56 results about "Watermark robustness" patented technology

The invention discloses a digital image watermark detection method based on support vector machine correction, which comprises the following steps of: extracting the phase information of an image in a high-frequency sub-band after the to-be-detected image is subjected to plural pyramid transformation, calculating the energies of Gaussian-Hermite moments on the four directions of a second level of high-frequency sub-band and taking the energies as training features; and training a sample in combination with the minimum square support vector machine theory to obtain a training model, thereby carrying out geometric correction. As the plural pyramid transformation and the support vector machine theory are combined, the digital watermark robustness can be effectively improved through the method, and meanwhile, the imperceptibility and the amount of information of a digital watermark are ensured. The digital image watermark detection method has the characteristics of simple calculation and no original carrier required in the watermark extraction, and the practicability of the watermark for digital image copyright protection is enhanced.

The invention discloses a method for embedding and extracting multiple zero watermarks of a digital image. Multiple zero watermarks are embedded and extracted in a discrete wavelet transform and discrete cosine transform composite domain. The multiple zero watermarks are embedded by constructing a binary watermark secret key by using a coefficient in an original digital image composite domain and then carrying out XOR operation with multiple actual binary digital watermarks, and the multiple zero watermarks are extracted by constructing a binary watermark private key by using a coefficient in a to-be-tested digital image composite domain and combining with zero watermark information. The method has the advantages of excellent watermark robustness and capability of resisting conventional image processing attack; and meanwhile, the method for embedding has the advantages that a plurality of binary digital watermarks are registered in a copyright database without any changing an original digital image, therefore, the embedded binary digital watermarks are not completely detected, the contradiction between the robustness and the difficulty in detection of the digital watermarks is balanced, and multiple requirements of application of the digital watermarks are met.

The invention discloses an H.264 / AVC compressed-domain video watermark embedding and extracting method. Binaryzation is conducted on watermark images, and binary watermark information is obtained through Arnold transformation; compressed encoding is conducted on original video by using an H.264 / AVC encoder, 16 4*4 luminance blocks of each macro block are extracted out according to the H.264 / AVC video encoding standard, and integer discrete cosine transformation is conducted on each 4*4 luminance block; according to the coefficient of each 4*4 luminance block obtained after integer discrete cosine transformation, each 4*4 luminance block is divided into a texture block body and a non-texture block body; according to the binary watermark information, the local energy difference of each macro block is modulated; entropy encoding is conducted on the 4*4 luminance blocks obtained after binary watermarks are embedded to obtain video streaming with the watermarks. According to the H.264 / AVC compressed-domain video watermark embedding and extracting method, video information (texture and energy) is fully considered, and the robustness and invisibility of the watermarks are improved; operational complexity is low, no complex parameter choice exists, and higher real-time performance is achieved; the original video is not needed in the watermark extraction, and the H.264 / AVC compressed-domain video watermark embedding and extracting method belongs to a blind watermarking method.

The invention relates to a watermark embedding and extraction method and device based on a dynamic BP nerve network; the method comprises the following steps: using improved Arnold transformation to scramble an original watermark image; embedding the scrambled watermark image into a dynamic BP nerve network hidden layer, and training to obtain a carrier image with an embedded watermark; the method combines the improved Arnold transformation with the dynamic BP nerve network, thus obtaining the carrier image with the watermarks; even after various attacks, the pixel sizes in the embedded positions can be still extracted, thus correctly detecting a watermark signal; in addition, the method can balance the contradictions between the image watermark robustness and imperceptibility; in addition, in a watermark image extraction process, i.e., the reverse process of the watermark embedding process, the watermark image can be finally obtained through the combination between the dynamic BP nerve network and the improved Arnold transformation.

