第1个回答 2008-12-12
In order to improve the transparency of the color image watermark and robustness, this combination of human visual masking properties, given the digital watermark embedded method to determine the adaptive position (i.e. the wavelet coefficient method in selecting). Specific procedures as follows:
(1) the wavelet coefficients F, including the brightness factor F wavelet decomposition level said fernandez, F says high direction, Fernandez said the wavelet coefficients position.
(2) choosing coefficient of visual important for the watermark is embedded. Brightness factor for keyword, F within all of the wavelet coefficients were descending, choose a sort F, before the wavelet coefficients as digital watermark embedded position.
2.5 adaptive quantization interval
Understanding is not interested in a basic requirements of digital watermark algorithm, thus watermark characteristics and the sensitivity to adapt. The characteristics of human visual system to determine the embedding strength can not visible in the watermark based on effectively improve the robustness. In order to make the adaptive quantization modulation in image, we according to the selection of wavelet coefficients of mean and variance s to determine d, quantitative intervals. A selection of wavelet coefficients is the value of collection and F, F for automatic weighing parameters and F F. O F the minimum and maximum F, can use F F by wavelet coefficients quantifying sets, for spacing F here F choice of wavelet coefficients of quantity and F, F is the user setting maximum and minimum quantization interval. Mean and variance F F was used to adjust to quantify the watermark interval embedded in the image to be better. Parameters controlling F quantitative weighing background, F used to control the quantitative weighing structure.
第2个回答 2008-12-12
In order to improve the color watermark image of transparency and robustness of this article fully integrated human visual masking properties, gives the location of the digital watermark embedded in the adaptive method (that is, select the method of wavelet coefficients). Concrete steps are as follows:
(1) Wavelet coefficients calculated brightness of the F-F, which said the F-level wavelet decomposition, F, said high-frequency sub-band direction; F said that the location of the wavelet coefficients.
(2) Select an important factor for visual watermark embedding. F brightness factor for keywords, all sub-band of wavelet coefficients in descending order of selection before the F-wavelet coefficients as a digital watermark embedding location.
Adaptive quantization intervals of 2.5 to determine.
Can not be perceived requirements of digital watermarking is an essential characteristics of the algorithm must therefore Watermark and the sensitivity of the human eye to adapt. Use of the human visual system to determine the characteristics of the embedded watermark strength can ensure the invisible watermark on the basis of effectively improve its robustness. In order to quantify the adaptive modulation carrier in the image, we choose according to the wavelet-coefficient d and the average variance to determine the quantitative s interval. Has been set up to choose the wavelet coefficients of the right to set the value, of which F and F, respectively for the F and F weighing parameters. To derive the maximum value of F and F minimum F, can be the use of F-wavelet coefficients quantify the interval F, where F for the alternative set the number of wavelet coefficients, F and F are the user to set the maximum and minimum Quantify the interval. F and F mean variance was used to adjust the spacing to allow quantification of the embedded watermark to better adapt to the carrier in the image. F parameters used to control the context of the quantitative weight, F used to quantify the weight of the control structure.
第3个回答 2008-12-12
For transparency and Shandong stick sex improving the colour watermark image, the main body of a book shelter characteristic property combining with optesthesia of person eye sufficiently, self fit in with having given a figure out watermark implantation location ascertains method (being that the minor wave modulus chooses method). Concrete step as follows: (1) calculates small wave modulus F brightness factor F , F among them indicate that the minor wave breaks up arrangement of ideas , F expresses the direction that high frequency son brings along; F expresses minor wave modulus position. The important modulus of (2) choice optesthesia is used for watermark implantation. Brightness factor F is keywords , the simultaneous inner minor wave modulus carries out a descending sort on all son , the F minor wave modulus choosing front implants location as figure watermark. That ascertaining forbidden perception is that the figure watermark demands a basic characteristic property of 2.5 certainly fit in with quantizations gap, the watermark algorithm must fit in with each other with sensibility of person eye as a result. Make use of the implantation intensity coming to ascertain a watermark characteristic of human being optesthesia system to be able to improve whose Shandong stick sex on invisible guarantee watermark basis effectively. Come we to ascertain the quantization gap for making quantization modulate the certainly fit in with Yu carrier image, mean value d and method difference s small wave modulus collection according to choice. Set up right value minor wave modulus collection quilt choosing, F and F among them are that F and the F weighing are parametric respectively. Seek out F maximum value F and minimum value F , can make use of F to get quantization gap F that the minor wave modulus gathers right away, here, F is to provide the quantity that the alternative minor wave modulus gathers , that F and F part for is the maximal and minimal quantization gap that the consumer sets up. Method difference F and mean value F are applied to adjust the quantization gap to make the embedded watermark adapt to the Yu carrier image much better. Parameter F is used to control the background quantization weighing , F is used to control architectural quantization weighing.