This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
With todays modern technology-driven world, multimedia have brought about tremendous changes in our social life. Now, multimedia puts business, communication, education, socialization, virtual reality, entertainments at our fingertips. Human beings are completely surrounded by multimedia environments and consequently multimedia has created new weighty security challenges and great innovative opportunities.
This report reveals the aspects regarding the data hiding through multimedia elements like audio signals, video signals, text and images keeping mind in both practical and theoretical approaches. Digital information hiding is the new way of secret communication technology. One of the fundamental approaches of data hiding in multimedia is steganography and the modern approach is Digital watermarking.
A steganography is the technique of hiding the messages using such a way that no one else can interpret the content of the message except the sender and the recipient. Steganography has an origin from the Greek word "Steganos" which means covered or protected message and is still in use. There are plenty of techniques that are familiar to us ranging from invisible ink to pin punctures. The use of steganography has been passed for generations but the main applications were military and intelligence purposes.
The digital watermarking plays a very essential role in identifying the ownership of the copyrights of the embedded signals such as audio signals or image data. A digital watermarking is nothing but a process of hiding computer aided informative data in a carrier signal. A digital watermark is a way of marking covertly which is added into another noise tolerant signal. The main applications of digital watermarking are to verify the authorization or integrity of the carrier signal and to identify the owner of the signal.
Table of contents
1. Abstractâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦.â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 1
2. Introduction â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 3
2.1 Classification of multimedia â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 3
3. Literature Reviewâ€¦â€¦â€¦â€¦â€¦â€¦..â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦. 4
4. Research Sectionâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦. 5
4.1 Data hiding in Imagesâ€¦â€¦â€¦â€¦.â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦.. 6
4.2 Steganography in JPEG imagesâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 6
4.3 Data hiding in Videoâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦. 8
4.4 Data hiding in Audio â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 9
5. Conclusion and Recommendationâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦. 10
6. Referencesâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦.. 10
List of figures
The basic framework of data hiding in multimediaâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦5
2.1 Block diagram of embedding and extraction process in binary imagesâ€¦â€¦â€¦â€¦.6
2.2 (a) Test Imageâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦ 7
2.2 (b) Stego imageâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦. 8
3.1 Block diagram of proposed video data hiding systemâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦.8
3.2 Strategies for handling frame jitteringâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦.9
4.1 Audio signal Embedding processâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦.9
The term Multimedia is nothing new to the human beings. Since the existence of human beings multimedia has always around us. Every person, no matter in which age, has somewhere involved in multimedia. People ranging from all ages has enjoyed multimedia elements such as presentations, Animations, cartoon films and nowadays, video tape, compact disks, digital video disc, blue ray and many more since their childhood days. An earlier person used to deal with Analogue-multimedia and now as the technology is changing at the light speed now they start entertaining their selves by the creative and innovative digital-technology. The main merit of Analogue multimedia over the digital multimedia is that digital media allows editing, combining, distributing of informative data which is beyond the limitation of analogue multimedia. The multimedia are sometimes called as "mixed multimedia" or the system that provides the way of communication by using a wide range of media effects. The word multimedia is a combination of two words; multi and media; multi means plural or many and media resembles medium of communication. In Simple words, The Multimedia are nothing but the meaningful combination of text, audios, graphics, videos, images, animation which can be used for the widest range of applications such as for communication entertainments, educations, trainings, businesses, e-commerce, and ultimately for the web applications. Furthermore, the term interactive multimedia means the rich multimedia, which is used to convey the informative data in a very entertaining way.
