Study of Content Based Image Retrieval Techniques

Abstract

Content Based Image Retrieval (CBIR) is one of the most popular and interesting research areas because of the proliferation of video and image data in digital form. Fast and accurate retrieval of image from large databases is an important problem that needs to be addressed. Basically, CBIR is on developing technologies to bridge the semantic gap that currently prevents wide-deployment of image content-based search engines. Image search engines currently in use such as Google Images and Yahoo! Image search, are based on textual annotation of images. Here, images are manually annotated with keywords, which depends totally on the person's perception, and then retrieved using text-based search methods. This method is both time-consuming and prone to errors. Hence, such search engines result in retrieving many non-relevant images. To overcome such drawbacks of text based image retrieval, CBIR is introduced where the visual content of an image, such as color, texture and shape, is extracted automatically; the retrieval of images is totally dependent on these features. In this paper, we will try to study the basic block diagram of CBIR and the different techniques currently being used, such as color histogram and edge histogram.