Secure Spread Spectrum Watermarking for Multimedia

SecureSpread Spectrum

Watermarking for Multimedia Ishani Vyas

CS590 Winter 2008

Agenda To explain about watermarking

characteristics How to insert watermark in spread spectrum Structure of the Watermark Experimental Result

Characteristics of Watermarking Unobtrusive Perceptually invisible or presence should not interfere with work being

protected Robust Difficult to remove. May be removed sufficient knowledge of the process of

insertion. Available of Partial Knowledge-Result in degradation in data fidelity.

Watermark should be robust to Common Signal Processing-should be retrievable even if its operations

are applied to the data. For example D/A&A/D conversion,resampling ,requantization.

Common geometric distortions (image and video data)- Watermarks in image and video data should be immune from geometric image operations such as rotation, translation, cropping and scaling

Collusion and Forgery- Watermark should be robust to collusion by multiple individuals who each posses a watermarked copy of data.

Cont Watermark should be robust to combining copies of

same data set to destroy the watermarks Universal – apply same digital watermark algorithm to all

three media under consideration. Unambiguous- Retrieval of the watermark should

unambiguously identify the owner.

Why need of Spread Spectrum watermarking The previous methods have not explicitly identified the perceptually

most significant components of a signal as the destination for the watermark.

Key insight of this paper is that the watermark must be placed in perceptually significant regions of the data despite the risk of potential fidelity distortions.

Apply watermarking in significant spectral components of a signal. How common signal transformations, (i.e compression, quantization

and manipulation) affects the frequency spectrum of a signal. Watermark should be embedded in the data`s perceptually

significant frequency components. But how to insert watermark into significant components of

frequency spectrum without any distortions.

Watermarking in frequency domain

Lossy compression is an operation that usually eliminates perceptually non-salient components of an image or sound.

To preserve watermark in the face of such an operation, watermark must be placed in the perceptually significant regions of the data.Data loss usually occurs among the high frequency components .

Geometrics distortions specific to image and video.(Rotation,scaling ,translation,cropping)- minimum 4-9 points between the original and the distorted watermark image..Again loss of data at high frequency spectral regions of the image.

Common signal distortion(D/A&A/D, resampling ,requantization, including dithering and recompression and common signal enhancements to image contrast/or color)

Spread Spectrum coding of watermark How to insert the watermark into most significant region

of an spectrum without such alteration become noticable.

Structure of Watermark

Experimental Result

Watermarked image

Experiment 1:Uniqueness of Watermark

Experiment 2:Image Scaling

Experiment3: JPEG coding distortion

Experiment 4:Dithering Distortion

Experiment 5:Clipping

Experiment 6:Print ,Xerox or scan

Experiment 7:Attack by watermarking watermarked image

Experiment 8:Attack by collusion