Content-based image copy detection

Abstract

This paper proposes a novel method to detect copied versions of digital images. The proposed copy detection scheme can be used as either an alternative approach or a complementary approach to watermarking. A test image is reduced to an 8 x 8 sub-image by intensity averaging, and the AC coefficients of its discrete cosine transform (DCT) are used to compute distance from those generated from the query image, of which a user wants to find copies. A challenge is that the replicated image may be processed to elude copy detection or enhance image quality. We show ordinal measure of DCT coefficients, which is based on relative ordering of AC magnitude values and using distance metrics between two rank permutations, is robust to various modifications of the original image. The optimal threshold selection scheme using the maximum a posteriori criterion is also described. The efficacy of the proposed method is extensively tested with both cluster-free and cluster-based detection scheme. (C) 2002 Elsevier Science B.V. All rights reserved.