Presentation is loading. Please wait.

Presentation is loading. Please wait.

Information hiding in stationary images staff corporal Piotr Lenarczyk Military Uniwersity of Technology Institute of Electronics and Telecomunication.

Similar presentations


Presentation on theme: "Information hiding in stationary images staff corporal Piotr Lenarczyk Military Uniwersity of Technology Institute of Electronics and Telecomunication."— Presentation transcript:

1 Information hiding in stationary images staff corporal Piotr Lenarczyk Military Uniwersity of Technology Institute of Electronics and Telecomunication

2 Watermarking „Digital image watermarking means embedding information into image in such a way that it is imperceptible to a human observer, but can be easily detected by special computer algorithm.” Ingemar J. Cox

3 Applications of watermarking Owner Identification Proof of Ownership Copy Control Device Control Broadcast Monitoring Content Authentication

4 Method of changing Discrete Cosine Transform coefficients in frueqency domain

5 2-Dimensional Discrete Cosine Transform DCT 2D M,N = 8 m,n – discrete index of pixel position k,l discrete index of DCT coefficient

6 DCT 2D Basis Functions Basis functions matrices of Discrete Cosine Transform for block 8x8 pix 2 dimensional Discrete Cosine Transform coefficients Standard quantization matrix used in JPEG algorithm 1611101624405161 12 141926586055 1413162440576956 1417222951878062 182237566810910377 243555648110411392 49647887103121120101 7292959811210010399 Q=

7 162 161162157163161 162 161162157163161 162 161162157163161 162 161162157163161 162 161162157163161 164 158155161159 160 163158160162159156 159 155157158159156157 1283,54,773,24-0,20,25-0,55-4,595,64 7,95-0,79-4,931,962,98-3,83,32 -5,03-0,3-1,551,72-0,68-0,451,85-2,2 2,261,141,70,92-0,76-1,370,211,14 -1,15-0,34-1,361,751,12-1,48-0,67 1,210,46-1,8-0,11-2,010,711,690,76 -1,70,263,11,641,63-2,25-1,2-0,9 1,33-0,42-2,38-1,59-0,881,970,550,65 Matrix A represents block of 8 x 8 pixel values Matrix B represents DCT 2D calculated on matrix A: A = B = 0,55 23,55 Original image Watermarked image

8 Perceptual transparency Data payloadRobustness

9 Original pictureWatermarked picture for scalling coefficient k=23 Watermarked picture for scalling coefficient k=3000

10 BCH code influence on Bit Error Rate

11 Algorithm Characteristic Supported formats BMP 16, 24 bits PNG, JPG 24 bits Minimal picture size [pix x pix]256 x 256 Data payload [bits]130 Error correctionBCH – correction capability of 14 bits Redundancy of embedded informationAt least four times Method advantagesRobust against multiply rotation of 90 degrees Robust against high-pass filtering Robust against picture distortion like text embedding, painting Robust against lossy JPG compression - approximately 70% JPEG format popularity Program execution time [sec] BMP format, picture size 400 x 400 pix Coder - 0,12 Decoder - 0,07 Program execution time [sec] JPG format, picture size 400 x 400 pix Coder - 0,63 Decoder - 0,29

12 Thank you for your attention. Questions?


Download ppt "Information hiding in stationary images staff corporal Piotr Lenarczyk Military Uniwersity of Technology Institute of Electronics and Telecomunication."

Similar presentations


Ads by Google