2.1Classification of Multimedia
Basically, multimedia can be broadly classified on the basis of timeline and origin as:-
Captured media is analogous to the informative data gathered from the real world For instance, still images or dynamic image, different forms of sounds etc. Whereas processed media refers to the media which is synthesized by means of computing devices such as text, graphics and computer animation. On the other hand, Continuous Media resembles to the media which is based on time. Sound rolling images, animation are the best examples of continuous media. Continuous media are also named as dynamic media or time dependant media whilst, the media which includes space and dimension are known as non-linear media. Mostly, Non-linear media is many times referred as non-time based media or non-temporal media. Nevertheless; there are lots of attacks and countermeasures for multimedia. Plenty of applications are designed for hiding multimedia contents like ownership protection, patent ownership, password protection, copy or access restrictions and authentications. The most essential factor is to test the robustness and security of multimedia contents hiding through the attacks. In a broad sense it has more significance than the designing of the multimedia contents.
3. Literature Review
"With the advances of the digital information revolution and the societal changes they have prompted, it has become critical to facilitate the secure management of content usage and delivery across communication networks". (MIN WU and Bede Liu, 2003). The studies projects that users are very much keen to enjoy the conveniences provided by multimedia. In this digital world the lightning fast sharing of digital multimedia data has achieved the mountain peak by means of ubiquitous network environment. Users are always in thirst of finding the cheapest way to share the media information while ignoring the chances of violating copyrights. Consequently, the multimedia designers are compelled to find out the countermeasures to secure the multimedia contents from the intruders. (MIN WU and Bede Liu, 2003).
Steganography is the art and science of communicating in such a way that the very existence of communication is not revealed to a third party. In order to communicate without being detected, the data-hider must obey following two conditions. Perceptual constraint. The perceptual distortion between the original and stego image should not be more than a certain maximum amount, D1, for some perceptual distance measure. Statistical Constraint. The embedding process should not modify the statistics of the host signal more than a very small number, epsilon, for some statistical distance measure. (Ken Sullivan and Kaushal Solanki, 2005). The research elicits that steganography is nothing but the multiple techniques for sharing of confidential informative data, keeping in mind that the information must not be interpreted by someone else who is not authorized to do so. For that one must be very particular about two ground rules while designing the multimedia contents. The first one is the perceptual constraints in which the multimedia designer must be able to identify the distortion between the stage image and original image and the designer must keep in mind that the distortion must not exceed than the particular limit of the perceptual distance. Moreover, the designer must also keep an eye on the statistical constraints which means that the embedded tactics should not change the original signal in all manners. It should be modified in a very typical way for a few statistical distance measures. (Ken Sullivan and Kaushal Solanki, 2005).
Digital audio, video, images, and documents are flying through cyberspace to their respective owners. Unfortunately, along the way, individuals may choose to intervene and take this content for themselves. Digital watermarking and steganography technology greatly reduces the instances of this by limiting or eliminating the ability of third parties to decipher the content that he has taken. (Morgan Kaufmann, 2008). In the era of digital globalization it can be crystal clearly understand that the multimedia elements like audio, video, images, graphics, animation, text can be shared in just a matter of nanoseconds to the respective recipient. But as said by someone, every option has its pros and cons and every coin has its two sides; The Multimedia is not an exception. In the few years the security of such communication becomes the noticeable issues. Nevertheless, the technology like digital watermarking and steganography immerged and the security flaws are resolved in some aspects by restricting the unauthorized parties to decrypt the information data.(Morgan Kaufmann, 2008)
4. Research Section
As both the practical and theoretical facts studied in the above literature review we can figure out the basic framework of data hiding in multimedia as shown in below. Here, the sender sends the original copy of informative multimedia or original media let's say (I0) (It can be audio, images, graphics, video or any form of multimedia elements) (sometimes I0 is also commonly known as covered media or host media.) After that, the embedded module inserts some set of secondary data which is also called as embedded data (b) for the embedding process. The output of the embedded module may be visually identical with the Io but it will contain the hidden data in it (b).Generally, the output of the embedded module is also referred as marked media and noted by (I1). The embedded process introduced a new term Embedding distortion which is nothing but the difference between the marked media (I1) and the original media (I0).
