AN OPTIMUM WATERMARKING METHOD IN YCC COLOR SPACE BASED ON HD-SVD IN THE DWT DOMAIN

Abstract

The tremendous development of computer networks and multimedia platforms must have constructed an alarming situation for copyright infringement technology that can secure legal possession of contextual information, given the popularity of digital material. However, watermarking for digital photographs is one method that is being created to safeguard digital photos against illicit alteration. Image processing is a fast-developing discipline with increasing areas of science and engineering. Understanding color is crucial in the creation and configuration of color perception. RGB, CMYK, YCC, and YIQ are among some of the coordinate structures that have been developed for various purposes. We employ the YCC color model for watermarking throughout this thesis. We might switch between RGB, YCC, and YIQ color spaces. Extensively in various multimedia applications, the YCC color space is recommended. This thesis aims to present a practical method for embedding a watermark inside a digital color photograph to recognize who owns it. In conclusion, we introduced non-blind solid watermarking approaches depending on DWT frequency domain transformation within the YCC color space. Singular value decomposition is used to incorporate the original watermark picture. After an R-level discrete wavelet treatment, the singular value matrices are integrated inside the Y plan of the host photographs. Then the color picture is decomposed into four spectrum sub-bands and a time scale using the discrete wavelet transform. The just-noticeable distortion and singular vectors of the low frequency (LL) sub-band of discrete wavelet transform are utilized to determine the embedding element for the Y plan. Correspondingly, the singular value decomposition (SVD) and the Hessenberg decomposition (HD) are tremendous computational decompositions that improve the two fundamental assessment metrics of successful watermarking transparency and durability. Whenever such two decompositions are combined, they accommodate each other 's weaknesses, producing exceptionally successful watermarking. The peak signal-to-noise ratio of the watermarked picture and the retrieved watermark's normalized correlation coefficient were studied. The experimental outcome verified the applicability of our proposed approach.