The invention relates to a vector point spatial data full-blind watermarking method based on grid dividing. The transform domain watermarking algorithms have the characteristic of being high in robustness. At present, most transform domain watermarking algorithms specific to linear vector space data and planer vector space data are characterized in that a watermark is embedded with a geometric object as a unit, but geometric objects of point spatial data are all independent points, and the situation that the transform domain watermarking algorithms are directly applied to the vector point spatial data watermarking can be difficultly achieved. If the whole of the point spatial data serves as a geometric object and a watermark is embedded in the geometric object, extraction of the watermark can be influenced by adding and deleting spatial points. The vector point spatial data full-blind watermarking method based on grid dividing achieves the watermark embedding and watermark extraction aims of the vector point spatial data under transform domains. The vector point spatial data full-blind watermarking method is advanced and scientific, the watermark is high in robustness and good in nonvisibility, and meanwhile the data accuracy after the watermark information is embedded can be guaranteed. An experiment shows that the vector point spatial data full-blind watermarking method has good robustness on attacks such as adding and deleting point operations, cutting, data object sequence scrambling and geometric transformation operations of the vector point spatial data, and has the good use value.

The invention provides a self-adaptive audio blind watermark method based on an auditory model. Arnold transformation, dimension reduction processing and chaotic transformation are performed on a watermark signal so that watermark information is obtained; a human ear auditory masking threshold is calculated; a watermark embedding segment and an embedding position are self-adaptively selected; embedding of a digital watermark is performed; audio segment recombination is performed, and all the audio segments without embedding of the watermark and the audio segments with embedding of the watermark are recombined so that audio signals containing all the watermarks are formed; and the digital watermark is extracted. The watermark information of large energy is embedded under the premise that the watermark signal does not influence the audio quality so that self-adaptive selection of the watermark embedding segment, the watermark embedding position and the watermark embedding strength can be realized, the imperceptibility of the algorithm can be guaranteed and the watermark robustness and the watermark embedding capacity can be enhanced. The human ear auditory characteristic can be utilized in a more detailed way so that the method has better hiding performance, the security of the watermark can be further guaranteed and the attack of common signal processing can be effectively resisted.

The invention discloses a digital image watermark method based on a visual attention mechanism. The digital image watermark method comprises the following steps of: firstly partitioning an image remarkable area by a visual attention mechanism module, obtaining a high-frequency area relative to the remarkable area by wavelet transform, continuously carrying out watermark embedding on dithering quantification modulation of coefficients of the high-frequency area, and dynamically regulating quantified step length depending on a JND (just noticeable difference) visual module in a quantification modulation process. Through the digital image watermark method based on the visual attention mechanism, disclosed by the invention, the step length is dynamically quantified and the watermark is embedded by combining the characteristics of human vision; when the robustness of the watermark is stabilized, the imperceptibility of the watermark is greatly balanced; and simultaneously, the influence of an image background area subjected to operations, such as filtering, noise pollution and the like, in the correctness of a watermark extracting result is further reduced to a certain extent by the introduction of a vision attention mechanism module; and the robustness of watermark detection is further improved.

The invention provides a digital watermarking algorithm for the enhancement of the robustness of a watermarking system, wherein the algorithm comprises the steps of coding specific marking information as a quick response matrix code (QR (Quick Response) code), as a digital watermark, and making use of discrete wavelet transform to embed the QR code into an intermediate frequency coefficient after secondary wavelet decomposition of a carrier image. Therefore, under the premise of assurance of the invisibility of the watermark, the enhancement of the watermark robustness is achieved. The simulation experimental result of a watermark attack indicates that as the digital watermark, the QR code is capable of largely enhancing the robustness of the watermarking system. The digital watermarking algorithm can be widely used in the fields of copyright protection, anti-counterfeiting and the like, and has certain practical value.

The invention discloses an image digital watermark embedding system and method based on an SVM (support vector machine) and DWT (discrete watermark transform). According to the method, first, watermark evaluation parameters are set, including similarity NC for measuring watermark robustness and a peak signal-to-noise ratio (PSNR) for measuring watermark invisibility; a weight beta between the two parameters is determined according to a demand; an SVM method is utilized to perform machine learning on a training set, and a regression prediction model about watermark embedding strength is established to predict optimal watermark embedding strength of a to-be-embedded original image; and last, DWT is adopted to embed a digital watermark into the original image.

