An efficient image retrieval using central tendency

Abstract

This paper aims to improve the effectiveness of retrieving images on the basis of color content. The paper is organized as follows. Firstly, a mean based technique with reduced feature vector is proposed. Secondly, a new idea to extract the features of an image based on the measure of central tendency is proposed. The proposed mean based technique is compared with existing color averaging techniques for retrieval speed and memory requirement. Finally the proposed mean based technique is compared with the measures of central tendency for retrieval efficiency. The proposed CBIR techniques are tested on generic image database and indexed image database. Experimental results shows that proposed method based on the measures of central tendency gives better performance in terms of higher precision and recall values with less computational complexity.