title of on the implementation of a information hiding design based on saliency map a.basu, t. s....

1
Title of On the Implementation of a Information Hiding Design based on Saliency Map A.Basu, T. S. Das and S. K. Sarkar/ Jadavpur University/ Kolkata/ India/ [email protected], [email protected], [email protected] Swanirbhar Majumder/NERIST (Deemed University)/ Itanagar/ India/ [email protected] In this paper, an adaptive spatial domain image watermarking scheme is proposed which embeds watermark information to the uneven bit depth salient image pixels. Watermarked image thus produced has better visual transparency with respect to human visual system (HVS) with high payload capacity. In proposed scheme, salient pixels are determined using the bottom-up Graph-Based Visual Saliency (GBVS) model. Experimental results reveal that proposed scheme has less perceptual error as well as improved robustness than existing spatial domain embedding scheme. The visual attention model consists of two steps: first forming activation maps on certain feature channels, and then normalizing them in a way which highlights conspicuity and admits combination with other maps. The model is simple and biologically plausible insofar as it is naturally parallelized The Performance Results with Comparisons: In this paper, a new spatial domain adaptive image watermarking scheme is proposed which embeds watermark information in the least significant bit depth pixels with respect to a well known selective visual attention model GBVS. It is experimentally shown for a significant percentage of images the saliency distribution of pixels remain same even after embedding in multiple bit plane LSB embedding. Experimental result also reveals that the perceptual transparency error due to embedding in proposed scheme is less than that of normal LSB embedding scheme. In the attacked environment, this blind proposed scheme can be easily used to make the method visually more robust. INTRODUCTION RESULT CONCLUSIONS REFERENCES Nov 3-5 2011 1.Kutter, M., et. Al. “Image watermarking techniques, Proceedings ofthe IEEE, Special Issue on Identi cation and Protection of Multimedia Information, 1999. 2.Itti, L., Koch C., Niebur E.: A model of saliency based visual attention for rapid scene analysis, IEEE Trans. on PAMI , vol. 20, pp. 1254- 1259,1998. 3.O.LeMeur, D.Thoreau, P.LeCallet and D.Barba, A Spatio Temporal Model of the Selective Human Visual Attention, 4.Mickael Guironnet, Nathalie Guyader, Denis Pellerin and Patricia Ladret, Static and Dynamic Feature based Visual Attention Model: Comparison to Human Judgment. 5.Christopher Wing Hong Ngau, Li-Minn Ang and Kah Phooi Seng, Bottom-up Visual Saliency Map Using Wavelet Transform Domain, 978-1-4244-5540-9/10 ©2010 IEEE. 6.Arijit Sur, et. al, “A New Image Watermarking Scheme using Saliency Based Visual Attention Model”, IEEE 978-1-4244-4859-3/09/©2009 7.C.C. Chang , et. al, Finding optimal least- significant-bit substitution in image hiding by dynamic programming strategy, Pattern Recognition 36 (2003) 1583–1595 8.C.C. Chang, , et. al, A block based digital watermarks for copy protection of images, Proceedings of Fifth Asia-Pacific Conference on Communications/Fourth Optoelectronics and Communications Conference, Beijing, China, Vol. WATERMARK ENCODER & DECODER Sl. No. Method TPE PSNR Capacity (bpp) Salient Watermarking Model 1 Sur et. al [6] 14.88x10 -5 N.A 0.0017 2 Our Method 0.0485 42.6287 4 Non Salient Watermarking Model 1 LSB substitutio n [7] N.A 38.3425 3 2 GA[8] N.A 38.3220 3 3 IP LSB [9] N.A 35.0751 4 4 Optimal LSB [9] N.A 34.9065 4 5 PWLC [10] N.A 48.3500 0.68 Sl . Paramete r Result 1 SNR 32.4235 dB 2 MSE 10.0539 3 PSNR 42.6287 dB 4 IF 0.99943 5 MD 7 6 AD 2.3418 7 AVD 0.043198 8 NAD 0.018951 9 NMSE 0.0005723 4 10 LMSE 0.051607 11 SER 11.058 12 KLD 0.80179 13 UIQI 0.9979 14 SSIM 0.9557 15 MSSIM 0.9904 16 VSNR 32.2061 17 VIF 0.8235 18 PVIF 0.7821 19 IFC 7.6012 20 NQM 26.1997 21 WSNR 37.0613 22 SC 0.999986 23 PQS 3.662338 24 TPE 0.0485 Performance analysis results for imperceptibility: CONTACT: Name: S. Majumder EMAIL: [email protected]

Upload: lesley-haynes

Post on 31-Dec-2015

213 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Title of On the Implementation of a Information Hiding Design based on Saliency Map A.Basu, T. S. Das and S. K. Sarkar/ Jadavpur University/ Kolkata/ India