Methods and apparatuses for embedding watermark information into multimedia contents like audio, imagery or video, and for detection of such watermarks from embedded multimedia content. A two step embedding approach is proposed wherein the watermark bits are first used to construct a modulation signal. Said modulation signal is used to modify certain selected characteristics of the media content. A detector is then used to first extract the modulation signal from embedded media content, and then to extract watermark information from the modulation signal extracted. This current invention enhances watermark robustness against classes of attacks that prior arts are vulnerable to resist. Current invention is novel and useful in any multimedia application fields, including but not limited to the identification, distribution tracking and copyright protection of audio, imagery and video contents.

The invention discloses a DWT-SVD-ICA-based digital audio watermarking algorithm. The algorithm is performed according to the following steps: the first step, preprocessing is performed on an original audio signal, and a watermark is embedded in a selected appropriate segment; the second step, when the watermark is embedded, three-level wavelet conversion is performed on the original audio signal, singular value decomposition is performed on an approximate component of the wavelet conversion, and then watermark embedment is completed by utilizing a random mixing method; and the third step, when the watermark is extracted, wavelet conversion and singular value decomposition are performed on the audio signal embedded with the watermark, and then watermark blind detection is achieved by utilizing a negentropy maximization fixed point algorithm. Compared with a traditional ICA-based audio watermarking algorithm, the DWT-SVD-ICA-based digital audio watermarking algorithm is advantageous in that watermark robustness is improved to some extent, that is to say, the watermark robustness is high; and imperceptibility of the watermark can be guaranteed, and blind extraction of the watermark can also be guaranteed.

The invention is a robust digital watermark embedding and detecting method based on the supported vector machine, which belongs to multi-media information security area. The invention consists of watermark embedding and watermark detecting. In the process of watermark embedding, a randomly generated referencing watermark is embedded to host media data together with real watermark; in the process of watermark detecting, the well trained supported vector machine can detect watermark with embedded watermark key. The supported vector machine is trained to study the attack with natural image characteristics, referencing watermark data and embedded watermark position data.

It is an adaptive expanding transform jitter modulation watermarking method. The expanding transform jitter modulation watermarking has the advantages both quantitative index modulation watermarking and spread spectrum communication system, and it has a good performance against attack. However, the original STDM method uses the fixed quantization step, which takes no account of human perceptual model, so its performance is greatly limited. The invention introduces a human perception model / signal quality evaluation method to the original STDM, used to measure the distortion introduced by watermarking and compute a weighted sequence, and the use of weighted plastic watermarking idea changes host signals, to make the local variable capacity of the host signal apply for the human perception system; re-designing the quantization step to meet the distortion conditions determined by un-perceptibility. Using the method described in the invention, the embedded watermark can fully use the local characteristics of carrier signal, adaptive design the quantization step, greatly enhance the watermark robustness, and at the same time achieve a good watermark un-perceptibility.

The invention discloses a robust image watermark detecting method based on fuzzy classification. The robust image watermark detecting method comprises the following steps: in a watermark-containing image correcting process, extracting an eight-circumcircle lower PHT matrix (low 7-order), obtained by circumcircle lower PHT decomposition, of a to-be-detected image, taking the eight-circumcircle lower PHT matrix as a training characteristic to combine with a fuzzy support vector machine theory, training a sample to obtain a training module for performing geometric correction. The fuzzy vector machine FSVM has very good learning capacity and the matrix obtained by PHT decomposition has high precision, so that watermark information under different attacks can be correctly extracted, and the watermark robustness is improved. Moreover, the method has the characteristics of simple calculation, no need of an initial carrier during watermark extraction, and the like, so that the copyright protection practicability of digital image works is strengthened.

