Color Image Retrieval Methods Based on Euler Vector

A compact representation of image features is highly desirable for efficient management of the image database, search, and retrieval. A method of image retrieval using Euler vector is proposed. The method combining gray-coded bit-planes with augmented bit-planes for image retrieval is divided into two ways. The first is to get the united Euler vector as image features, and make use of distance measures to the retrieval of similar color images to query images after Gaussian normalization. The other uses the gray-coded bit-planes based Euler vector and the augmented bit-planes based stacked Euler vector as the image features to compute two corresponding distances between the query image and the object ones, and then assigns a weight to each of the two distances to obtain a new weighted distance with which to retrieve the similar images. Experimental results are reported to demonstrate that the retrieval performance is improved significantly.