Title of On the Implementation of a Information Hiding Design based on Saliency MapA.Basu, T. S. Das and S. K. Sarkar/ Jadavpur University/ Kolkata/ India/ [email protected], [email protected], [email protected]

Swanirbhar Majumder/NERIST (Deemed University)/ Itanagar/ India/ [email protected]

In this paper, an adaptive spatial domain image watermarking scheme is proposed which embeds watermark information to the uneven bit depth salient image pixels. Watermarked image thus produced has better visual transparency with respect to human visual system (HVS) with high payload capacity. In proposed scheme, salient pixels are determined using the bottom-up Graph-Based Visual Saliency (GBVS) model. Experimental results reveal that proposed scheme has less perceptual error as well as improved robustness than existing spatial domain embedding scheme.

The visual attention model consists of two steps: first forming activation maps on certain feature channels, and then normalizing them in a way which highlights conspicuity and admits combination with other maps. The model is simple and biologically plausible insofar as it is naturally parallelized

The Performance Results with Comparisons: In this paper, a new spatial domain adaptive image watermarking scheme is proposed which embeds watermark information in the least significant bit depth pixels with respect to a well known selective visual attention model GBVS. It is experimentally shown for a significant percentage of images the saliency distribution of pixels remain same even after embedding in multiple bit plane LSB embedding. Experimental result also reveals that the perceptual transparency error due to embedding in proposed scheme is less than that of normal LSB embedding scheme. In the attacked environment, this blind proposed scheme can be easily used to make the method visually more robust.

INTRODUCTION RESULT CONCLUSIONS

REFERENCESNov 3-5

20111. Kutter, M., et. Al. “Image watermarking techniques, Proceedings

ofthe IEEE, Special Issue on Identification and Protection of Multimedia Information, 1999.

2. Itti, L., Koch C., Niebur E.: A model of saliency based visual attention for rapid scene analysis, IEEE Trans. on PAMI , vol. 20, pp. 1254-1259,1998.

3. O.LeMeur, D.Thoreau, P.LeCallet and D.Barba, A Spatio Temporal Model of the Selective Human Visual Attention,

4. Mickael Guironnet, Nathalie Guyader, Denis Pellerin and Patricia Ladret, Static and Dynamic Feature based Visual Attention Model: Comparison to Human Judgment.

5. Christopher Wing Hong Ngau, Li-Minn Ang and Kah Phooi Seng, Bottom-up Visual Saliency Map Using Wavelet Transform Domain, 978-1-4244-5540-9/10 ©2010 IEEE.

6. Arijit Sur, et. al, “A New Image Watermarking Scheme using Saliency Based Visual Attention Model”, IEEE 978-1-4244-4859-3/09/©2009

7. C.C. Chang , et. al, Finding optimal least-significant-bit substitution in image hiding by dynamic programming strategy, Pattern Recognition 36 (2003) 1583–1595

8. C.C. Chang, , et. al, A block based digital watermarks for copy protection of images, Proceedings of Fifth Asia-Pacific Conference on Communications/Fourth Optoelectronics and Communications Conference, Beijing, China, Vol. 2, 1999, pp. 977–980.

9. Cheng-HsingYang, Inverted pattern approach to improve image quality of information hiding by LSB substitution, Pattern Recognition 41 (2008) 2674 – 2683.

10.Chang-Lung Tsai, , et. al, Reversible data hiding and lossless reconstruction of binary images using pair-wise logical computation mechanism, Pattern Recognition 38 (2005) 1993 – 2006.

WATERMARK ENCODER & DECODER

Sl.No.

Method TPE PSNRCapacity

(bpp)Salient Watermarking Model1 Sur et. al [6] 14.88x10-5 N.A 0.00172 Our Method 0.0485 42.6287 4Non Salient Watermarking Model

1LSB substitution [7]

N.A 38.3425 3

2 GA[8] N.A 38.3220 33 IP LSB [9] N.A 35.0751 44 Optimal LSB [9] N.A 34.9065 45 PWLC [10] N.A 48.3500 0.68

Sl. Parameter Result1 SNR 32.4235 dB2 MSE 10.05393 PSNR 42.6287 dB4 IF 0.999435 MD 76 AD 2.34187 AVD 0.0431988 NAD 0.0189519 NMSE 0.00057234

10 LMSE 0.05160711 SER 11.05812 KLD 0.8017913 UIQI 0.997914 SSIM 0.955715 MSSIM 0.990416 VSNR 32.206117 VIF 0.823518 PVIF 0.782119 IFC 7.601220 NQM 26.199721 WSNR 37.061322 SC 0.99998623 PQS 3.66233824 TPE 0.0485

Performance analysis results for imperceptibility:

CONTACT:Name: S. MajumderEMAIL: [email protected]