You are on page 1of 1

The scope of intercepting, copying and redistributing multimedia in its exact original form, has broadened with

the exponential growth of network technology and World Wide Web [1]. Therefore, digital theft of intellectual
property and copyright protection have become a prime concern in the owner's mind. Thus, Digital Watermarking
techniques have been proposed as an effective method to build secure way for legal distribution of data [2].
Digital Watermarking is the process of inserting certain authentication data (ownership details, logo etc.) into a
multimedia file without affecting its perceptual quality i.e. invisibly [3]. In the case of a copyright infringement,
this authentication data (watermark) can then be extracted from the multimedia and used as an authoritative proof
of one's ownership. This helps in identifying piracy, preventing alteration and overall preserving integrity of the
data.
Here we are discussing watermarking of video files rather than individual images. Digital video
watermarking is a technique which is utilized to ensure the verification and approval rights for the video files. In
this process, a binary image is embedded into the video frames as a watermark so as to protect the ownership of
original video file [4]. Video watermarking is more challenging and time consuming due to frame redundancy,
vulnerability to attacks such as frame dropping and temporal redundancy [5].
Two vital properties of a watermark are imperceptibility and robustness [6]. Imperceptible watermarks
are those which are visually redundant i.e. cannot be seen by the naked eye, however, they can be recovered from
the multimedia file using the original watermarking algorithm. Robustness refers to the resistance of watermark
against a variety of intentional (printing, scanning, noise etc.) and non-intentional (compression, filtering,
cropping, rotation etc.) attacks. Imperceptibility is assured by evaluating optimal scaling factor (SF) and
appropriate embedding position. Robustness is achieved by ______[7]. The main challenge in watermarking is to
combine both imperceptibility and robustness. Therefore this makes watermarking problem belong to the class of
optimization problems [8]. Thus, here Cuckoo Search, a metaheuristic algorithm has been used for this purpose.
Video Watermarking techniques can be broadly divided into two major domains- frequency domain and
spatial domain. In the Spatial domain, watermarking is performed by directly modifying each image pixel's
intensity value in a video frame [9]. In the Frequency domain, pixel values are converted into transform domain
coefficients using a predetermined function (DCT, DWT, DFT etc.) and then they are modified to embed a
watermark in imperceptive frequency areas of the image [10]. While the spatial domain is computationally easier
to implement, it is more susceptible to distortions, less secure and overall creates fragile watermarks.
Comparatively, frequency domain creates more robust and imperceptible watermark since it diffuses the
watermark continuously in the image [11]. Frequently used transform are Discrete Fourier Transform (DFT),
Discrete Cosine Transform (DCT) and Discrete Wavelet Transform (DWT). DWT is superior in creating more
robust watermark over other two methods. DWT is more computationally efficient, it easily identifies areas in
original video frame where the watermark can be inserted imperceptibly and finally, it is compatible with the
Human Visual System (HVS) [12]. In this paper, we have utilized 2-level DWT (2D- DWT) which decomposes
the video frame into four sub-bands, the lower resolution component (LL) and other detail components LH, HL
and HH. This is illustrated in the Fig 1.

You might also like