Abstract: This paper proposes a new robust image watermarking algorithm based on independent component analysis (ICA). The algorithm performs binary extraction on the original image to obtain four sub-images, and then performs independent component analysis on these four sub-images to obtain image features. Among the image features with the largest amount of information, its DCT coefficient is selected to embed the watermark. When selecting the embedding domain, the human visual system (HVS) is considered to improve the robustness of the algorithm while ensuring invisibility. Experimental simulation shows that the proposed algorithm can realize blind extraction and has good robustness to common image processing and watermark attacks, such as filtering, scaling, JPEG compression, noise addition, histogram equalization, etc. Keywords: robust watermark, ICA, HVS
You Might Like
Recommended ContentMore
Open source project More
Popular Components
Searched by Users
Just Take a LookMore
Trending Downloads
Trending ArticlesMore