2. Overview: DigitalWatermarking Least Significant BitAlgorithm Discrete CosineTransform Algorithm Metrics for quality LSB and DCT: Comparison Conclusion References 3 May 20142 Digital Image Protection UsingAdaptive WatermarkingTechniques 3. Digital Watermarking: The concept 3 May 20143 Digital Image Protection UsingAdaptive WatermarkingTechniques 4. Digital Watermarking: Types Visible watermark: 3 May 20144 Digital Image Protection UsingAdaptive WatermarkingTechniques Fig.1.VisibleWatermark 5. Digital Watermarking: Types (cont.) Invisible watermark: 3 May 20145 Digital Image Protection UsingAdaptive WatermarkingTechniques Fig.2.InvisibleWatermark 6. Digital Watermarking: Classification 3 May 20146 Digital Image Protection UsingAdaptive WatermarkingTechniques Fig.3.Classification ofWatermarking techniques 7. Least Significant Bit Technique 10110101 01001101 11001101 00010011 00010100 01001010 10110100 01001100 11001101 00010010 00010100 01001011 3 May 20147 Digital Image Protection UsingAdaptive WatermarkingTechniques Fig.4.The LSB Substitution technique Embed 001001 Secret Data Original Image encoded format Embedded data 8. 8 Secret data Convert to bit stream Conversion of data Encode Data Embedding Extract Decode Data Original Source Image Load Source image Convert to stego image Extracted secret data Digital Image Protection UsingAdaptive WatermarkingTechniques 3 May 2014 Fig.5.The LSB Substitution procedure 9. Discrete Cosine Transform Algorithm 3 May 20149 Digital Image Protection UsingAdaptive WatermarkingTechniques Fig.6.The DCT algorithm Fragment Use DCT Block selection Coefficient selection EmbedWatermarks Apply inverse DCT 10. Discrete Cosine Transform Algorithm 3 May 201410 Digital Image Protection UsingAdaptive WatermarkingTechniques Fig.7.The DCT region Low band frequency region Middle band frequency region High band frequency region 11. 3 May 201411 Digital Image Protection UsingAdaptive WatermarkingTechniques Fig.8.The MBEC algorithm Select the FM region Select two points say Pi(u1,v1) & Pi(u2,v2) Pi(u1,v1)>= Pi(u2,v2) Encode 0 Encode 1 A A True False 12. 16 11 10 16 24 40 51 61 12 12 14 19 26 48 16 55 14 13 16 24 40 57 69 56 14 17 20 29 51 87 80 62 18 22 37 56 68 10 9 10 8 77 24 35 55 64 81 10 4 11 3 92 49 64 78 87 10 3 12 1 12 0 10 1 72 92 95 98 11 2 10 0 10 3 99 E.g.: 3 May 2014 Digital Image Protection UsingAdaptive WatermarkingTechniques12 16 11 10 16 24 40 51 61 12 12 14 19 26 48 16 55 14 13 16 24 40 57 69 56 14 17 20 29 51 87 80 62 18 22 37 56 68 10 9 10 8 77 24 35 55 64 81 10 4 11 3 92 49 64 78 87 10 3 12 1 12 0 10 1 72 92 95 98 11 2 10 0 10 3 99 Fig.9.An illustration of DCT algorithm Transform Embed Inverse Transform 16 11 10 16 24 40 51 61 12 12 14 19 26 48 16 55 14 13 16 24 40 57 69 56 14 17 20 29 51 87 80 62 18 22 37 56 68 10 9 10 8 77 24 35 55 64 81 10 4 11 3 92 49 64 78 87 10 3 12 1 12 0 10 1 72 92 95 98 11 2 10 0 10 3 99 16 11 10 16 24 40 51 61 12 12 14 19 26 48 16 55 14 13 16 24 40 57 69 56 14 17 20 29 51 87 80 62 18 22 37 56 68 10 9 10 8 77 24 35 55 64 81 10 4 11 3 92 49 64 78 87 10 3 12 1 12 0 10 1 72 92 95 98 11 2 10 0 10 3 99 13. MBEC: The actual thing. 3 May 2014 Digital Image Protection UsingAdaptive WatermarkingTechniques13 16 11 10 16 24 40 51 61 12 12 14 19 26 48 16 55 14 13 16 24 40 57 69 56 14 17 20 29 51 87 80 62 18 22 37 56 68 109 108 77 24 35 55 64 81 104 113 92 49 64 78 87 103 121 120 101 72 92 95 98 112 100 103 99 14 = 14; Encode 0 20 < 22; Encode 1 Fig.10.MBEC encoding 14. Metrics for quality: Peak Signal to Noise Ratio(PSNR) PSNR> 0.35 dB PSNR = 10 log10 . A2 1 N x M [f(i,j) f(i,j)]2 Normalized Cross-correlation(NC) NC>0.75 3 May 201414 Digital Image Protection UsingAdaptive WatermarkingTechniques 15. LSB & DCT : Comparison LSB Substitution technique DCT technique Spatial domain technique Frequency domain technique Provides limited ROBUSTNESS Provides better ROBUSTNESS Embeds little INFORMATION Embeds more INFORMATION Lesser NOISE immunity Highly immune to NOISE SIMPLE implementation COMPLEX implementation 3 May 201415 Digital Image Protection UsingAdaptive WatermarkingTechniques Table.1.LSB and DCT technique comparison 16. Conclusion DigitalWatermarking proves crucial. A combination of more than one technique is required. DCT proves to be advantageous. 3 May 201416 Digital Image Protection UsingAdaptive WatermarkingTechniques 17. References [1].Yusuf Perwej, Firoj Parwej ,Asif Perwej -The International Journal of Multimedia & Its Applications (IJMA)Vol.4, No.2, April 2012,An AdaptiveWatermarkingTechnique for the copyright of digital images and Digital Image Protection. [2] Lin Liu -A Survey of DigitalWatermarkingTechnologies. [3]. Petitcolas, F.A. P.,Anderson, R. J.: Kuhn, M. G., Information Hiding A Survey, Proceedings of the IEEE, Special Issue on Protection of Multimedia Content. 3 May 201417 Digital Image Protection UsingAdaptive WatermarkingTechniques 18. THANKYOU! 3 May 201418 Digital Image Protection UsingAdaptive WatermarkingTechniques 19. 3 May 201419 Digital Image Protection UsingAdaptive WatermarkingTechniques
