Authors : Dalila Goudia (LIRMM-SIMPA) Marc Chaumont (LIRMM, France) William Puech (LIRMM, France) Naima Hadj Said (SIMPA, Algeria) EUSIPCO 2011
• Generalities – Data hiding – Joint data hiding and compression approach – JPEG2000 standard – Trellis Coded Quantization (TCQ) • Joint JPEG2000 compression & data hiding scheme – The TCQ-based data hiding strategy – The proposed joint scheme – The embedding and extraction algorithms • Experimental evaluations – Protocol 1: data hiding performances – Protocol 2 : compression performances • Conclusion EUSIPCO 2011
• Generalities – Data hiding – Joint data hiding and compression approach – JPEG2000 standard – Trellis Coded Quantization (TCQ) • Joint JPEG2000 compression & data hiding scheme – The TCQ-based data hiding strategy – The proposed joint scheme – The embedding and extraction algorithms • Experimental evaluations – Protocol 1: data hiding performances – Protocol 2 : compression performances • Conclusion EUSIPCO 2011
Secret Secret Key message message Host Embedding Transmission Detector Key document algorithm via network • Content description and meta data enrichment applications • Embed the maximum amount of data (payload) in the host image without perceptually distorting it • The information embedded must be recovered without error during the extraction stage 1 EUSIPCO 2011
Joint data hiding and compression Content description applications Compliant Good Robustness Low High syntax to visual complexity Payload bitstream compression quality 2 EUSIPCO 2011
Rate control stage Forward Pre- Tier 1 Tier 2 wavelet Quantization Processing encoder encoder transform codestream JPEG2000 original image ROI processing ISO/IEC 15444- 1, “Information Technology - JPEG2000 Image Coding System-Part 1: Core Coding System”, 2000 JPEG2000 features • Good compression performances, and smooth transmission from lossy to lossless • Progressive transmission • Regions of interest • Flexible file format • Error Resilience • … 3 EUSIPCO 2011
Partitioning of a scalar quantizer into 4 subsets conbined D 0 0 to form 2 union quantizers: D 2 & A D D A D D D 2 0 0 2 1 1 3 1 D 0 D 1 2 D 3 D 3 3 D 1 D 2 4 Subsets D i of the 2 union quantizers A 0 and A 1 are used to D 0 label the branches of a trellis D 0 5 Quantization is performed by running the Viterbi D 2 algorithm to find the optimal path (minimum distortion D 3 6 path) through the trellis D 1 The least significant bit (LSB) of the TCQ indices D 1 7 determine the path through the trellis D 3 4 EUSIPCO 2011
• Generalities – Data hiding – Joint data hiding and compression approach – JPEG2000 standard – Trellis Coded Quantization (TCQ) • Joint JPEG2000 compression & data hiding scheme – The TCQ-based data hiding strategy – The proposed joint scheme – The embedding and extraction algorithms • Experimental evaluations – Protocol 1: data hiding performances – Protocol 2 : compression performances • Conclusion EUSIPCO 2011
• The data is hidden during the quantization process without any additional stage for hiding data. • Data hiding strategy derived from the QIM (Quantization Index Modulation) principles • Integration into a TCQ approach (trellis) • Quantizers are modulated according to the data to hide • Data is embedded only in the significant wavelet coefficients which have a better chance of survival after JPEG2000 rate allocation stage. • Selected coefficients are quantized with the associated quantizer 5 EUSIPCO 2011
The QIM principles applied to TCQ union quantizers of JPEG2000 x m= 0 m= 1 D 0 D 2 D 0 D 2 D 0 D 2 D 0 D 2 D 0 D 2 A 0 -8 Δ -6 Δ -4 Δ -2 Δ Δ 3 Δ 5 Δ 7 Δ 9 Δ ˆ 0 x = 0 -4 -3 -2 -1 0 1 2 3 4 5 q ( A ) 0 = 1 D 3 D 1 D 3 D 1 D 3 D 1 D 3 D 1 D 3 D 1 A 1 -9 Δ -7 Δ -5 Δ -3 Δ - Δ 2 Δ 4 Δ 6 Δ 8 Δ ˆ 0 x -5 -4 -2 -1 0 1 2 3 4 -3 ( ) q A 1 • Union quantizer A 0 : if the bit to embed is the bit 0, then the quantizer D 0 is used to quantize the wavelet coefficient. Otherwise the quantizer D 2 is used. • Union quantizer A 1 : if the bit to embed is the bit 0, then the quantizer D 1 is used to quantize the wavelet coefficient. Otherwise the quantizer D 3 is used. 6 EUSIPCO 2011
