Fig 6 - uploaded by Scott David Dexter
Content may be subject to copyright.
Detector response of the watermark embedding with JND adjustment is plotted against JPEG compression with increasing quality factor, along with the detection threshold and the maximum response among 999 fake watermarks.
Source publication
A robust image watermarking scheme in curvelet domain is proposed. The curvelet transform directly takes edges as the basic representation element; it provides optimally sparse representations of objects along edges. The image is partitioned into blocks and curvelet transform is applied to those blocks with strong edges. The watermark consists of a...
Context in source publication
Context 1
... have tested our proposed watermarking scheme on ten different grayscale images (Lena, Baboon, Boat, Airplane, Goldhill, etc.) of dimension 512 ×512. They have been partitioned into blocks of n’×n’ pixels with n’ =64, thus obtaining 64 blocks. We will only give the results for the standard image “Lena”. The blocks with edge strength higher than 100 (T 1 ) are selected for watermark embedding. We choose all the significant coefficients in scale 2, the embedding strength is limited by JND threshold. A watermark W composed of m= M×N (6144×49) elements is added to curvelet coefficients. The original “Lena”, the watermarked “Lena” with PSNR=39.73 and the absolute difference between them are shown in Fig. 4. The fidelity of the image is well maintained because the watermark energy is compacted into noise- unperceivable blocks with strong edges and the introduced distortion is below the just noticeable distortion described in previous section. Fig. 5. shows only the embedded watermark yields a high response far above the threshold and the responds of fake watermarks are all below the threshold. The parameter k in Equation (23) is determined to be 5.2 to satisfy the requirement of false positive less than 10 -7 . We have investigated the robustness of our watermarking system. Take the attack of JPEG compression as example, 1000 watermarks were tested in our experiments. In Fig 6., the response of the detector of the embedded watermark is plotted against JPEG compression with increasing quality factor of the image (from 5% to 90%), along with the detection threshold and the max response of 999 fake watermarks. The detection threshold is obtained by Equation (23) when the factor k =5.2. Thus the probability of false positive P (z>T z | the image is not marked with W) is also less than 10 -7 . Observation shows the embedded watermark survives the severe image compression when the quality factor is 10% and up. Table 1. reports the performance of the robustness of our watermarking system against a wide range of image attacks including JPEG compression, Gaussian noise addition, cropping, histogram equalization, contrast adjustment, low pass filtering, Gaussian blur, Gamma correction and sharpening. The synchronization step in watermarking detection is extremely important because our scheme requires accurate identification of the embedding blocks. Fig. 7. shows a cropped “Lena” rotated by 30o and scaled down by scaling factor 0.75. We compute the radon transform over the edge map of the corrupted “Lena” and obtain the main axis at θ ’ =152o and max( ρ ’ ) = 135. The vector E* is constructed based on Equation (6) where n=8. Referring to reference parameters, we have θ =122o and max( ρ ) =181. The distortion is evaluated and the correctness of the estimation is confirmed by δ (0.89)>T (0.85) defined in Equation (18). We inverted the corrupted watermarked image back to its original state ...
Citations
... The blind watermarking problem in our proposed model is that the optimal coefficients in DCurT are in most extreme security level, compared with DWT, DST, RDWT, SVD, and some different strategies. 11,43,44 For instance, deviation in the work of Thanki et al 29 is 0.66%, and it is compared with a proposed model closer to 0.30% deviation, and at that point, second execution compared with our work is those by Mankar et al 45 and Bajaj,46 it is 0.92% heartiness of blind watermarking. The proposed plan not only beats more than the computational time of quick existing watermarking methods yet additionally vaguer than watermarking schemes based on DST and DCurT with RGO approaches. ...
Watermarking scheme is not helpful in conveying the original image continually so as identifying the owner's mark from the watermarked image. In the current situation, our chosen topic is significant in real‐time applications like fraud recognition, emergency clinic, and government divisions. In this study, an optimal multiblind watermarking model is proposed for the watermark detection process. Our proposed model is a combination of intelligent domain transforms like Discrete Shearlet Transform and Discrete Curvelet Transform (DCurT). The imperceptibility necessity of the plan is accomplished utilizing optimal coefficients that are performed by applying in DCurT with metaheuristic optimization model that is Grasshopper Algorithm. It is played by a chasing behavior of gathering of grasshoppers and cooperation process, here, Random Grasshopper Optimization is utilized for watermarking. The secret image is embedded with a lower band of optimal coefficients with DCurT, and here, the band is DCur (1, 5). The secret data is embedded in the host images to make it secure, and then, the extraction of the embedding process takes place inversely. For experimentation analysis, 20 digital images are considered, and different attacks are applied in the proposed watermarking model. Thus, the watermarked image looks lossless compared with the host image. Moreover, recent literary works and domain transform are also utilized for strength investigation of the proposed watermarking model. Highlights An optimal multiblind watermarking model is proposed. A combination of intelligent domain transforms is introduced. Random Grasshopper Optimization is utilized for watermarking. The presented model outperforms the state of art methods.
