Pranab Kumar Dhar

Robust FFT Based Watermarking Scheme for Copyright Protection of Digital Audio Data

Digital watermarking has drawn extensive interest as a means of protecting digital contents from unauthorized copying. This paper proposes a robust audio watermarking scheme based on the Fast Fourier Transformation (FFT) for copyright protection of digital audio data. The watermarking scheme transforms the original audio into the frequency domain. The magnitude spectrum of the original audio is divided into an arbitrary number of segments and the energy of each segment is calculated. Watermarks are then embedded into most prominent peaks of the highest energy segment of the original audio. Watermarks are extracted by performing the inverse of the watermark embedding process. Simulation results indicate that our proposed watermarking scheme resists various attacks such as noise addition, cropping, re-sampling, re-quantization, and MP3 compression. Moreover, it outperforms Coxs method in terms of imperceptibility, while keeping comparable robustness with the Coxs method. Its SNR (signal-to-noise ratio) values range from 20 dB to 31 dB, in contrast to Coxs method whose SNRs range from only 11 dB to 23 dB.

A blind audio watermarking method based on lifting wavelet transform and QR decomposition

This paper presents a blind audio watermarking method based on lifting wavelet transform (LWT) and QR decomposition (QRD) for audio copyright protection. In our proposed method, initially the original audio is segmented into non-overlapping frames. Watermark information is embedded into the largest element of the upper triangular matrix obtained from the low frequency LWT coefficients of each frame. A blind watermark detection technique is introduced to identify the embedded watermark under various attacks. Simulation results indicate that the proposed watermarking method is highly robust against different attacks. In addition, it has high data payload and provides good imperceptible watermarked sounds. Moreover, it shows superior performance than the state-of-theart watermarking methods reported recently.

An Efficient Audio Watermarking Algorithm in Frequency Domain for Copyright Protection

Digital Watermarking plays an important role for copyright protection of multimedia data. This paper proposes a new watermarking system in frequency domain for copyright protection of digital audio. In our proposed watermarking system, the original audio is segmented into non-overlapping frames. Watermarks are then embedded into the selected prominent peaks in the magnitude spectrum of each frame. Watermarks are extracted by performing the inverse operation of watermark embedding process. Simulation results indicate that the proposed watermarking system is highly robust against various kinds of attacks such as noise addition, cropping, re-sampling, re-quantization, MP3 compression, and low-pass filtering. Our proposed watermarking system outperforms Cox’s method in terms of imperceptibility, while keeping comparable robustness with the Cox’s method. Our proposed system achieves SNR (signal-to-noise ratio) values ranging from 20 dB to 28 dB, in contrast to Cox’s method which achieves SNR values ranging from only 14 dB to 23 dB.

A semi-fragile watermarking method using slant transform and LU decomposition for image authentication

Semi-fragile watermarking has been widely used for content authentication. In this paper, we propose a semi-fragile watermarking method using slant transform (ST) and LU decomposition (LUD) for image authentication. Initially, the watermark image is scrambled using an algorithm to enhance the confidentiality of the image. The host image is separated into red, green, and blue channels. The red channel is divided into 8×8 non-overlapping blocks. ST is applied to each block to get the ST coefficients. LU decomposition is then applied to these coefficients to obtain the lower and upper triangular matrix. Watermark data is embedded into the upper triangular matrix of ST coefficients using an embedding equation. Simulation results indicate that proposed method produces high quality watermarked image, provides secure embedding, extraction and authentication. Moreover, it effectively detects malicious tempering on image while being resistant to various content preserving manipulations. Furthermore, it shows superior performance than the state-of-the-art methods in terms of imperceptibility.

Enhanced Edge Localization and Gradient Directional Masking for Moving Object Detection

Moving object detection has been widely used in intelligent video surveillance system. This paper proposes a new moving object detection method based on enhanced edge localization mechanism and gradient directional masking. In our proposed method, initially gradient map images are generated from the input image and the background image using gradient operator. The gradient difference map is then calculated from gradient map images. Finally, the moving object is extracted by using appropriate directional masking and thresholding. Simulation results indicate that the proposed method outperforms conventional edge based methods under different illumination conditions including indoor, outdoor, and foggy cases to detect moving object. In addition, it is computationally faster and applicable for real-time processing.