The invention provides a geometrical-attack-resistant video zero-watermark generating method and a geometrical-attack-resistant video zero-watermark generating device. The method or the device comprises the steps or units of a step or unit for extracting a key frame image in a video through a pixel frame averaging method; a step or unit for processing the key frame image, performing singular value decomposition on a low-frequency sub-band of the key frame image after processing and obtaining a singular value decomposition coefficient after decomposition; a step or unit for encrypting a watermark image and generating a watermark crytograph; and a step or unit for performing XOR processing on the watermark crytograph and the singular value decomposition coefficient, forming a zero watermark and registering the water mark into a third-party database. The geometrical-attack-resistant video zero-watermark generating method and the geometrical-attack-resistant video zero-watermark generating device have advantages of realizing no direct digital watermark embedding in a video, generating the watermark in the database, improving watermark robustness of the video, effectively resisting geometrical attacks such as translation, rotation and zooming, balancing contradiction between watermark robustness and video definition in embedding a traditional digital watermark, and realizing no modification for any video data.

The invention provides a scheme self-adaptive digital watermark embedding and extracting method based on the camouflage technology. The method comprises the main steps that multiple sets of subschemes of relevant operations having both the camouflage property and the identity property are built; an index number is set for each set of subschemes, and an initial scheme sequence is obtained by arraying the subschemes in an ascending order; the BKDR Hash value of secret key information is calculated, and is used for carrying out sorting operation on the initial subscheme sequence to obtain a self-adaptive scheme sequence; according to a self-adaptive scheme sequence table, watermark information is embedded in sequence in a circulation mode. The scheme self-adaptive digital watermark embedding and extracting method has the advantages that the tamper localization function is achieved, the purpose of improving the safety of formatted text document watermark robustness is achieved by means of the camouflage technology, and the problem that watermark information is likely to be illegally extracted and then cracked due to the method of embedding watermarks by adjusting formats of formatted text document watermarks is solved.

The invention discloses a dual zero watermark method of a vector map, wherein the dual zero watermark method comprises: extracting feature points and non-feature points from the vector map to obtain the feature points and the non-feature points of the vector map; after a series of data mining processes is performed on the feature points, obtaining an angle value sequence of a triangle composed ofthe feature points; after a series of data mining processes is performed on the non-feature points, obtaining a binary table of the non-feature points; obtaining a watermark sequence after a series ofprocesses is performed on an initial watermark; performing XOR operations on the watermark sequence and the angle value sequence and on the watermark sequence and the binary table separately, and then performing graphical and merging process to obtain the dual-watermark vector map. In the embodiment of the present invention, the vector map uses dual watermarks so as to solve the poor watermark robustness in case of feature point attack. The present invention also provides a dual zero watermark system of the vector map, a storage medium and a server.

The invention relates to a vector geographic space data total blindness watermarking method based on error reduction. Through the method, the problem that errors in discrete fourier transformation (DFT) domain watermarking are hard to control is solved. The vector space data usually conducts geometric transformation in use. Due to the fact that the DFT itself has the characteristics of invariance of translation, rotation, zooming and the like, a watermarking algorithm based on DFT has large advantages on resistance of geometric transformation. Watermarks can be embedded in a DFT range coefficient and a phase coefficient. If the watermarks are embedded in the transformed coefficient directly, large effects can be caused on original data, and especially small change of the phase coefficient can cause large change of the original data. By amplifying the transformation coefficient, then adding the watermarks and reducing the transformation coefficient, effects of the watermarks on the original data can be greatly reduced. The method is advanced and scientific, watermarking robustness is high, invisibility is good, and meanwhile precision of data with embedded watermark information can be guaranteed.

The invention provides a method and a system for achieving zero-watermark of a video based on wavelet transform and Radon transform. The method comprises the steps of extracting key frames in the video, and ensuring that similarities between the key frames are minimal; transforming digital images signals in the key frames into a frequency domain from a time domain by using the discrete wavelet transform to decompose into high-frequency signals and low-frequency signals, filtering out the high-frequency signals, performing the Radon transform on the low-frequency signals after the low-frequency signals are transformed into the time domain from the frequency domain, and computing a characteristic matrix and a characteristic value according to a matrix generated by the Radon transform; encrypting a watermark image to generate a watermark cipher text; and implementing XOR processing on the watermark cipher text and the computed characteristic value, and thus forming the zero-watermark and registering the zero-watermark to an intellectual property information database. According to the method, a speed problem of an existing video zero-watermark construction algorithm can be overcome on the basis of not modifying the original video, and geometric attacks such as rotating, scaling and translating are effectively resisted while watermark robustness of the video is ensured.

