Digital Image Security Using Multiple Watermarking

Abstract

For the confidentiality, authenticity and integration or for security of digital content data hiding techniques are used. Data hiding is a phenomenon of hiding information. Various methods have been developed for data hiding as of now but each have some limitations and some advantages too. According to the level and kind of application one or more data hiding methods is used. Data hiding can be done in audio, video, text, and image and in other form of information. In this project a data hiding technique based on multiple watermarking in frequency domain is proposed which can be used to authenticate the digital content and its author‟s ownership. Multi-level Discrete Wavelet Transformation (DWT) is used, to firstly transform the host image from spatial domain to frequency domain because of the beneficial properties of better spatial localization, high robustness against attacks and better perceiving of Human Visual System (HVS) of this transformation domain. After transformation multiple watermarks and their copies are embedded into the transformed coefficients of host image. These watermarks are related to each other one is text watermark which is authentication information, second is encoded QR code of this authentication information and third is CRC-16 code of this authentication information. The purpose of embedding multiple copies of these watermarks is to strengthen the robustness of watermark. Ultimate aim of this technique to extract at least one watermark in intact form either image watermark or text watermark. Extracted watermark is then compared with extracted CRC-16 code to prove the authenticity of digital content. Bit correction technique is also imposed on text watermarks to decode most accurate information. Related work for techniques used for data hiding in digital image is also described in this report.