A novel image authentication method based on singular value decomposition

Abstract

Fragile watermarking methods have been widely used for image authentication. In order to increase the fragility of digital watermarking techniques, cryptologic hash functions, parity bits and secret sharing methods are generally used. In this article, a new image authentication method based on singular value decomposition (SVD) with a new block-based data hiding technique is proposed. The block size can be determined by the user in the proposed method. In this work, 2 x 2, 4 x 4, 8 x 8 or M x M size of blocks are used. Features are extracted from each block using SVD. Extracted features are quantizated by block size. In the stage of watermark generation, watermarks are generated by using secure pseudo random generators. The generated watermark is encrypted by extracted features and encrypted watermark is embedded into cover image using the proposed data hiding function. This method consists of preprocessing, feature extraction, quantization, watermark generation, watermark encryptiom, watermark embedding, watermark extraction, watermark decryption and tamper detection. The visual quality of the suggested method is measured by peak signal to noise ratio (PSNR) and tamper detection capability is measured by bit error rate (BER). Experimental results have demonstrated that the proposed method is successful.