The invention discloses a color image watermark detection method based on invariant-moment geometric correction. According to the method, in a process of correcting an image containing a watermark, amplitude values of eight low-order quaternion exponential moments of an image to be detected are calculated and are used as training features, and the least square support vector machine theory is combined for training the sample to obtain a training model, so that the geometric correction is carried out, the quaternion exponential moments and the support vector machine theory are combined, a system has the advantages that the digital watermark robustness can be effectively improved, and meanwhile, the watermark information quantity and the imperceptibility of the digital watermark are ensured. In addition, the method has the characteristics of simplicity in calculation and the like, and the practicability of the method for being used for digital image mark copyright protection is enhanced.

The invention provides a digital watermarking method based on 2DPCA (two-dimensional principal component analysis), belonging to the field of digital watermarking. The method comprises the following steps of: pre-processing an image: performing various non-linear geometrical transform to the primary image to obtain a new image after each transform, extracting an image characteristic vector matrix and an image principal component matrix; watermarking pre-processing; watermarking embedding; recovering the image; and watermarking extracting. As the characteristic vector matrix which is used for extracting the image principal component matrix is unchanged, even though each valve of the image principal component matrix is changed, the fluctuation range of the data in the image principal component matrix is approximately similar; i leading out principal components are selected from the principal component matrix of the primary image; namely, weight values of i different principal components representing image information are accumulated, and the accumulation sum is within the range of 0.4-0.9; and the i leading out principal components are subjected from discrete cosine transform or Fourier transform. The invention, under the precondition of ensuring robustness and invisibility of the digital watermarking, effectively resists attack from the non-linear geometry.

The invention discloses a vector data watermark anti-counterfeiting method based on geographical semantics support. The method comprises the steps as follows: creating a fidelity description knowledge rule base of vector geographical data based on geographical semantics, creating a watermark carrier model according to the fidelity description knowledge rule base, embedding watermark information based on a typical watermark algorithm, extracting the watermark information, and checking the watermark information. According to the method, the watermark information is embedded into an application-sensitive data characteristic set based on the fidelity description knowledge rule base of the geographical data, so that the watermark robustness is effectively improved; by the typical watermark algorithm, a watermark embedding model taking high fidelity as a solving target guarantees the usability and the invisibility of watermark version data and improves the practicality of the algorithm; through dimensionless treatment of a watermark carrier, a geometric attack is invalid for the watermark model; and a method for adopting repeated embedding in a watermark embedding process and electing in a detecting process effectively improves the resistance of the watermark model to a common watermark attack.

The invention discloses a self-encoding color image robust watermark processing method based on visual perception, which comprises the following steps of: training a watermark embedding module in an AGWNet by using a carrier image training set and an original watermark image training set to obtain structure loss and visual perception loss of a watermark-containing image and a carrier image; inputting the watermark-containing image into a noise attack module in the AGWNet to generate an attacked watermark image; inputting the attacked watermark image into a watermark extraction module in the AGWNet for training to obtain cross entropy loss of the extracted watermark and the original watermark image; calculating the overall loss according to the structure loss, the visual perception loss and the cross entropy loss, and obtaining a trained AGWNet when the overall loss is smaller than a threshold value in the iteration of the overall loss; and embedding or extracting the watermark by using the trained AGWNet. According to the method, the watermark robustness can be improved while the quality of the watermark-containing image is ensured.

The invention discloses a natural language watermark robustness testing method which comprises the following steps of: reading parameters required for system operation; based on a watermark embedding algorithm to be tested, embedding watermark information into an original carrier text to generate a watermark text; based on active attack methods, performing an active attack on the watermark text to generate an attacked text; based on a watermark extraction algorithm to be tested, extracting new watermark information from the attacked text; and calculating an error bit rate. According to the invention, different active attack methods are obtained according to an adversarial model of the watermark algorithm, the watermark text is attacked based on the active attack methods, the new watermark information is then extracted from the attacked watermark text, and the error bit rate is finally obtained. The experiment proves that the method disclosed by the invention can effectively and objectively evaluate the robustness of a natural language watermark algorithm to be tested.

