Self-Embedding Authentication Watermarking with Effective Tampered Location Detection and High-Quality Image Recovery
Abstract
:1. Introduction
2. Related Work
3. Proposed Method
3.1. Description of Symbol Definitions
- : the weight of the original image
- : the height of the original image
- : the original image
- : the watermarked image
- : the tampered image
- : the recovered image
- : the size of a block
- : the total number of the blocks in the image
- : each block of the image
- : the pixel value of each block
- : the secret key
- : the pseudo- random number that is generated by
- : the map** block
- : the mean value of each block
- : the recovery data of each block
- : the authentication data of each block
- : the recovery data that is generated from
- : the authentication data that is generated from
- : the size of the authentication data
- : the number of bits that will be embedded into each pixel
- : the table that is marked whether it has been tampered with
- : the map** table of
- : the enlarged image
3.2. Block-Wise Detection
3.3. Pixel-Wise Detection
4. Experimental Results and Comparison
4.1. Digital Image Tamper Detection
4.2. Comparison with Other Methods
5. Conclusions
Author Contributions
Acknowledgments
Conflicts of Interest
References
- Feng, J.; Potkonjak, M. Real-time watermarking techniques for sensor networks. Proc. SPIE 2003. [Google Scholar] [CrossRef]
- Wong, J.L.; Feng, J.; Kirovski, D.; Potkonjak, M. Security in sensor networks: Watermarking techniques. In Wireless Sensor Networks; Raghavendra, C.S., Sivalingam, K.M., Znati, T., Eds.; Springer: Boston, MA, USA, 2004. [Google Scholar]
- Shi, X.; ** embedding strategy. Signal Process. 2017, 138, 280–293. [Google Scholar] [CrossRef]
- Kim, C.; Shin, D.; Yang, C.-N. Self-embedding fragile watermarking scheme to restoration of a tampered image using AMBTC. Pers. Ubiquit. Comput. 2018, 22, 11–22. [Google Scholar] [CrossRef]
- Tai, W.-L.; Liao, Z.-J. Image self-recovery with watermark self-embedding. Signal Process. 2018, 65, 11–25. [Google Scholar] [CrossRef]
- Wang, C.; Zhang, H.; Zhou, X. A Self-recovery fragile image watermarking with variable watermark capacity. Appl. Sci. 2018, 8, 548. [Google Scholar] [CrossRef]
- Katzenbeisser, S.; Petitcolas, F.A.P. Information Hiding Techniques for Steganography and Digital Watermarking (Artech House Computer Security Series); Artech House: Norwood, MA, USA, 2000. [Google Scholar]
- Keys, R. Cubic convolution interpolation for digital image processing. IEEE Trans. Acoust. Speech Signal Process. 1981, 29, 1153–1160. [Google Scholar] [CrossRef] [Green Version]
i | 1 | 2 | 3 | 4 | … | |
---|---|---|---|---|---|---|
generation | ↓ | ↓ | ↓ | ↓ | ↓ | ↓ |
4 | 1 | 2 | 3 | |||
… | ||||||
Associated blocks | ↓ | ↓ | ↓ | ↓ | ↓ | ↓ |
FPR (%) | PSNR(w) (dB) | PSNR(r) (dB) | ||||
---|---|---|---|---|---|---|
2 | 4 | 3 | 16 | 0.17335 | 41.28 | 44.48 |
3 | 3 | 3 | 16 | 0.13257 | 41.77 | 44.51 |
4 | 4 | 2 | 24 | 0.17335 | 47.32 | 44.68 |
Image | PSNR(w) (dB) | Tamper Rate | 10% | 20% | 30% | 40% | 50% |
---|---|---|---|---|---|---|---|
Lena | 47.20 | PSNR(r) (dB) | 49.47 | 44.39 | 41.23 | 38.58 | 36.61 |
FNR (%) | 0 | 0 | 0 | 0 | 0 | ||
FPR (%) | 0.173 | 0.391 | 0.669 | 0.787 | 0 | ||
Baboon | 47.29 | PSNR(r) (dB) | 38.69 | 35.55 | 33.95 | 32.93 | 32.13 |
FNR (%) | 0 | 0 | 0 | 0 | 0 | ||
FPR (%) | 0.173 | 0.391 | 0.669 | 0.787 | 0 | ||
Peppers | 47.23 | PSNR(r) (dB) | 42.84 | 40.54 | 38.32 | 36.76 | 35.17 |
FNR (%) | 0 | 0 | 0 | 0 | 0 | ||
FPR (%) | 0.173 | 0.391 | 0.669 | 0.787 | 0 | ||
Airplane | 47.33 | PSNR(r) (dB) | 46.59 | 44.54 | 42.83 | 40.32 | 36.79 |
FNR (%) | 0 | 0 | 0 | 0 | 0 | ||
FPR (%) | 0.173 | 0.391 | 0.669 | 0.787 | 0 | ||
Tiffany | 47.54 | PSNR(r) (dB) | 45.81 | 41.87 | 40.08 | 38.36 | 36.79 |
