is tremendous increase in use of digital images on social media sites for sharing or interpreting the information. It makes the task of retrieving images from millions of stored images is very complex and challenging. Basically, image retrieval can be done using methods like Content based Image Retrieval (CBIR) and Tag/keyword based Image Retrieval (TBIR). CBIR is common method of image retrieval depends on visual content of an image. In TBIR method, images are retrieved based on the tag information associated with it. Most of the time, the tags associated with the images aren’t adequate and are very context specific and hence poses limitation with the TBIR method. There are some good techniques proposed by various researchers for efficient retrieval of images. The focus here is to understand the techniques and problems associated with the social image understanding viz. tag/keyword assignment, tag/keyword refinement and retrieval of images.