Digital image Watermark embedding and extraction using oppositional fruit Fly algorithm

Abstract

The proposed method discusses digital image watermark embedding and extraction utilizing discrete wavelet transform and singular value decomposition (DWT-SVD) as well as oppositional fruit fly algorithm. Initially, for partitioning the original input image into several sub-bands DWT-SVD is utilized. After decomposition variance value is calculated for all sub bands and the band with low variance is selected for further process. Then Oppositional Fruit Fly Algorithm (OFA) is used for optimal position selection of selected low variance blocks for embedding the watermark bits. In order to perform the embedding process at first, the conversion of watermark image into its corresponding bits occurs. Then the generated watermark bit is added together with gold code and the combination is utilized for performing watermark embedding. The method used for retrieving the watermarked image from the original image is called watermark extraction. The performance of the proposed image watermarking scheme is analyzed through various metrics such as the Mean square error (MSE), Normalize correlation (NC) and the peak signal to noise ratio (PSNR). The proposed scheme maintains the embedding quality with an average PSNR value of 47.49 dB.