A Content Based Indexing system For Image Retrieval

Abstract

Abstract newlineOver five to six years the internet has grown immensely and has become the most newlinepopular and cost effective medium for storing and providing electronic media. The newlinelargest amount of multimedia data is stored on the internet creating numerous newlineproblems for users when trying to find information. It is important to remember that newlinemost indexing and searching programs rely solely on processing textual information. newlineBut this technique is not appropriate and effective when we search the desired image newlinefrom the different image database. newlineThe image processing technique can be used to solve such problem as it improves the newlinequality of indexing and searching results. newlineThis idea has raised a new area of research and development widely known as CBIR newlinecontent base image retrieval. In this technique the features of the image are newlineautomatically extracted from the image itself that help in retrieval process. newlineThe measures uses in CBIR are of low level attributes that includes image color newlinecomposition, color histogram and texture. Whereas region based searches focus on newlinebetter measures that indentify the know objects in the region of an image, on the basis newlineof that image can be retrieved from a database. newlineIndexing is the one among the significant factors of image database system newlinethat enhances the accuracy and speed of retrieval process. Many researches in this newlinearea have been conducted for last many years. newlineSome research focuses on improving tree structure up to it selects best image features newlinethat can be used for indexing. These researches have their own pros and cons and newlinecannot be used universally effective in each area. newlineThe aim of our research work to indentify the suitability of image features in various newlinesearching and indexing process. For this purpose in our research work we have used newlinesome of the establishes techniques for extracting image features from the image and newlinecombining these features for effective retrieval process from big image databases. In newlinethis process we have indentify best combinations for different image data set newlinerespect