The Video Watermarking Techniques Computer Science Essay

Published Date: 02 Nov 2017

Gopika Mane#, G. G. Chiddarwar*

#ME-II

Sinhgad College of Engineering

Vadgaon (Bk), Pune. Maharashtra

India

[email protected]

*Assistant Professor

Sinhgad College of Engineering

Vadgaon (Bk), Pune. Maharashtra

India

[email protected]

Abstract— The protection and illegal redistribution of digital media has become an important issue in digital era. This is due to the popularity and accessibility of the Internet now days by people. This results in recording, editing, replication of multimedia content. It is necessary to protect digital information against illegal manipulations and distributions. Digital watermarking technique is process of embedding noise-tolerant signal such as audio or image data in carrier signal. This technique provides a robust solution to the problem of intellectual property rights of online contents. This paper reviews different aspects and techniques of digital watermarking for protecting digital contents.

Keywords— Attacks, Content protection, Digital properties, DWT, SVD, Security, Watermarking techniques.

Introduction

The development and success of the internet has created a new challenge to protect digital content from piracy. Digital watermarking is a new technique that provides an ultimate and robust solution to this problem. It involves direct embe ding of additional information into the original content or host signal without perceptible difference between the watermarked and the original content [1]. Such watermark is extremely difficult to remove without altering or damaging the original signal. Watermarking techniques consist of two parts: a part to embed a signal to the original content and a part to detect if a given piece of signal hosts a watermark and subsequently retrieves the conveyed signal. Digital watermarking is very important because it protects the rights of owners of a signal; it also provides copyright identification, user identification, authenticity determination and automated monitoring.

Multimedia and network security issues are classically handled through cryptography, however, cryptography ensures confidentiality, authenticity, and integrity only when a message is transmitted through a public channel such as an open network. It does not protect against unauthorized copying after the message has been successfully transmitted. Digital watermarking is an effective way to protect copyright of multimedia data even after its transmission.

Video watermarking introduces some issues which is not present in image watermarking. Due to large amounts of data and inherent redundancy between frames, video signals are highly suscepectible to pirate attacks, including frame averaging, frame dropping, frame swapping, statistical analysis, etc [1]. Applying a fixed image watermark to each frame in the video leads to problems of maintaining statistical and perceptual invisibility. Furthermore, such an approach is necessarily video independent; as the watermark is fixed while the frame changes. Applying independent watermarks to each frame also presents a problem. Regions in each video frame with little or no motion remain the same frame after frame. Motionless regions may be statistically compared or averaged to remove independent watermarks [2]. In addition, video watermarking schemes must not use the original video during watermark detection as the video usually is in very large size and it is inconvenient to store it twice.

Watermark embedding scheme can either embed the watermark into the host signal or to a transformed version of the host signal. Transform domain watermarking is a scheme that is used to transform image frequency domain in such a way to modify the transform coefficient. Some common transform domain watermarking for image data can be Discrete Cosine Transform (DCT) based [2, 3] or Discrete Wavelet Transform (DWT) based [4]. This scheme is very useful for taking advantage of perceptual criteria in the embedding process for designing watermark techniques. Spatial domain watermarking on the other hand has the capability of performing some transformation directly on image pixels. The use of perceptual models is also an important component in generating an effective and acceptable watermarking scheme for audio just as it is used in image watermarking [3, 4].

This paper is organized five sections. The subsequent section explains the important aspects of video watermarking. Section II focuses the widespread techniques of video watermarking where various domains of video watermarking are explored and a robust algorithm in each domain is considered.

Aspects of video watermarking

Video sequence is a collection of consecutive and equally time spaced still images. Apparently any image watermarking technique can be extended to watermark videos, but in reality video watermarking techniques need to meet other challenges. Watermarked video sequences are very much susceptible to pirate attacks such as frame averaging, frame swapping, statistical analysis, digital-analog (AD/DA) conversion, and lossy compressions.

