作者: Jing-Ming Guo
DOI: 10.1016/J.SIGPRO.2007.12.015
关键词: Image processing 、 Digital watermarking 、 Information hiding 、 Encoder 、 Dither 、 Decoding methods 、 Pixel 、 Computer vision 、 Artificial intelligence 、 Halftone 、 Mathematics
摘要: This work proposes a novel technique, named embeddable cells selection (ECS), for embedding flexible amounts of data in ready-made dithered halftone image, while still achieving good quality results. The encoder embeds the secret high-frequency regions with nearly same number black and white pixels low pixel connections. A blind decoding technique without prior knowledge original image but some side information is adopted decoder. inverse halftoning second round are key steps locating embedded bits. Experimental results demonstrate that an objective capacity reasonable complexity obtained. Moreover, correct rate 100% maintained, host can also be reconstructed Furthermore, by recording positions, this method guard against distortions caused tampering, cropping, print-and-scan attacks.