... Tao et al. [22] proposed a method for embedding watermarks into the curvelet coefficients using the spread spectrum [9]. The method was capable of blind detection and robust to signal distortion. ...
... Figure 8 shows typical example images. We then compared the proposed method with Tao's [22], Zebbiche's [25], Zhang's [26], and Nguyen's [17] blind curvelet domain watermarking techniques. We also compared it with Makbol's method [14], a watermarking technique that uses discrete wavelet transformsingular value decomposition. ...
Watermarking inserts invisible data into content to protect copyright. The embedded information provides proof of authorship and facilitates the tracking of illegal distribution, etc. Current robust watermarking techniques have been proposed to preserve inserted copyright information from various attacks, such as content modification and watermark removal attacks. However, since the watermark is inserted in the form of noise, there is an inevitable effect of reducing content visual quality. In general, most robust watermarking techniques tend to have a greater effect on quality, and content creators and users are often reluctant to insert watermarks. Thus, there is a demand for a watermark that maintains maximum image quality, even if the watermark performance is slightly inferior. Therefore, we propose a watermarking technique that maximizes invisibility while maintaining sufficient robustness and data capacity to be applied in real situations. The proposed method minimizes watermarking energy by adopting curvelet domain multi-directional decomposition to maximize invisibility and maximizes robustness against signal processing attacks with a watermarking pattern suitable for curvelet transformation. The method is also robust against geometric attacks by employing the watermark detection method utilizing curvelet characteristics. The proposed method showed very good results of a 57.65 dB peak signal-to-noise ratio in fidelity tests, and the mean opinion score showed that images treated with the proposed method were hardly distinguishable from the originals. The proposed technique also showed good robustness against signal processing and geometric attacks compared with existing techniques.
... Tao et al. [4] proposed a method to embedding watermarks into the curvelet coefficients using the spread spectrum [5]. The method was capable of blind detection and was robust to signal distortion. ...
... This is the inverse step of Step 2 in Algorithm 1. 3: Obtain correlations for all pairs and find the pair with highest correlation. 4: Rotate the image using information from that pair. For example, if the pair found in step 3 is (3,11), inverse rotate the image by 360°/n s ×2. ...
... Figure 8 shows some typical example images. We then compared the proposed method with Tao's [4] and Zhang's [2] blind curvelet domain watermarking techniques. ...
Watermarking inserts invisible data into content to protect copyright. The embedded information provides proof of authorship and facilitates tracking illegal distribution, etc. Current robust watermarking techniques have been proposed to preserve inserted copyright information from various attacks, such as content modification and watermark removal attack. However, since the watermark is inserted in the form of noise, there is an inevitable effect of reducing content visual quality. In general, more robust watermarking techniques tend to have larger effect on the quality, and content creators and users are often reluctant to insert watermarks. Thus, there is a demand for a watermark that maintains maximum image quality, even if the watermark performance is slightly inferior. Therefore, we propose a watermarking technique that maximizes invisibility while maintaining sufficient robustness and data capacity enough to be applied for real situations. The proposed method minimizes watermarking energy by adopting curvelet domain multi-directional decomposition to maximize invisibility, and maximizes robustness against signal processing attack by watermarking pattern suitable for curvelet transformation. The method is also robust against geometric attack by employing watermark detection method utilizing curvelet characteristics. The proposed method showed very good results of 57.65 dB peak signal-to-noise ratio in fidelity tests, and mean opinion score showed that images treated with the proposed method were hardly distinguishable from the originals. The proposed technique also showed good robustness against signal processing and geometric attacks compared with existing techniques.
... In 1999, Candes and Donoho [10] introduced a new multi-scale transform called the curvelet transform which can use a few samples to represent edges and other singularities along curves much more efficiently than traditional transforms like wavelet [11]. After this, certain watermarking schemes based on curvelet domain have been proposed [12,13]. ...
... According to [29], the embedding model for spreadspectrum-based algorithms can be expressed in Equation (13). ...