Watermarking systems can be characterized by a number of defining properties including embedding effectiveness, fidelity, data payload, blind or informed detection, false positive rate, robustness, security, cipher and watermark keys, modification and multiple watermark, cost, tamper resistance, unobtrusiveness, ready detection, unambiguous, sensitivity, and scalability. Some of them are common to most practical applications. In this section, such general requirements will be listed and briefly discussed and focus will put on video watermarking.

Perceptual Transparency

Invisibility is that degree that an embedded watermark remains unnoticeable when a user views the watermarked contents. However this requirement conflicts with other requirements such as tamper resistance and robustness, especially against lossy compression algorithms. to survive the next generation of compression algorithms, it will probably be necessary for a watermark to be noticeable to a trained observer which is asked to compare the original and the marked version of the video.

Robustness

Robustness is the resilience of an embedded watermark against removal by signal processing. The use of music, images and video signals in digital form, commonly involves many types of distortions, such as lossy compression. For watermarking to be useful, the mark should be detectable even after such distortions occurred. Robustness against signal distortion is better achieved if the watermark is placed in perceptually significant parts of the signal.

Due to large amounts of data and inherent redundancy between frames, video signals are highly susceptible to pirate attacks, including frame averaging, frame dropping, frame swapping, statistical analysis, interpolation etc [1].

Capacity

Capacity is that amount of information that can be expressed by an embedded watermark. Depending on the application at hand, the watermarking algorithm should allow a predefined number of bits to be hidden.

Review on Video Watermarking Techniques

Many digital watermarking schemes have been proposed in the literature for still images and videos. Most of them operate on uncompressed videos [8, 11, 13], while others embed watermarks directly into compressed videos [12, 13]. Video watermarking applications can be grouped as security related like Copy control [8], fingerprinting, ownership identification, authentication, taper resistance etc. or value added applications like legacy system enhancement, database linking, video tagging, digital video broadcast monitoring [8], Media Bridge etc.

A classification map of the existing video watermarking techniques is presented in Figure 1. They can be divided into 3 main groups based on the domain that the watermark is embedded.

Fig. 1 Classification map of existing digital video watermark techniques

Spatial Domain Watermarking

The spatial domain watermarking techniques embed the watermark by modifying the pixel values of the host image/video directly. Low computational complexities and simplicity are main strengths of pixel domain methods. As a result they have proven to be most attractive for video watermarking applications where real-time performance is a primary concern.

1) Least Significant Bit Modification: In this technique, the Least Significant Bit of each pixel is used to embed the watermark or the copyright information. In this technique cover image is used to store the watermark, in which we can embed a smaller object multiple times. An extension to basic LSB substitution would be to use a pseudo-random number generator to determine the pixels to be used for embedding based on a given "seed" or key.

LSB modification proves to be a simple and fairly powerful tool for stenography, however lacks the basic robustness that watermarking applications require.

2) Correlation-Based Techniques [14]: The most straightforward way to add a watermark to an image in the spatial domain is to add a pseudorandom noise pattern to the luminance values of its pixels. A Pseudo-random Noise (PN) pattern W (x, y) is added to the cover image I (x, y), according to the given below:

I w (x, y) = I (x, y) + k *W(x, y)

Where k denotes a gain factor and Iw the resulting watermarked image. Increasing k increases the robustness of the watermark at the expense of the quality of the watermarked image. To retrieve the watermark, the same pseudo-random noise generator algorithm is seeded with the same key, and the correlation between the noise pattern and possibly watermarked image is computed. If the correlation exceeds a certain threshold T, the watermark is detected, and a single bit is set. This method can easily be extended to a multiple-bit watermark by dividing the image into blocks and performing the above procedure independently on each block.

Frequency Domain Watermarking

Most of watermarking techniques [6-8], the watermark will be embedded into the frequency domain instead of the spatial domain for the robustness of the watermarking mechanism. Discrete Cosine Transformation (DCT), Discrete Fourier Transformation (DFT) and Discrete Wavelet Transformation (DWT) are the three main methods of data transformation in this domain. The main strength offered by transform domain techniques is that they can take advantage of special properties of alternate domains to address the limitations of pixel-based methods or to support additional features. Generally, the main drawback of transform domain methods is their higher computational requirement.

