In this algorithm, a large amount of extracted features should be compressed and passed as side information for making the operation reversible. Though compared with first method section 2. When the watermarking information is embedded into the frequency domain the technique is called the frequency domain watermarking. In extraction phase watermarked image is correlated with the carrier signal. During embedding the image is tiled in a specific dimension and the message is embedded to these tiles. The resultant L-ary symbols are then converted into their original decimal values using the method given in section 3.

Unsuccessful extracted watermarked image is passed through affine correction and weighted extraction phases, and finally if it still fails to extract data successfully the watermarked image is declared as inauthentic. The embedding capacity can be increased using more than one color-pair sample pair. The message embedding procedure is the following equation. Additional information can be compressed, if necessary, before concatenation. Relative address coding RAC 75 Figure 4. Lossless zooming, contraction and expansion of images have the characteristic of retrieving the original image from the product changed or watermarked image. We discuss all the previously proposed algorithms in these two broad categories.

In order to reach the goal, we can either take all the blocks into account or leave some blocks intact.

During decoding the payload watermark signal tehsis removed subtracted from the watermarked image in the restoration step. An unauthorized party may insert false multimedia files or documents in the multimedia system as a threat to authentication.

(PDF) Reversible Watermarking using the Human Visual system | Mohammad Awrangjeb –

Due to its embedding property it offers high capacity and a constant PSNR of Thesid to operating in spatial domain these algorithms are generally simple. The images are taken from very large distances. The second proposed reversible watermarking algorithm section 4.


From the table we see the PSNR becomes very low — average The residual is then compressed to have the capacity to insert the payload and hash, h.

Any type of simple change could give inappropriate knowledge to researchers. Extract watermark payload from I1 2. For even quantization factor the embedding is trivially invertible, but for odd quantization factor a flag should be included to the message to be embedded to reconstruct the original JPEG revedsible during verification. A variety of attacks to the watermarked image may result unavailable or reduction in availability of secret data.

In this algorithm, a large amount of extracted features should be compressed and passed as side information for making the reversihle reversible. Removal attacks aim to remove or seriously degrade the watermark data so that a detector fails to detect it.

reversible watermarking thesis

There are a lot of cryptographic protocols that provide entity authenticity. Finally, thanks to NUS for its financial assistance without which it would have been impossible for me to continue study in abroad.

In addition, I am grateful to all the staffs, specially Ms. But image quality is greatly reduced to This is based on lossless compression of biased bit-streams derived from the quantized JPEG coefficients.

Such information can be embedded into the watermarked image as a part of the watermark payload and transferred secretly. The later is further categorized as spatial domain and frequency domain reversible data hiding.


It is also called private or oblivious watermarking [11,13].

reversible watermarking thesis

Smooth images are more sensitive to noise than textured images. In this method, wateermarking capacity is limited by the frequency of the most frequent gray-level in the image. We restore the original pixel values by replacing the LSBs with extracted original features. In addition, acceptable visual quality is another central requirement.

The authors propose an improvement by shortening the JBIG header. The watermark information can be encrypted using a secret key in order to prevent active attacks. This is due to extracting high compressible features. Generic watermarking system 3 Figure 1.

Watermarked image is the image found after embedding payload to the cover work. VLSI circuit image 22 Figure 2. It is actually incorporating the effect of luminance into JND value, since the distortion in an image is more noticeable in the mid-gray region and sensitivity changes parabolically as the gray value fluctuates on the both sides of mid- gray level.

The residual is then compressed to have the capacity to insert the payload and hash, h.

reversible watermarking thesis

If I w’ is authentic, we continue the process and replace Y with X. Therefore, the embedding capacity is quite high.