D 0 0 D 2 The choice of the branch to traverse is D 2 determined by the value of the bit to be 1 D 0 embedded D 1 2 The trellis is pruned only at the transitions D 3 which correspond to the selected coefficients D 3 3 D 1 The trellis pruning is similar to Miller et al. D 2 scheme (DPTC) 4 D 0 D 0 5 D 2 D 3 6 D 1 D 1 7 D 3 7 EUSIPCO 2011
D 0 0 The trellis is pruned only at the transitions which correspond to the selected D 0 1 coefficients 2 D 1 • if the bit to embed is the bit 0 : D 0 and D 1 are used to quantize the wavelet coefficient. 3 D 1 Trellis structure : remove the red branches 4 D 0 at the considered transition D 0 5 D 1 6 D 1 7 8 EUSIPCO 2011
D 2 0 The trellis is pruned only at the transitions which correspond to the selected D 2 1 coefficients D 3 2 • if the bit to embed is the bit 1 : D 2 and D 3 are used to quantize the wavelet coefficient. 3 D 3 Trellis structure : remove the blue branches 4 D 2 at the considered transition 5 D 2 D 3 6 D 3 7 9 EUSIPCO 2011
• Computation of selection threshold τ IBP for each code-block • Coefficients are selected if their TCQ indices have their absolute magnitude bits greater than τ IBP • Data is hidden in the least significant bits (LSB) of the TCQ indices of the selected coefficients Note : in order to avoid destruction of those LSBs (path in the trellis) by the JPEG2000 R-D optimisation stage, they are moved to a higher bit plane position. 10 EUSIPCO 2011
D 0 D 0 D 0 D 0 D 0 D 0 0 D 2 D 2 D 2 D 2 D 2 D 2 D 2 D 2 D 2 D 2 D 2 D 2 1 D 0 D 0 D 0 D 0 D 0 D 0 D 1 D 1 D 1 D 1 D 1 D 1 2 D 3 D 3 D 3 D 3 D 3 D 3 D 3 D 3 D 3 D 3 D 3 D 3 3 D 1 D 1 D 1 D 1 D 1 D 1 D 2 D 2 D 2 D 2 D 2 D 2 4 D 0 D 0 D 0 D 0 D 0 D 0 D 0 D 0 D 0 D 0 D 0 D 0 5 D 2 D 2 D 2 D 2 D 2 D 2 D 3 D 3 D 3 D 3 D 3 D 3 6 D 1 D 1 D 1 D 1 D 1 D 1 D 1 D 1 D 1 D 1 D 1 D 1 7 D 3 D 3 D 3 D 3 D 3 D 3 Transition 1 2 3 4 5 11 EUSIPCO 2011
D 0 D 0 D 0 D 0 D 0 D 0 0 D 2 D 2 D 2 D 2 D 2 D 2 D 2 D 2 D 2 D 2 D 2 D 2 1 D 0 D 0 D 0 D 0 D 0 D 0 D 1 D 1 D 1 D 1 D 1 D 1 2 D 3 D 3 D 3 D 3 D 3 D 3 D 3 D 3 D 3 D 3 D 3 D 3 3 D 1 D 1 D 1 D 1 D 1 D 1 D 2 D 2 D 2 D 2 D 2 D 2 4 D 0 D 0 D 0 D 0 D 0 D 0 D 0 D 0 D 0 D 0 D 0 D 0 5 D 2 D 2 D 2 D 2 D 2 D 2 D 3 D 3 D 3 D 3 D 3 D 3 6 D 1 D 1 D 1 D 1 D 1 D 1 D 1 D 1 D 1 D 1 D 1 D 1 7 D 3 D 3 D 3 D 3 D 3 D 3 Transition 1 2 3 4 5 Selection no yes no yes no 11 EUSIPCO 2011
Message 0 1 D 0 D 0 D 0 D 0 D 0 0 D 2 D 2 D 2 D 2 D 2 D 2 D 2 D 2 D 2 D 2 1 D 0 D 0 D 0 D 0 D 0 D 1 D 1 D 1 D 1 D 1 2 D 3 D 3 D 3 D 3 D 3 D 3 D 3 D 3 D 3 D 3 3 D 1 D 1 D 1 D 1 D 1 D 2 D 2 D 2 D 2 D 2 4 D 0 D 0 D 0 D 0 D 0 D 0 D 0 D 0 D 0 D 0 5 D 2 D 2 D 2 D 2 D 2 D 3 D 3 D 3 D 3 D 3 6 D 1 D 1 D 1 D 1 D 1 D 1 D 1 D 1 D 1 D 1 7 D 3 D 3 D 3 D 3 D 3 Transition 1 2 3 4 5 Selection no yes no yes no 11 EUSIPCO 2011
Message 0 1 D 0 D 0 D 0 D 0 D 0 0 D 2 D 2 D 2 D 2 D 2 D 2 D 2 D 2 D 2 D 2 1 D 0 D 0 D 0 D 0 D 0 D 1 D 1 D 1 D 1 D 1 2 D 3 D 3 D 3 D 3 D 3 D 3 D 3 D 3 D 3 D 3 3 D 1 D 1 D 1 D 1 D 1 D 2 D 2 D 2 D 2 D 2 4 D 0 D 0 D 0 D 0 D 0 D 0 D 0 D 0 D 0 D 0 5 D 2 D 2 D 2 D 2 D 2 D 3 D 3 D 3 D 3 D 3 6 D 1 D 1 D 1 D 1 D 1 D 1 D 1 D 1 D 1 D 1 7 D 3 D 3 D 3 D 3 D 3 Transition 1 2 3 4 5 Selection no yes no yes no 11 EUSIPCO 2011
message to hide …1001101. . . . Rate- TCQ Pre- Forward Verification of EBCOT distorsion Original quantization Data hided processing Wavelet the presence of (Tier1 optimization with Data image image Transform the hidden data encoder) (Tier2 Hiding bistream encoder) τ IBP Selection Coefficients thresholds selection computation The joint JPEG2000 encoder/data hiding embedding scheme. 12 EUSIPCO 2011
message to hide …1001101. . . . Rate- TCQ Pre- Forward Verification of EBCOT distorsion Original quantization Data hided processing Wavelet the presence of (Tier1 optimization with Data image image Transform the hidden data encoder) (Tier2 Hiding bistream encoder) τ IBP Selection Coefficients thresholds selection computation 1. Computation of the selection thresholds τ IBP 13 EUSIPCO 2011
Recommend
More recommend