1) Discrete Fourier Transform [14]: This approach first extracts the brightness of the to-be-marked frame, computing its full-frame DFT and then taking the magnitude of the coefficients. The watermark is composed of two alphanumerical strings. The DFT coefficient is altered, then IDFT. Only the first frame of each GOP is watermarked, which was composed of twelve frames, leaving the other ones uncorrupted. It is good robustness to the usual image processing as linear/non-linear filtering, sharpening, JPEG compression and resist to geometric transformations as scaling, rotation and cropping.

2) Discrete Cosine Transform [14]: The Discrete Cosine Transform allows an image to be broken up into different frequency bands, making it much easier to embed watermarking information into the middle frequency bands of an image. The middle frequency bands are chosen such that they have minimize they avoid the most visual important parts of the image (low frequencies) without over-exposing themselves to removal through compression and noise attacks (high frequencies).

Fig. 2 Process of Discrete Cosine Transform (DCT)

Following are steps carried out in encoding procedure of DCT:

The image is broken into N*N blocks of pixels.

D (i, j) = C (i) C (j)

p (x,y) is x,yth element of an image represented by matrix p.N is size of block that DCT is done on. The equation calculates one entry (i, j) of the transformed image from pixel values of original image matrix.

Working from left to right, top to bottom, the DCT is applied to each block which is accomplished by matrix multiplication.

Each block’s elements are compressed through quantization means dividing by some specific value.Qunatization is achieved by dividing each element in transformed image matrix by corresponding element in quantization matrix

The array of compressed blocks that constitute the image is stored in a drastically reduced amount of space. It is carried out using zig-zag sequence.

3) Discrete Wavelet Transform: The DWT decomposes input image into four components namely LL, HL, LH and HH where the first letter corresponds to applying either a low pass frequency operation or high pass frequency operation to the rows, and the second letter refers to the filter applied to the columns.

The basic idea in the DWT for a one dimensional signal is the

following. A signal is split into two parts, usually high frequencies and low frequencies. The edge components of the signal are largely confined to the high frequency part. The low frequency part is split again into two parts of high and low frequencies. This process is continued an arbitrary number of times, which is usually determined by the application at hand.

Let

H (ω) =∑ hke−jkω;

k

and G(ω) = ∑ hke−jkω;

k

be a lowpass and a highpass filter, respectively. A signal, x[n] can be decomposed recursively as

Fig. 3 The model of DWT decomposition

cj−1, k = ∑ hn−2kcj, n

n

dj−1, k = ∑ gn−2kcj, n

n

for j = J+1, J,…., J0 where cJ+1,k = x[k], k Є Z, J+1 is the high resolution level index, and J0 is the low resolution level index.

In the encoding part of DWT while watermarking, we first decompose an image into several bands with a pyramid structure as shown in Fig. 3 and then add a pseudo-random sequence (Gaussian noise) to the largest coefficients which are not located in the lowest resolution.

4) Contourlet Transform (CT): DWT offers multistage and time frequency localization of the image. However, it fails to represent the image effectively, if the image contains smooth contours in different directions. CT addresses this problem due to its inherent characteristics, viz., directionality and anisotropy.

Laplacian pyramid and the directional filter Bank is combined in Contourlet transform. Figure 4 shows a multiscale and directional decomposition using a combination of a Laplacian pyramid (LP) and a directional filter bank (DFB). Bandpass images from the LP are fed into a DFB so that directional information can be captured. The scheme can be iterated on the coarse image. The combined result is a double iterated filter bank structure, named contourlet filter bank, which decomposes images into directional subbands at multiple scales.

The result of process described above returns decomposition of input video frames with different frequency bands. Watermark is also decomposed using CT and mapped with low pass coefficients of host video frames. Inverse Contourlet Transform (ICT) is applied to the modified sub bands to get final watermarked video.

