HSV model based data hiding in video for watermark applications

Abstract

This manuscript proposes a new data hiding approach that is used in watermark applications in video by transforming the RGB model to HSV model. This method initially estimates the number of frames needed to embed the data (watermark). Then two sets of RGB (Red, Green, Blue) coefficients (R1, G1, B1), (R2, G2, B2) are converted to HSV (Hue, saturation, values) Coefficients (H1, S1, V1) and (H2, S2, V2). The ‘Value’ Coefficients V1 and V2 are used to embed the watermark, since there exists a strong correlation between the adjacent ‘Value’ Coefficients. The same process is repeated on adjacent HSV coefficients till the watermark is fully embedded. After embedding the data HSV coefficients are again converted back to RGB coefficients. During the extraction phase, the data is extracted by transforming the RGB coefficient to HSV coefficients. One bit of information can be extracted from two adjacent HSV coefficients. Experimental outcomes show that the proposed watermarking approach is efficiently against attacks, viz noise, filtering, etc. Also, the proposed method performs better than traditional watermarking methods with the help of embedding rate (bpp), Structural similarity index measurement (SSIM), Visual quality (PSNR), Normalized cross-correlation (NC).