Fig 1.1 Basic framework of data hiding in multimedia (Min Wu, 2001)
Generally, the hidden data (b) is nothing but the collection of bits, which usually depends from where the application is coming. It may be encoded character string, or from the patterns or from some executable agents or from military applications .The most essential factor in most of the cases is to extract the hidden data from the marked media (I1) in bit by bit accurate forms. Sometimes, the modification of hidden data is possible in such applications where the original media and the hidden media are partially identical like image in image or video in video applications. Consequently, in such applications like ownership or copyright protection, exact bit to bit decoding of hidden data from the distorted tested media is needed. Such kind of data hiding is well known as watermarking or robust data hiding.
A Finger printing is also a kind of watermarking which is generally used where various labels are hidden in the identical multimedia content before sharing it to more than one receiver while these hidden labels are used to identify the respective receiver.
Furthermore, the marked media is tested to the different varieties of processing and security attacks before passing it to the detector module. After this testing process the data is considered as a test media which is denoted by (I2). The difference between the tested media (I2) and the marked media (I1) is called as noise and the data extracted from the tested media is known as extracted data which is noted as (b^).
4.1 Data hiding in Images
Basically, For the purpose of data hiding in binary images, there are two ways to modify binary images. First one is by modifying the values of each and every pixel separately and another one is by changing the values of a group of pixels.
In the first technique, we can simply change the black pixel of the image into white pixel and white pixel into black one and in the other technique we changes such attributes of images such as the thickness of the strokes, their positions and the relative curvatures which usually depends on the type of image. Types of images range from text to signatures. In this second approach of data hiding in binary images only the limited amount of informative data can be hidden because of blind detection (without using the original image in detection) and invisibility there are very few attributes in the image that can be modified for the purpose of data hiding.
On the other hand, in first approach an image is split into several blocks and for the purpose of data hiding several bits are modified in each block by changing a few pixels in the respective block. To make this approach crystal clear three ground issues should be kept in mind while using this approach.
The main issue is about selection of pixels for modification so that as possible the image should look visually natural.
Particular means for modification of data in each and every block through flappable pixels.
The purpose of modifying the equal number in each block and to enhance the efficiency.
The following figure 1.2 helps to understand the concept embedding and extraction process.
Fig 2.1 block diagram of embedding and extraction process in binary images (Min Wu, 2001)
The above block diagram shows the step by step process of embedding and the extraction process of binary images for the purpose of authentication and annotation.
4.2 Steganography in JPEG images
Steganography in JPEG (joint photographic experts group) images plays an important role in embedding a secret data into the least significant bits from the counted discrete cosine transform X and Y co-efficient. One of the demerits of this process is that only few informative data can be embedded into the cover image.
In this approach the secret informative data can be embedded into the mid part of the quantized DCT co-efficient frequency. This stegnographic system changes the LSBs in sequence; consequently, it causes distortion which is detected by steganalysis techniques. The histogram of color frequencies changes because of embedding of high entropy data. Another technique is for embedding the informative message bits by means of Pseudo random number generator which selects DCT coefficients randomly.
T-codes belong to the family of variable length codes. We believe that the best variable length codes are the Huffman codes just because they are easy to design moreover it provides a better efficiency if the source statistics are identified. But still there are some limitations for it like, when it is used in serial communication usually, loss of synchronisation occurs in a complex synchronisation process and the length and the result are quite difficult to recognise. To overcome to this problem T-codes are introduced.One of the best merits of T-codes is that they are self-synchronising and due to that the decoder can recover synchronisation automatically in case if few bits are modified or lost in a T-code encoded stream. In addition, best T-codes are able to gain self-synchronisation followed by the lock loss within 1.5 characters. Thus, t-codes replaced the Huffman codes for the jpeg and jsted images algorithm. Moreover, this approach is able to send stenographic messages in lossy pattern which are robust in with respect to detection or attack.
