An Analysis of Imperceptibility and Robustness Performance in CRT Image Watermarking based on Color Space Theory

Abstract

Color images have more information and more storage space than gray images. The process of computing color images is also more complex compared to gray images. Nevertheless, human vision is more sensitive to color images compared to gray images, so color image processing is so important. Image watermarking is a method of copyright protection by embedding copyrights in the image. The embedding process can cause changes in the pixel value of the image, so it needs to be analyzed and measured the quality of watermarked images both visually or imperceptibility and watermark durability. In general, color images use the RGB color space model and color space conversion is not done when embedding watermarking. But several studies have suggested converting RGB color space to YCbCr color space for a specific purpose. This research proposes to analyze the robustness and imperceptibility performance in the R, G, B and Y color spaces in the CRT method. CRT is one of the methods of image watermarking in the spatial domain, where this method was previously also used as a cryptographic algorithm. So that the CRT method excels in durability and security watermarks. Based on the results of testing with the CRT method, the color space R has an advantage in the aspect of imperceptibility compared to other color spaces, the difference is very thin with the G and B color spaces, but when compared with the Y color space the difference reaches 3dB based on the PSNR calculation. But for testing the robustness of the color space Y is far superior compared to other color spaces even though extraction without attack, not all watermarks can be extracted perfectly.