In this article, a robust blind watermarking scheme using wave atoms with multiple descriptions is proposed. In the presented scheme, the watermark image is embedded in the wave atom transform domain. One of the sub-bands is used to carry watermark data. The experimental results indicate that superiority of the proposed method against common attacks such as JPEG compression, Gaussian noise addition, median filtering, salt and pepper noise, etc., compared with the existing watermarking schemes using multi-scale transformations.
... In order to obtain better imperceptibility, Human Visual System (HVS) is normally added to the watermarking algorithm and many schemes related to HVS were proposed [12][13][14][15][16]. Based on the characteristic of curvelet transform, human visual system model based on the curvelet transform can be developed by utilizing the frequency and orientation properties. Several papers using HVS model based on the curvelet transform were also proposed [17][18]. As curvelet Transform can decompose the image into several bands, we will make use of this property to embed the watermark within those bands and develop a suitable HVS model. ...
In this paper, six robust non-blind watermarking schemes based on curvelet transform are proposed. Single band watermarking method was proposed in Ref. 1. This paper develops the single band watermarking method and adds Human Vision System (HVS) to form six different multi-bands watermarking methods. With the increasing redundancy of watermark, the robustness of the algorithm will be investigated and comparative studies with the single band watermarking will be shown. The experimental results demonstrate that the proposed algorithms have great robustness against various imaging attacks.
... In order to obtain better imperceptibility, Human Visual System (HVS) is normally added to the watermarking algorithm and many schemes related to HVS were proposed [12][13][14][15][16]. Based on the characteristic of curvelet transform, human visual system model based on the curvelet transform can be developed by utilizing the frequency and orientation properties. Several papers using HVS model based on the curvelet transform were also proposed [17][18]. As curvelet Transform can decompose the image into several bands, we will make use of this property to embed the watermark within those bands and develop a suitable HVS model. ...
... In order to obtain better imperceptibility, Human Visual System (HVS) is normally added to the watermarking algorithm and many schemes related to HVS were proposed [12][13][14][15][16]. Based on the characteristic of curvelet transform, human visual system model based on the curvelet transform can be developed by utilizing the frequency and orientation properties. Several papers using HVS model based on the curvelet transform were also proposed [17][18]. As curvelet Transform can decompose the image into several bands, we will make use of this property to embed the watermark within those bands and develop a suitable HVS model. ...
In this paper, a robust image watermarking scheme based on curvelet transform is proposed. Considering the frequency and orientational sensitivity of the human visual system(HVS), we present a HVS model in curvelet domain,which is used to control the embedding strength. The watermark is embedded to the selected coefficients of the curvelet transform, and original image is not required in the extracting process. The experimental results demonstrate that the proposed algorithm can provide excellent robustness against most image processing methods.
... In order to obtain better imperceptibility, Human Visual System (HVS) is normally added to the watermarking algorithm and many schemes related to HVS were proposed [12][13][14][15][16]. Based on the characteristic of curvelet transform, human visual system model based on the curvelet transform can be developed by utilizing the frequency and orientation properties. Several papers using HVS model based on the curvelet transform were also proposed [17][18]. As curvelet Transform can decompose the image into several bands, we will make use of this property to embed the watermark within those bands and develop a suitable HVS model. ...
In this paper, a digital watermarking algorithm used in the curvelet domain is proposed. Wrapping of specially selected Fourier samples is employed to implement Fast Discrete Curvelet Transforms (FDCT) to transform the digital image to the curvelet domain. Based on the capacity of the watermark and the impact of the variation of the curvelet coefficients, the proper positions to embed the watermark were chosen. The experimental results show that watermark is robust to most of the signal processing operations, including Common Image Processing Attacks like Histogram Equalization, Brighter Filter, Darker Filter, Increase Contrast, Decrease Contrast, Median Filtering, JPEG Compression attack, Gaussian Noise attack and Laplacian Filtering.
In this paper, a hybrid and blind watermarking scheme is proposed for the protection of copyright of digital images. The scheme based on hybridization of two advance transforms such that discrete curvelet transform (DCuT) and redundant discrete wavelet transform (RDWT). The motivation behind using these two transforms combination to improve the imperceptibility of the watermarking scheme. The impercep- tibility requirement of the scheme is achieved using hybrid coefficients which are achieved by applying single level RDWT to the high frequency curvelet coefficients of the cover image. The watermark infor- mation is inserted by modifying the coefficients of the wavelet coefficients of LH subband using PN se- quences according to watermark bit and gain factor. The security of the proposed scheme is achieved by applying Arnold scrambling to watermark image before embedding. Experiments of the proposed scheme are conducted on various types of natural images. Experiments results show that, compared with existing schemes, the proposed scheme is robust to various attacks while having high imperceptibility. Also, the proposed scheme is performed better than many existing schemes.