Digital audio signal watermarking in real-time is difficult in an environment that has limited processing power. According to the invention, the channels in a data block-based audio multi-channel signal are prioritised with respect to watermarking importance, whereby the channel priority can change for different input signal data blocks. For a current input signal block, the most important channel is watermarked and the required processing time is determined. If this required processing time is shorter than a predefined application-dependent threshold, the next most important channel is marked and the additionally required processing time is determined, and so on. Due to the block-based nature of the audio watermarking including block overlap / add and due to the sensitivity of the resulting audio quality against blocking artefacts, several problems are solved in order to lead to acceptable performance and quality. The invention optimises the trade-off between watermark robustness and security on one hand and real-time processing constraint on the other hand.

The invention discloses a novel digital watermarking method capable of maintaining fidelity and robustness. An embedded algorithm serves as main content, and at first, an original image is segmented into subblocks which are distributed in eight rows and eight columns and are not mutually covered; then discrete cosine transformation is carried out on each subblock; finally, maximization of integrity of watermark information extracted from figures which contain watermarks and are attacked serves as a goal, and under the constraint that the PSNR is larger than 25dB, the watermark weighted value of each subblock is optimized to obtain the optimal solution which serves as the final watermark embedding scheme. On the condition of guaranteeing fidelity of digital watermarking, watermark robustness is optimized by means of an evolutionary optimization algorithm, the fidelity and robustness are taken into full consideration, digital watermarking quality is improved, degradation of images formed after the watermarks are embedded is difficult to perceive compared with that of the original image, and good robustness is achieved against attacks such as compression, filtering and zooming rotational shearing.

According to a method provided by the invention, a watermark is embedded through discrete Fourier transform based on features such as video resolution and a code rate. The algorithm complexity is low.The watermark robustness, the video fidelity and the watermark capacity of a video are balanced. The watermark embedding speed and extraction speed are high.

The invention discloses a spatial domain color digital image blind watermarking method fusing discrete Fourier transform, and the method combines the advantages of high operation speed of a spatial domain digital watermarking algorithm and high robustness of a frequency domain digital watermarking algorithm. According to the unique characteristics of the discrete Fourier transform direct current components, the correlation principle between the direct current components of the adjacent pixel blocks is utilized, different quantization step lengths are adopted in different channels, the embedding and blind extraction of digital watermarks are completed in the airspace, and real discrete Fourier transform is not needed. According to the invention, a color digital watermark image is embedded into a color carrier image; on the premise that the watermark invisibility is good, the watermark robustness and the algorithm real-time performance are high, the problems that a large-capacity color image digital watermark algorithm is low in operation speed and poor in robustness are solved, and the method is suitable for occasions where digital media copyright protection is conducted rapidly andefficiently.

The invention discloses an image digital watermark embedded strength regression prediction model modeling method. The method herein includes the following steps: firstly setting watermark evaluation parameters which include a similarity NC which is intended for balancing watermark robustness and a peak signal-to-noise ratio which is intended for balancing watermark concealment, and based on requirements, determining a weight [beta] between two parameters, and using the SVM method to conduct machine learning on a training set. The method herein establishes a regression prediction model related to watermark embedded strength, can predict optimal watermark embedded strength of a to-be-embedded original image, and eventually uses DWT conversion to embed digital watermark to the original image.

The invention discloses a self-adapting watermark embedding method, based on local quality evaluation, belonging to the digital watermark technique, while present watermark technique will cause distortion, the embedded watermark technique adjusts the strength of water mark according to one global quality index, without considering the local character of signal, and for confirming local signal quality, the global watermark strength will be reduced to decrease robustness. The invention blocked evaluates the quality of signal, and self-adapting adjusts the embedded strength of watermark in each block according to the local distortion condition. With said invention, the embedded watermark can fully utilize the local character of signal, to improve the robustness of watermark and attain better local unknown ability.