A modified New Reno for performance enhancement of TCP in wireless network

Transmission Control Protocol (TCP) is the dominant transport protocol in the Internet and supports many of the most popular Internet applications, such as the World Wide Web (www), file transfer and e-mail. TCP assumes congestion which is the primary cause of packet loss and uses congestion control mechanisms such as Tahoe, Reno and New Reno to overcome this congestion in wireless network. These TCP variants take longer time to detect and recover packet loss. In order to improve retransmission scheme, we propose a modified version of New Reno that shows better performance than previous TCP variants because of utilizing faster retransmission scheme as well as transferring more packets to the destination.

Baleen Whale Sound Synthesis using a Modified Spectral Modeling

Spectral modeling synthesis (SMS) has been used as a powerful tool for musical sound modeling. This technique considers a sound as a combination of a deterministic plus a stochastic component. The deterministic component is represented by the series of sinusoids that are described by amplitude, frequency, and phase functions and the stochastic component is represented by a series of magnitude spectrum envelopes that functions as a time varying filter excited by white noise. These representations make it possible for a synthesized sound to attain all the perceptual characteristics of the original sound. However, sometimes considerable phase variations occur in the deterministic component by using the conventional SMS for the complex sound such as whale sounds when the partial frequencies in successive frames differ. This is because it utilizes the calculated phase to synthesize deterministic component of the sound. As a result, it does not provide a good spectrum matching between original and synthesized spectrum in higher frequency region. To overcome this problem, we propose a modified SMS that provides good spectrum matching of original and synthesized sound by calculating complex residual spectrum in frequency domain and utilizing original phase information to synthesize the deterministic component of the sound. Analysis and simulation results for synthesizing whale sounds suggest that the proposed method is comparable to the conventional SMS in both time and frequency domain. However, the proposed method outperforms the SMS in better spectrum matching.

Implementation and performance analysis of real-time digital filter for audio signals

A real-time digital filter can perform filtering operation on real-time signals. This paper presents the implementation and performance analysis of real-time FIR (finite impulse response) digital filter on audio signals. Digital filters (low pass, high pass and band pass) implemented using DSP block sets are used to remove the higher and lower frequency components of real-time voice signals. This paper evaluates the performances of these digital filters in terms of the filtered output and the frequency response curves. Experimental results indicate that a maximum number of ripples appear in stop band for Kaiser window, while a minimum number of ripples appear in triangular window for different filters. Thus, real-time FIR digital filter using triangular window gives the optimal result to process the audio signals.

Bangla text compression based on modified Lempel-Ziv-Welch algorithm

Text compression algorithm performs compression at the character level. Bangla text has some unique features such as no distinct upper and lower case letter, consonant cluster (CC) and consonant with dependent vowel sign (CV) etc. The conventional Lempel-Ziv-Welch (LZW) algorithm is not suitable for compressing Bangle text. Therefore, in this paper, we propose a modified LZW (MLZW) algorithm which can compress Bangla text effectively and efficiently. In our proposed method, a dictionary with Unicode ranges from 1–90 is used for Bangla characters. The compression process is started with checking the input character. If input character is a part of CC or CV, then CC or CV is considered as a character and search it in the dictionary. If the character to be encoded is already in dictionary, encode it with the dictionary index. Otherwise, the character is added to the dictionary and is encoded with its corresponding dictionary index. Simulation results indicate that the proposed MLZW algorithm compresses Bangla text effectively and efficiently. We observed that the proposed MLZW provides higher compression rate approximately 3% for dictionary index and 33% for output sequence compared with LZW algorithm.

Image watermarking in LWT domain based on nonnegative matrix factorization and singular value decomposition

We present a blind watermarking method in lifting wavelet transform (LWT) domain based on nonnegative matrix factorization (NMF) and singular value decomposition (SVD) for image copyright protection. The watermark image is preprocessed first using a Gaussian map in order to enhance the confidentiality. LWT is then applied to the original image to get sub-bands and the low frequency sub-band is divided into blocks. NMF is performed on each of these blocks to get the nonnegative matrix and weight matrix. Weight matix of each block is decomposed into three matrices using SVD. The largest singular value of each weight matrix is selected for embedding watermark using a quantization function. Simulation results indicate that the proposed method shows high robustness against different attacks. Moreover, it outperforms state-of-the-art methods in terms of invisibility and robustness.