Reversible Data Hiding Method Based on Prediction-Error and Pixel-Value Difference for Digital Images

• Main Authors: Su, Chun Te, 蘇峻德
• Other Authors: Liu, Fong Hao
• Format: Others
• Language: zh-TW
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/52628362049346099233

碩士 === 國防大學管理學院 === 資訊管理學系 === 99 === Digital media including texts, images and videos, will affect modern life. With the rapid development of the Internet, digital images in a network environment, have the special properties such as easy accessing, distributing and modifying. To further encourage more people create the art and cultural works, multimedia authentication and security have become very important issues currently. Information hiding is a method to embedding information in the cover-media, the encapsulated data was difficult to be detected in covered information. The information hiding technique is now not only just for the transmission of confidential information, but also for books copyright authentication, digital content protection, the broadcast monitoring, and digital forensics. With the advance of reversible data hiding in recent years, people can restore the original covered image without distortion. Moreover, it produces greater benefits for the Multimedia information hidden application. In this paper, a reversible data hiding based on prediction-error and pixel-value difference was proposed. It detects the differences among image regions and the characteristics of embedded data to select the appropriate range of reversible algorithms and embedded value. The results of the proposed approach, the hiding capacity is 0.24ppb, the PSNR (Peak Signal-to Noise Ratio) higher than 53.81dB, without prejudice to the case of human vision, allowing digital media content security has increased significantly.