Colour image retrieval based on mean vector and covariance tests

In this paper, statistical test of hypothesis based method is proposed, which retrieves manifold color images such as structured, semi-structured, textured, scaled, and rotated. The proposed method, first, performs preprocessing and then employs the test statistic: test for equality of covariance, and generalized component test. If the query and target images pass the covariance test, then it proceeds to test the mean vectors of the two images; otherwise, the test is dropped. If the query and target images pass both tests, then it is concluded that the images belong to same class. Otherwise, they belong to different class. The performance of the proposed method is evaluated using the image database and its feature database. The images have been collected from various datasets to maintain heterogeneity of the images. The proposed method yields better results.