A robust and high-efficiency blind watermarking method for color images in the spatial domain

To respond the digital infringement quickly and effectively in the fifth-generation (5G) new environment, a novel spatial-domain watermarking method combining discrete Tchebichef transform (DTT) is proposed in this paper. Based on the energy concentration property of DTT, its direct current (DC) component is calculated directly by image pixels, then is quantified to embed the color watermark in the spatial domain by using the variable quantization steps, and the embedded watermark can be extracted from the DC component in the spatial domain. The contributions and novelty of this paper are summarized as follows: 1) the mechanism combining DTT and spatial domain is analyzed and derived to quickly implement watermark embedding and blind extraction; 2) different quantization steps are used in different channels to improve the watermark imperceptibility; 3) the overflow detection is set up to ensure the accurate and complete extraction of the watermark. The objective performance of the proposed method is shown as follows: 1) the average of peak signal-to-noise ratio (PSNR) is greater than 41 dB; 2) the average of structural similarity index metric (SSIM) is more than 0.97; 3) all normalized cross-correlation (NC) values without attacks remain 1; 4) the average of NC after various attacks is above 0.96; 5) the total execution time is within 0.5 seconds; 6) the maximum embedding payload keeps 0.0625bpp; 7) the key space reaches to 2206. The obtained simulation results manifest that the proposed method provides strong robustness, high real-time performance, and reliable security around the premise of the better imperceptibility and satisfactory embedding payload, which is suitable for fast and efficient copyright protection occasions in 5G environment.