Embedding Algorithm for T-codes (V.Sathya, IEEE conference)
Input: secret message and the cover image
Stepl. Encode the message using the T Codes
Step2. Divide the cover image into 8x8
Step3. Calculate DCT coefficients for
Step4. Quantize the coefficients
Step5. while complete message not
Step5.1 get next DCT coefficient
Step 5.2 if DCT 0 , DCT 1 and
DCT = -1 then
Step 5.2.1 get next bit from
Message Fig 2.2(a) Test Image
Step 5.2.2 replace DCT LSB
with message bit
Step6. De-quantize and take inverse DCT
to obtain stego-image
Output: Stego- image
Extracting algorithm for T-codes (V.Sathya, IEEE conference)
Fig 2.2(b) Stego image
Step 1. Divide the stego image into 8x8 blocks
Step2. Calculate DCT coefficients for each block
Step3. Quantize the coefficients
Step4. while secret message not completed do
4.1 get next DCT coefficient
4.2 If DCT 0 , DCT land
DCT = -1 then
Concatenate DCT LSB to
Step5. Decode secret message bits using
Output: Secret message
4.3 Data hiding in Video
In the multimedia domain, video plays a central role of interests among the people. A video is nothing but the proper sequence of image frames therefore; data hiding video techniques are analogous in some aspects of the data hiding images techniques. One video is considered better than 1000 words. In addition, we can embed more informative data in video as compare to any other multimedia elements. The main requirement of video data hiding is by the video providers like movie producers and News agencies for the purpose of ownership protection, access control (piracy), alteration detection and fingerprinting.
The figure 3.1 below elicits the step by step process of video data hiding system.
Fig 3.1block diagram of proposed video data hiding system (N. Murali, IEEE conference)
As the consecutive video frames viewed identically and each frame can be considered as a standalone unit it is easy to insert or delete some frames or replace the positions of adjacent frames without any observable visual difference except in case of fast moving videos and scene changes. Though, these modifications are the most potential attacks in such robust data hiding systems we can control it by inserting redundancy or searching for frame jitter invariant field.
Furthermore, to manage the frame jittering we introduced two strategies which are demonstrated in following fig. 3.2. Here, in first approach we a video is divided into several segments each of which contains an identical number of consecutive frames and after that embed the same data in every frame of individual segment.
The main advantage of this strategy is that it permits us to tolerate frame dropping that involves a few numbers of separated frames. To receive topmost detection accuracy, redundancy plays an important role to struggle with noise from intruders' attacks and processing. The process of extraction can be accomplished by means of providing more weights to the frames which are having less distortion. Though, the first approach is capable of doing all these but still it is unable to process frame repositioning, frame inserting, or frame deleting from the larger units. To overcome all this listed issues, the other approach has come into existence which simply embeds a short version of segment index in each and every frame. This information is known as frame synch which helps us to detect and locate the frame jittering. This strategy can improve the robustness against the intruder attacks of frame repositioning and frame deleting through redundancy approach.
In simple terms, we use the same image data hiding strategy to each video frame and embed the same user data and the frame synch index in each frame of the similar segments to overcome frame jittering.
Fig 3.2 Strategies for handling frame jittering
4.4 Data hiding in Audio
The audio masking is nothing but the process of data embedding in audio signals which exploits imperfection of the human auditory system. As we know the phenomenon of audio that when the loud audio signal (referred as masker) is audible other weak audio signal is inaudible, this depends on the temporal and spectral features of both the masked signal and masker. The quantisation noise is hidden below the masking threshold in perceptual compression technique whereas; in general data hiding application the embedded signal is hidden.
Fig 4.1 Audio signal Embedding process
The human auditory system is designed in such a way that it can receive over a range of power greater than one billion to one and the range of frequencies greater than one thousand to one.
5. Conclusion and Recommendations
In this paper, we have studied a number of robust methods and techniques in multimedia elements and come to the conclusion that among all the above techniques, data hiding in video has great merits and scope. As far as the video data hiding has been tested, we can hide a large amount of data through this approach and has been found to be used world widely by the movie producers and by the news agencies in order to avoid piracy and to prevent ownership protection from the hackers and to send the data to the destination in a safe manner. In this research we didn't only focus on the weakness of existing data hiding strategies but also we have overviewed what data hiding in multimedia can do and what are the limitations of it .
On the other hand, audio data hiding has very limited scope in future as it has limitations of covert communication and inaccessible information storage.