Content based Image Retrieval using Color, Texture, Shape and Active Re-Ranking Method

Background/Objectives: The most effective and efficient tool for managing large image database is Image Retrieval. Content Based Image Retrieval is strategy for recovering images based on the content of the image. Color, Shape, Texture are said to be Content of an image. Methods/Statistical Analysis: The necessity for Content Based Image Retrieval has been increasing over a decade in different domains such as Data mining, Medical Imaging, Education, Crime prevention etc. It is still an research area where research is going on how to recover the images based on their content. The existing system has a flaw that is image retrieving is based on the keywords presented along with the image. Findings: The images retrieved based on the keywords are inaccurate and it also consumes more time. In this technique we present query image as an input and we get related images as the output which match the content of the query image. Here, the findings are not only based on the color, texture and shape but also trace the underlying points of the image. At first the pictures are recovered based on the color, then took after by texture and finally tracing the underlying graphical structure. Application/Improvements: The proposed system is efficient in recovering images based on the content of the image presented as query. Retrieving images based on underlying graphical structure will help in removal of many irrelevant images and makes the system efficient.