Fig. 4 The contourlet filter bank

Watermarks Based on MPEG Coding Structures

Video watermarking techniques that use MPEG-1, -2 and -4 coding structures as primitive components are primarily motivated by the goal of integrating watermarking and compression to reduce overall real-time video processing complexity. Compression in block-based schemes like MPEG-2 is achieved by using forward and bi-directional motion prediction. There are a number of MPEG-2 and -4-based techniques that have been proposed, including approaches based on GOP modification [69], high frequency DCT coefficient manipulation, DCT block classification.

In [5], video object watermarking is based on the structure of MPEG-4. This technique can be easily added to the embedding and detection schemes without changing the watermarking algorithm. It modifies some predefined pairs of quantized DCT coefficient in the luminance block of pseudo-randomly selected MBs. It is based on spread-spectrum techniques. Dividing the image into blocks of equal size, them binary sequence is generated using secret key, and then adds to the image.

In [6], watermarking procedure embed copyright protection into video sequences which is object based transparent. Each watermark is created by shaping an author and video dependent pseudo-random sequence according to the perceptual masking characteristics of the video. As a result, the watermark adapts to each video to ensure invisibility and robustness. Furthermore, the noise like watermark is statistically undetectable to prevent unauthorized removal.

In [7], Mobasseri proposed direct sequence watermarking using m-frames This scheme applies a direct sequence spread spectrum model to the watermarking of the digital video. First, video signal is modeled as a sequence of bit planes arranged along the time axis. Watermarking of this sequence is a two layer operation. A controlling m-sequence first establishes a pseudorandom order in the bit plane stream for later watermarking.

Other Watermarking Techniques

Watermark can be either directly inserted in the raw video data or integrated during encoding process or implemented after compressing the video data. In this section we briefly discuss some of video watermarking techniques present in literature

A novel collusion resistant (CR) video watermarking approach is proposed in [8]. This is a practical frame by frame video watermarking technique. Here a basic s × s watermark pattern is first created and this pattern is repeatedly embedded so that it is centered on a fixed number of selected points known as anchors in every video frame. The part of the video frame where the basic watermark is embedded is called the footprint. Anchor points are calculated using feature extraction algorithm. After generating these watermark frames within a given host frame, spatial masking is applied on it to ensue robustness and imperceptibility criteria. Then the scaled watermark is embedded in the host data using addition.

In Watermarking using CDMA modulation [9] one of the four least significant bitlanes are replaced by watermark planes. The bitlanes to be replaced are selected according to a random periodic quaternary sequence. The watermark plane is generated using 1D spread spectrum methodology.

In [10] a watermarking method using variable length code (VLC) swapping was proposed. This methodology was based on the observation that in the MPEG-2, H.26l VLC tables there are pairs of code words (r, l) → c0 and 1 (r, l ±1) → c1 such that length (c0) = length (c1), lsb (c0) ≠ lsb (c1). Such level-adjacent pairs are called label-carrying VLC (lc-VLC). A covert data bit U i is embedded into a frame by extracting eligible lc-VLC, ci ∈ {c0} ∪ {c1} , and swapping a codeword, if necessary such that to ensure lsb(ci ) =Ui . This process does not use any random key based component as a result of that this method is not robust against attacks.

A perceptual watermarking (PW) method explicitly model masking properties of the HVS and utilizes these models to analyse video sequence or frames to embed watermark in the optimal way. The five main properties of the HSV namely, frequency sensitivity, luminance sensitivity, contrast masking, edge masking and temporal masking can be exploited by video watermarking techniques [11], [12]

conclusion

This study has introduced a number of techniques for the watermarking of digital images, as well as touching on the limitations and possibilities of each. LSB substitution is not a very good candidate for digital watermarking due to its lack of even a minimal level of robustness. LSB embedded watermarks can easily be removed using techniques that do not visually degrade the image to the point of being noticeable. Embedding in the DCT domain proved to be highly resistant to JPEG compression as well as significant amounts of random noise. The wavelet domain as well proved to be highly resistant to both compression and noise, with minimal amounts of visual degradation. The counters proposed to geometric distortion typically rely on discovering the exact rotation, or shifting used in the attack, and then transforming the image back into its pre-attack state. Typically these techniques are computationally pricey, and unpredictable.