FNR (%) | 0 | 0 | 0 | 0 | 0 | ||
FPR (%) | 0.173 | 0.391 | 0.669 | 0.787 | 0 |
Methods | PSNR(w) (dB) | PSNR(r) (dB) | Condition of Restoration |
---|---|---|---|
Yang and Shen [21] | 40.7 | 32 | <50% |
Yang et al. [22] | 51.3 | 36 | <50% |
Qian et al. [24] | 37.9 | 35 | <35% |
Qin et al. [31] | 46 | 41 | <45% |
Kim et al. [32] | 43.7 | 33.6 | <50% |
Proposed method ( block-wise detection) | 43.73 | 36.62 | <50% |
Proposed method ( pixel-wise detection) | 41 | 37.26 | <50% |
Block-Wise | Block-Wise | Block-Wise | Pixel-Wise | |
---|---|---|---|---|
512_lena | 0.92739 | 0.93511 | 0.958791 | 0.94159 |
512_baboon | 0.97581 | 0.97843 | 0.986461 | 0.97489 |
512_peppers | 0.93037 | 0.93739 | 0.961136 | 0.93938 |
512_airplane | 0.92725 | 0.93173 | 0.95682 | 0.94193 |
512_tiffany | 0.91812 | 0.92558 | 0.955061 | 0.94091 |
512_lake | 0.94736 | 0.95261 | 0.970439 | 0.96453 |
Methods | FPR(%) | FNR(%) |
---|---|---|
Tong et al. [25] | 0.22 | 0 |
Chen et al. [26] | 0.25 | 0 |
Ansari et al. [28] | 0.5 | 0.01 |
Wang et al. [34] | 0.30 | 0 |
Proposed method ( block-wise detection) | 0.173 | 0 |
Proposed method ( pixel-wise detection) | 0 | 0 |
Block-Wise | Block-Wise | Block-Wise | Pixel-Wise | |
---|---|---|---|---|
Tamper Rate (%) | 14.35 | 14.35 | 14.35 | 14.35 |
Recovered PSNR(r) (dB) | 38.749796 | 38.855165 | 38.192091 | 39.2149 |
FNR(%) | 0 | 0 | 0 | 0 |
FPR(%) | 0.029 | 0.025 | 0.038 | 0 |
Block size | 2 × 4 | 3 × 3 | 4 × 4 | |
---|---|---|---|---|
Tamper Rate (%) | 14.35 | 14.35 | 14.35 | |
Recovered PSNR(r) | average | 41.55 | 41.54 | 40.61 |
highest | 47.55 | 48.09 | 48.18 | |
lowest | 37.45 | 37.41 | 36.77 | |
FNR(%) | average | 0.000082 | 0.000274 | 0.000562 |
highest | 0.000824 | 0.003803 | 0.006887 | |
lowest | 0.000000 | 0.000000 | 0.000000 | |
FPR(%) | average | 0.02850 | 0.02531 | 0.03764 |
highest | 0.02855 | 0.02548 | 0.03803 | |
lowest | 0.02801 | 0.02324 | 0.03526 |
Hsu and Tu [23] | Lo and Hu [27] | Singh & Singh [29] | Yin et al. [30] | Qin et al. [31] | Tai and Liao [33] | Proposed Block-Wise | Proposed Pixel-Wise | |
---|---|---|---|---|---|---|---|---|
Block Size | ||||||||
Block division | 5120 | 4096 | 16,384 | 4096 | 64,516 | 4096 | 4096 | 4096 |
Predictive coding and it inverse | 0 | 65,536 | 0 | 0 | 0 | 0 | 0 | 0 |
Arnold’s permutation | 0 | 0 | 0 | 0 | 0 | 4096 | 0 | 0 |
DWT | 4096 | 0 | 0 | 16,384 | 0 | 4096 | 0 | 0 |
DWT embedding | 0 | 0 | 0 | 0 | 0 | 4096 | 0 | 0 |
DCT | 0 | 0 | 16,384 | 0 | 0 | 0 | 0 | 0 |
Mean calculation | 4096 | 0 | 16,384 | 0 | 0 | 0 | 4096 | 4096 |
Watermark hashing | 1024 | 0 | 16,384 | 0 | 64,516 | 4096 | 4096 | 65,536 |
Hilbert curve transformation | 1024 | 0 | 0 | 4096 | 64,516 | 0 | 0 | 0 |
XOR | 20,480 | 0 | 0 | 4096 | 0 | 4096 | 4096 | 65,536 |
LSB embedding | 1024 | 0 | 16,384 | 0 | 0 | 0 | 4096 | 65,536 |
Histogram shifting and modification | 1024 | 65,536 | 16,384 | 4096 | 64,516 | 0 | 0 | 0 |
Total operation count | 37,888 | 139,264 | 81,920 | 32,768 | 258,064 | 62,464 | 20,480 | 204,800 |
© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Share and Cite
Lee, C.-F.; Shen, J.-J.; Chen, Z.-R.; Agrawal, S. Self-Embedding Authentication Watermarking with Effective Tampered Location Detection and High-Quality Image Recovery. Sensors 2019, 19, 2267. https://doi.org/10.3390/s19102267
Lee C-F, Shen J-J, Chen Z-R, Agrawal S. Self-Embedding Authentication Watermarking with Effective Tampered Location Detection and High-Quality Image Recovery. Sensors. 2019; 19(10):2267. https://doi.org/10.3390/s19102267
Chicago/Turabian StyleLee, Chin-Feng, Jau-Ji Shen, Zhao-Ru Chen, and Somya Agrawal. 2019. "Self-Embedding Authentication Watermarking with Effective Tampered Location Detection and High-Quality Image Recovery" Sensors 19, no. 10: 2267. https://doi.org/10.3390/s19102267