Data Hiding in Motion Vectors of Compressed Video

7/30/2019 Data Hiding in Motion Vectors of Compressed Video

1/26

Data Hiding in Motion Vect

Of Compressed Video

Guided By:

Prof. D R. Bhamare

Presented By:

1) Harshit A. Pa

2) Priyanka Pat

3) Sonali


2/26

Steganography

Hiding a Secret data in Digital Images without affecting thequality of Medium.

This image is called as StegoImage.

Steganography, Cryptography &Watermarking all are sameconcept only but some small differences should be there.

Original Image Stego image


3/26

Steganography Vs Cryptogra

In both concepts we are going to transmit some secret informone Cover Medium.

In steganography we are going to hide secret message in imchanging the quality of original, So no one understand someconfidential information in this image like that

But in Cryptography method we are going to change the dato Cipher text.

So any one can understand some secret process is going likecant retrieve.


4/26

Goal Of Steganography

1) The main goal of steganography is to communicate securely

completely undetectable manner and to avoid drawing susp

transmission of a hidden data.

2) It is not to keep others from knowing the hidden information, b

others from thinking that the information even exists.

Accuracy

Consistency

Minimize false-positives


5/26

Example

George obtains oranges daily yet eights are rubbish

Good year !

Eorge btains ranges:- cover objeGeorge obtains oranges daily..-st

Good year ! :-embedded messag


6/26

STEGANOGRAPHY

Technical

steganography

Cryptography

Linguistic

Steganography

Security

systems

Information

hiding

Steganography Wate

Audio TextVideoDigital images

Robust

Imperceptible Visible


7/26

Objectives

Optimize the high hiding capacity & imperceptibility

Design the Optimum Pixel Adjustment algorithm to hide the d

minimized embedding error

Hiding the data in spectral domain

Avoid the Overflow & Underflow of embedding data


8/26

BLOCK DIAGRAM


9/26

FRAME SEPERATION


10/26

Frame processing is the first step in the background subtrathe purpose of this step is to prepare the modified vremoving noise and unwanted objects in the frame in ordeamount of information gained from the frame and the algorithm.

Preprocessing is a process of collecting simple image procchange the raw input video info a format. This can bsubsequent steps. Preprocessing of the video is necessarydetection of moving objects.

Another key issue in preprocessing is the data format used background subtraction algorithm. Most of the algluminance intensity, which is one scalar value per each

color image, in either RGB or HSV color space, is becomingthe background subtraction algorithms.


11/26

Proposed Methods

LSB Embedding Algorithm

LSB Embedding in Spatial Domain

LSB Embedding in Transform Domain

Embedding data, which is to be hidden, into an image requires two files.

The first is an innocent looking image that will hold the hidden informati

cover image.

The second file is the messagethe information to be hidden

To data insertion.


12/26

LSB Embedding Algorithm

Image Pixel Value = (1,1) = 241 & (1,2) = 140Secret Data A (ASCII) value = 65

241 = 11110001, 140 = 10001100 65 = 01000001

LSB Initialization

In 11110001

1111MSB 0001LSB

To initialize this 4 bits to 0000 we have to do bit and operationA secret data we have to do bit oroperation.


13/26

Continued. So using this algorithm we can hide half the secret data bits in fir

pixel next half of the bits we can hide in second Pixel.

In extraction Part we have to do same thing in reverse process.

Ex:

To make it last 4 bits zero bitand with 240

Now (1,1) = 1111 0000, & (1,2) = 1000 0000

secret data 65 = 01000001

After Embedding (1,1) = 1111 0100 & (1,2) = 1000 0001


14/26

LSB Embedding in Spatial Doma

The basic idea in LSB is the direct replacement of LSBs of nois

bits of the cover image with the secret message bits.

Till now LSB is the most preferred technique used for data hid

is simple to implement offers high hiding capacity, and provid

way to control stego-image quality.

It has low robustness to modifications made to the stego-imag

pass filtering and compression and also low imperceptibility.


15/26

LSB Embedding in Transform Dom

The secret message is embedded into the high frequencyof the wavelet transform while leaving the low frequency

sub band unaltered.

In an adaptive (varying) hiding capacity function is emplo

determine how many bits of the secret message is to be e

each of the wavelet coefficients.

The advantages of transform domain techniques over spa

techniques are their high ability to tolerate noises and som

processing operations but on the other hand they are com

complex and hence slower


16/26

Drawbacks of Transform domain

Low hiding capacity and Complex computations

Embedding error will be there(Overflow & Underflow)


17/26

ADVANTAGES OF LSB INSERTION

1) A major advantage of the LSB algorithm is it is quick and easy

2) There has also been steganography software developed wh

around LSB color alterations via palette manipulation.

3) LSB insertion also works well with gray-scale images.


18/26

Types of Compression

Lossless Compression

Data can be completely recovered after decompression

Recovered data is equal to original

Applications: Medical imagery,

Lossy Compression

Data cannot be completely recovered after decompression

Some information is lost for ever

Compression ratio is more than lossless comp

Applications: commercial distribution (DVD)


19/26

GUI


20/26

Performance Analysis

Performance measurement for image distortion is well known as peak sig(PSNR). To establish an objective criterion for digital image quality, a par

PSNR (Peak Signal to Noise Ratio) is defined as follows

The performance in terms of capacity and PSNR (in dB) is defined as per

follows,

2

1010log (255 / )PSNR MSE


21/26

Result Overview

While Reconstructing the frames and make it video PSNR is more than 40

gives more better image quality and security as per need.


22/26

Advantages

It is used in the way of hiding not the Information but the reach that Information.

Difficult to detect .Only receiver can detect.

Can be applied differently in digital image, audio & video file

It can be done faster with the large no. of software's.


23/26

Disadvantages

Huge number of data equal to huge file size, so some onabout it.

In sending and receiving information can be leaked.

The confidentiality of information is maintained by the algoritalgorithms are known then its all over.

If this technique is gone in the wrong hands like hackers, terrothen this can be very much dangerous for all.


24/26

Applications

1. Teleconference or video phone

Very low delay (1/10 second is a standard)

2. Live Broad cast Video

Modest delay is tolerable(second is normal)

Error tolerance is needed

3. Video-in-a-can (DVD, Video on Demand)

Random access to compressed data is desired

Encoding can take a lot time

4. Decoding must always be at least the frame rate.


25/26

Conclusion

Steganography is not implemented in wider ways but it casecurity tool.

The system combines an secret data hiding technique usinpixel adjustment to increase the hiding capacity of the systto other systems.


26/26

Data Hiding in Motion Vectors of Compressed Video

Documents