Zero-Watermarking Algorithm for Medical Image Based on VGG19 Deep Convolution Neural Network

Table 1

Experimental results under ripple distortion attack.

	Distortion quantity (%)	150	300	450	600	750
Medical image A	PSNR (dB)	12.7889	11.9036	11.5765	11.2648	11.0596
	NC	0.87595	0.81021	0.81021	0.81106	0.74818
Medical image B	PSNR (dB)	16.6378	12.8092	12.8092	11.7588	11.0855
	NC	0.87595	0.87595	0.87595	0.68929	0.75075
Medical image C	PSNR (dB)	21.8267	17.8745	15.8378	14.6473	13.8828
	NC	0.81363	0.74989	0.81163	0.65842	0.59668
Medical image D	PSNR (dB)	21.3456	17.7586	16.1687	14.8927	14.0789
	NC	0.87509	0.87138	0.81021	0.81392	0.81392
Medical image E	PSNR (dB)	28.1822	25.2023	23.3165	21.9457	20.5296
	NC	0.87823	0.75104	0.68987	0.68872	0.68872