Abstract
We present new intelligent logo watermarking based on independent component analysis (ICA) in which a binary logo watermark is embedded in a host image in a wavelet domain. To improve robustness, an image adaptive watermarking algorithm is applied by a stochastic approach based on a noise visibility function (NVF). The algorithm design, evaluation, and experimentation are described. Experimental results show that the logo watermark is perfectly extracted by ICA with excellent invisibility and with robustness against various image and digital processing operators and almost all compression algorithms such as Jpeg, jpeg 2000, SPIHT, EZW, and principal component analysis (PCA) based compression.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
