Information Hiding Based on Green Noise Mask in Halftone Images

• Main Authors: Chia-Fan Lin, 林珈帆
• Other Authors: Tao-I Hsu
• Format: Others
• Language: zh-TW
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/20752303457181068752

碩士 === 世新大學 === 平面傳播科技學系 === 92 === 　This study aims to embed watermark based on the dot gain properties of outputted halftone images. Digital halftoning reduces the gray levels by quantizing contone images into binary images affecting the robustness and security of watermark embedding. In addition, the capacity of information hiding in halftone image is limited. 　The blue noise mask (BNM) is the state of art of digital halftone technique, which employs the characteristics of higher cut-off spectrum frequency to meet the perceptional quality of human visual system. However, images halftoned by BNM causes larger dot gain because of its frequency modulation distribution, which may result in outputted image darker and less pleasant in visual perception. Moreover, the green noise mask (GNM) inherited not only from BNM’s spectrum characteristic but also from dot gain aspect of the printer model by varying the dot size where provide the possibility of positions for watermark to be embedded in images halftoned by GNM. 　The study attempts to embed watermark by toggling the value of pixels in halftone image. The embedding positions are decided depend on the degrees of dot gain compensation. The performance of the method is demonstrated by a number of experiments that can keep embedded images’ visual qualities and watermarks’ security, robustness and imperceptibility.