DHT domain digital watermarking with low loss in image informations

Abstract

The paper proposes a digital image watermarking scheme that selects regions for data embedding based on information measure. Two valued kernels of Hadamard transformation cause smaller image information change during embedding compared to other transform domains such as DCT (discrete cosine transform), DFT (discrete Fourier transform), Fourier-Mellin and wavelet-based embedding. Moreover, the usage of Hadamard transform as signal decomposition tool offers advantages in terms of simpler implementation, low computation cost and high resiliency at low quality compression considering both JPEG and JPEG 2000 framework. Compression resiliency is further improved using adaptive negative modulation. These facts are validated comparing the performance with some other existing watermarking schemes as well as DCT domain implementation of the proposed scheme.

Copyright Protection of Color Imaging Using Robust-Encoded Watermarking

“…While several methods have been proposed to watermark grayscale images, only a few have been specifically designed for color images [8]. In the recent years, several color image watermarking methods have been proposed in the literature, these methods could be classified as those which are based on the frequency transform domain [10], pixel modification in the spatial domain [11], histogram modification [12] and image normalization [13]. Almost all the methods mentioned above have shown watermark robustness to geometrical distortion; however they cannot provide enough robustness against common signal processing, such as filtering, noise contamination and image compression neither some aggressive combinations of geometric attacks and common signal processing operations.…”

Increase of stability at jpeg compression of the digital watermarks embedded in still images

“…Поиск блоков 8×8, энтропия (E) которых больше заданного первого порога E 1 . Энтропия определяется по формуле [7] 1 exp(1 )…”

“…Такие алгоритмы осуществляют внедрение ЦВЗ в частотные коэффициенты дискретного косинусного преобразования (ДКП) [2,3] и дискретного вейвлет-преобразования [4][5][6]. В последние годы в частотных алгоритмах цифрового маркирования применяется дискретное преобразование Адамара (ДПА) [7][8][9][10][11][12][13][14], которое для дополнительного уменьшения вычислительной сложности определяется строчно-столбцовым образом:…”

A modular framework for color image watermarking

“…The idea driving the use of this module is to reduce the average number of required changes to the coefficients, dependent on the covering radius of the used code. Presently, we experimented the Golay code [24,12,8] (obtained from the Golay code [23,11,7] by adding a parity bit), the Hamming code [7,4,3], the Hadamard code [32,6,16], the BCH code [31,11,11] and the BCH code [31,6,15], where the first two elements of every triple [ , − , ] are related to the number of coefficients r to be used to convey q (syndrome) bits of the watermark. In this context it is also possible to use combinations of the previous codes to store in every block a number of bits not directly possible with a single code: for example, in order to store 15 bits, one can use either the Golay code [24,12,8] along with the Hamming code [7,4,3] applied to 24 + 7 = 31 coefficients, or 5 times the Hamming code [7,4,3] applied to 7 × 5 = 35 coefficients.…”

“…In the same vein as [49], Kougianos et al [25] have investigated several DCT-DWT domain dual (robust-fragile) watermarking methods which are embedded in a hardware processor in order to achieve low power usage, real-time performance, reliability, and ease of integration with existing consumer electronic devices. Other common transforms such as the Fourier transform [41], [42], the fractional Fourier Transform [38], the Hadamard transform [31] or the Schur transform [39] have been used in the cited robust watermarking schemes to lower the computation cost and improve the authentication assessment of the watermark during common image processing attacks.…”