Robust Watermarking Algorithm Based on Structured Forests Edge and SIFT

Liu Ying; Yang Xing; Zhu Tingge

doi:10.3788/LOP202158.0615006

Journals >Laser & Optoelectronics Progress >Volume 58 >Issue 6 >Page 615006 > Article

Laser & Optoelectronics Progress
Vol. 58, Issue 6, 615006 (2021)

Robust Watermarking Algorithm Based on Structured Forests Edge and SIFT

Liu Ying^1、2, Yang Xing^1、*, and Zhu Tingge^1、2

Author Affiliations

¹School of Communication and Information Engineering, Xi''an University of Posts & Telecommunications, Xi''an, Shaanxi 710121, China

²Key Laboratory of Electronic Information Application Technology for Scene Investigation, Ministry of Public Security, Xi''an, Shaanxi 710121, China

show less

DOI: 10.3788/LOP202158.0615006 Cite this Article Set citation alerts

Liu Ying, Yang Xing, Zhu Tingge. Robust Watermarking Algorithm Based on Structured Forests Edge and SIFT[J]. Laser & Optoelectronics Progress, 2021, 58(6): 615006 Copy Citation Text

show less

Fig. 1. Flow chart of SIFT operator extracting 1024 stable points

Download full size

Fig. 2. Flow chart of watermark embedding

Download full size

Fig. 3. Flow chart of watermark extraction

Download full size

Fig. 4. Cover images.(a) Lena; (b) Peppers; (c) Plane; (d) Mandrill

Download full size

Fig. 5. Watermark images.(a) Lena_W; (b) Peppers_W; (c) Plane_W; (d) Mandrill_W

Download full size

Fig. 6. NC values of the extracted watermark against different attacks

Download full size

Parameter	Lena	Peppers	Plane	Mandrill
PSNR /dB	55.39	56.18	59.43	58.33
NC	1	1	1	1
SSIM	1	1	1	1

Table 1. Test results of watermarked images without attacks

Algorithm	Lena		Peppers		Plane		Mandrill
Algorithm	PSNR /dB	NC	PSNR /dB	NC	PSNR /dB	NC	PSNR /dB	NC
Proposed algorithm	55.39	1	56.18	1	59.43	1	58.33	1
Algorithm in Ref.[8]	47.96	0.9775	47.34	0.9602	44.52	0.9830	48.34	0.9793
Algorithm in Ref.[22]	52.34	0.9889	53.21	0.9891	50.96	0.9883	54.69	0.9921

Table 2. Comparison results of different cover images without attacks

Attacks	Lena	Peppers	Plane	Mandrill
No attack	1	1	1	1
Salt & pepper noise(d=0.01)	0.9824	0.9867	0.9859	0.9825
Salt & pepper noise(d=0.03)	0.9585	0.9593	0.9591	0.9608
Gaussian noise (m=0, var is 0.005)	0.9893	0.9809	0.9826	0.9972
Gaussian noise (m=0, var is 0.01)	0.9601	0.9700	0.9717	0.9961
Cropping (20 columns)	0.9987	0.9954	0.9972	0.9964
Cropping (30 columns)	0.9979	0.9927	0.9963	0.9945
Cropping (40 columns)	0.9917	0.9936	0.9949	0.9908
JPEG compression (Q=50)	0.9954	0.9982	0.9963	0.9975
JPEG compression (Q=70)	1	0.9991	1	0.9984
Median filter (3×3)	0.9870	0.9889	0.9907	0.9861
Sharpening	0.9927	0.9842	0.9993	0.9878
Histogram equalization	0.9552	0.9982	0.9692	0.9576
Rotation (45°)	0.9829	0.9748	0.9599	0.9767
Average filter (3×3)	0.9629	0.9879	0.9879	0.9824
Contrast adjustment	0.9889	0.9936	0.9916	0.9962
Speckle noise (var is 0.01)	0.9907	0.9926	0.9946	0.9975
Speckle noise (var is 0.05)	0.9963	0.9935	0.9958	0.9963
Motion blur	0.9842	0.9879	0.9757	0.9861
Gaussian filter (3×3)	0.9761	0.9889	0.9898	0.9870

Table 3. NC values of the extracted watermarks against various attacks

Liu Ying, Yang Xing, Zhu Tingge. Robust Watermarking Algorithm Based on Structured Forests Edge and SIFT[J]. Laser & Optoelectronics Progress, 2021, 58(6): 615006

Download Citation

Set citation alerts for the article

Tools

Set citation alerts for the article

Save the article for my favorites

Paper Information