Color and Edge Directive Descriptor Feature Extraction Technique for Content Based Image Retrieval System

The development of multimedia technology in Content Based Image Retrieval (CBIR) System is one of the prominent area to retrieve the images from a large collection of database. It is practically observed that any one algorithm is not efficient in extracting all different types of natural images. Hence a thorough analysis of certain color, texture and edge extraction techniques are carried out to identify an efficient CBIR technique which suits for a particular type of images. The Extraction of an image includes feature description, index generation and feature detection. The low-level feature extraction technique is proposed in this paper are tested on Corel database, which contains 10 categories of natural image dataset, each category has 100 images, totally the database has 1000 images. The feature vectors of the query image are compared with feature vectors of the database images to obtain matching images. This paper proposes Color and Edge Directivity Descriptor (CEDD) feature extraction technique which extract the matching image based on the similarity of color and edge of an image in the database. The Image Retrieval Precision (IRP) and Recall value of the proposed technique is calculated and compared with that of the existing techniques. The algorithms used in this paper are Discrete Cosine Transform (DCT), Discrete Wavelet Transform (DWT) and Fuzzy Linking algorithm. The proposed technique results in the improvement of the average Precision and Recall value. Also CEDD is effective and efficient for image indexing and image retrieval. CBIR systems are based on color, texture, shape and images from the image data base for the given query edge information are available in the literature. The image, by comparing the feature of the query image. general applications of CBIR are consumer digital photo Edges in an image constitute an important feature to album, digital museum, MPEG-7 Content descriptor, represent their content of the image. Human eyes are general image collection for licensing and natural very sensitive to edge features for image perception. collections. Content Based Image Retrieval (CBIR) shows Histogram is used to represent an important edge feature. an important role in data retrieval techniques. In CBIR An edge histogram in an image space represents the system, the images are retrieved from large database directionality of the brightness changes and its based on the similarity in the characteristics of input frequency. The normative MPEG-7 edge histogram is query image rather than going into the details of designed to contain the 80 bins of local edge distribution. description and tags, annotations of any particular image. These 80 bin histograms are the standardized semantics It is set automatically extracting characteristic image are for MPEG-7 Edge Histogram Descriptor. The local compared automatically and examine the role of query histogram bins are not sufficient to represent global image, such as color, shape, texture, measures of similarity features of the edge distribution. The global edge and finally get output best matching image and its relation distribution is needed to improve the retrieval to the information. This paper describes an image retrieval performance of an image and images in the database. technique based on multi wavelet texture features. Texture Relevant images are retrieved according to minimum is an important feature of natural images. Features of an distance or maximum similarity measure calculated meaning of the image. CBIR system retrieves the relevant