Table 4 Imperceptibility analysis for various test images with different gain factors.

From: Convolutional neural network and wavelet composite against geometric attacks a watermarking approach

Host image

Gain factor

MSE

PSNR (dB)

SSIM

NC

Male

0.01

0.1128

57.6096

0.9970

1.0000

0.02

0.4455

51.6424

0.9988

0.9999

0.05

2.7517

43.7347

0.9934

0.9994

0.1

10.9883

37.7215

0.9780

0.9977

0.2

43.9348

31.7027

0.9364

0.9908

Cameraman

0.01

0.1160

57.4850

0.9998

1

0.02

0.4455

51.6424

0.9993

0.9999

0.05

2.7517

43.7347

0.9963

0.9997

0.1

10.9883

37.7215

0.9876

0.9986

0.2

43.9348

31.7027

0.9640

0.9945

Girl

0.01

0.1160

57.4850

0.9998

1

0.02

0.4455

51.6424

0.9991

1

0.05

2.7517

43.7347

0.9955

0.999

0.1

10.9883

37.7215

0.9858

0.9991

0.2

43.9348

31.7027

0.9607

0.9963

Chemical plant

0.01

0.1160

57.4850

0.9998

1

0.02

0.4455

51.6424

0.9940

0.999

0.05

2.7517

43.7347

0.9965

0.997

0.1

10.9883

37.7215

0.9875

0.9987

0.2

43.9348

31.7027

0.9607

0.9947

Back to article page