CONTENT BASED IMAGE RETRIVAL USING COMBINATION HISTOGRAM AND MOMENT METHODS

Abstract

In this work, we introduce content based image retrieval CBIR. One of the essential features is colour in an image processing and CBIR so we use the colour histogram and colour moment features in order to compare a inquiry image with the image in the database. The two ways, colour histogram and colour moment have achieved state-of-art results when we applied them to WANG database images. For testing purposes, We have used 100 images (10 images from each class).The mean retrieval of precision of histogram was 74.4 % and the average of colour moment was 72.4% when test every algorithm alone and the result be more efficient when combine them which be 75.1 % by using constant weight and the precision increase to 81.9% when make the weight of combination variable by the user. (Restate with marks)