Variable scaling factor based invisible image watermarking using hybrid DWT — SVD compression — Decompression technique

Abstract

Watermarking is a technique of embedding hidden information in multimedia data imperceptibly, such as image, video, audio and text data. Generally, an original image is transformed and coded watermark image data is embedded in frequency domain watermarking models. The proposed work explores the computational analysis of variable scaling factor (α) based invisible image watermarking process using the combination of Singular Value Decomposition (SVD) and Discrete Wavelet Transform (DWT) algorithms. The performance analysis of such an invisible image watermarking system is measured using percentage Compression Ratio (%CR), Peak Signal to Noise Ratio (PSNR), with respect to the embedded and extracted images correspondingly. The algorithms thus developed may find applications in copyright identification, finger printing, transaction tracking, content authentication, device control to name a few.