Pixel Modification Research Articles

P‐177: A Novel OLED Pixel Circuit for High‐Resolution or High‐Frequency Application

This essay proposed a solution for adverse display qualities caused by insufficient charging in fields of high resolution display or high frequency supports. Verified through experiments, results indicate that the modification of the OLED pixel driving circuit could meet the relatively satisfying demands of frequency up to 240Hz@FHD.

Half Difference Expansion Based Reversible Data Hiding Scheme for Medical Image Forensics.

Medical image authentication is an important area which attempts to establish ownership authentication and data authentication of medical images. In this paper, we propose a new reversible watermarking scheme based on a novel half difference expansion technique for medical image forensics. Conventional difference expansion based reversible watermarking scheme generates watermarked images with less visual quality, and the embedding rate was considerably less due to the high probability of overflow or underflow. In the proposed scheme, the quality of the watermarked image has been improved by modifying the traditional difference expansion based watermarking scheme, half of the difference between two pixels will be expanded during watermarking. The modification of pixels during watermarking is limited by expanding half of the pixel difference, which helps to obtain watermarked images with better visual quality and improved embedding rate due to less chance of overflow or underflow during watermarking. We also propose a tamper detection localization process to detect the tampered regions from the watermarked image. Experimental study of the proposed scheme on the standard medical images from Osrix medical image data set shows that the proposed watermarking scheme outperforms the existing schemes in terms of visual quality of the watermarked image and embedding rate. The overhead related to location map and parity information need to be addressed in future works to improve the proposed scheme.

Adversarial example generation with adaptive gradient search for single and ensemble deep neural network

Deep Neural Networks (DNNs) have achieved remarkable success in specific domains, such as computer vision, audio processing, and natural language processing. However, researches indicate that deep neural networks are facing many security issues (e.g., adversarial attack, information forgery). In the field of image classification, adversarial samples generated by specific adversarial attack strategies can easily fool deep neural classification models into making unreliable predictions. We find that such adversarial attack algorithms induce large-scale pixel modifications in crafted images to maintain the effectiveness of the adversarial attack. Massive pixel modifications change the inherent characteristics of generated examples and cause large image distortion. To address the mentioned issues, we introduce an adaptive gradient-based adversarial attack method named Adaptive Iteration Fast Gradient Method (AI-FGM), which focuses on seeking the input’s preceding gradient and adjusts the accumulation of perturbed entity adaptively for performing adversarial attacks. By maximizing the specific loss for generating adaptive gradient-based entities, AI-FGM calls for several gradient-based operators on the clean input to map crafted sample with the corresponding prediction directly. AI-FGM helps to reduce unnecessary gradient-based entity accumulation when processing adversary by adaptive gradient-based seeking strategy. Experimental results show that AI-FGM outperforms other gradient-based adversarial attackers in attacking deep neural classification models with fewer pixel modifications (AMP is 0.0017 with L2 norm in fooling Inception-v3) and higher success rate of invasion on deep neural classification networks in white-box and black-box attack strategy on public image datasets with different resolution.

Crafting adversarial example with adaptive root mean square gradient on deep neural networks

Deep Neural Networks have achieved remarkable success in computer vision, natural language, and audio tasks. It shows excellent ability in dealing with specific tasks with surpassing efficiency and accuracy. However, researches indicated that deep neural models are extremely vulnerable to crafted adversarial perturbation. In image classification domain, crafted images with adversarial perturbation can fool deep neural models into misclassifying. Specific researches revealed that adversarial examples crafted by attack methods show substantial pixel modification strength, which causes lower similarity between the clean and corresponding sample and makes the change in crafted samples visible. To address the issues mentioned above, we propose an adversarial attack method, which generates adversarial perturbation based on adaptive root mean square gradient strategy. In our proposed approach, we formulate adversarial perturbation based on an adaptive gradient at root mean square level during crafting adversarial sample. Due to the adaptive strategy, the proposed method searches the decision boundary between the original and the adversarial classes in latent space directly by searching the extremum of loss. It helps to generate adversarial samples with higher image quality and better transferability on fooling multiple deep neural models. We evaluate several state-of-the-art attack methods with proposed methods. Experimental results show that our approaches outperform modern techniques in crafting adversarial sample with slight pixel modification, and excellent efficiency in fooling classifiers in both no-targeted and targeted attack strategies.

Anisotropic distortion cost update strategy in spatial image steganography

In current adaptive steganography study, the distortion function is used to describe pixel modification distortion cost. Distortion cost can describe the influence of pixel modification to the cover. And it plays an important actor in adaptive stegaography. Based on previous researches, in this paper, an Anisotropic Distortion cost Update strategy (ADU strategy) is proposed to adaptively update the initial distortion cost (obtained by HILL .etc) to describe the interaction between multiple embedding modifications more accurately. Based on steanographic security theorty, it is proved that clustering modification strategy can improve steganographic security through theoretical derivation. And through study on steganalysis features, it is analyzed and proved that the determination of optimal modification method of the central pixel is anisotropic and much more complicated. Following these two proofs, the image gradient and thermal conductivity are used to quantify the anisotropy of 4 neighboring pixels. Then the optimal modification method of the central pixel is calculated. Finally, the distortion cost of the central pixel can be updated based on the optimal modification method. Experiments with the optimal embedding simulator show that the ADU strategy has better performance when the scaling factor is 2; the proposed ADU strategy can effectively improve the steganography schemes, especially for steganalytic performance with maxSRMd2 features.

JND-Guided Perceptually Color Image Watermarking in Spatial Domain

Watermarking is an effective solution for copyright protection and forensics tracking via hiding information into the image signal. Recently, spatial-embedding watermarking methods from different transform domain have been proposed rapidly and effectively protect the copyright of the color image. Compared to the individual frequency domain method, it has both advantages of spatial domain and frequency domain. Here, we proposed a novel spatial-embedding watermarking method based on an attended just noticeable difference (JND) model with color complexity, to achieve a good tradeoff between robustness and perceptual quality. In particular, at the spatial embedding level, the direct current (DC) coefficients are selected for embedding and the perceptual JND model is used to guide the amount of pixel modification to improve visual quality. Different from the previous JND model, we proposed an attended JND model that considers color complexity, which is more consistent of the human visual perception model. Compared with the other JND models, the proposed JND model is more suitable for the watermarking framework. Experimental results show that the proposed watermarking scheme has better performance than other existing watermarking schemes. This greatly benefits the practical implementations of the spatial-embedding watermarking methods.

Additive Watermarking in Filtered Satellite Images Using Alpha Blending

Development of computer network creates many benefits in digital imaging area which is applicable in satellite imaging too. But this easy access provides virtually unprecedented opportunities to pirate copyrighted satellite images. Maintaining privacy and authenticity in satellite information is an essential requirement but unfortunately it is not possible to do this. It is sometime necessary to identify the copyright information of satellite images. In this case, digital watermarking may be a better tool for this identification. To form digital watermarking, we have to incorporate a special bit pattern into the main information which is useful to identify proper users. In the proposed scheme, a novel method of watermarking application using alpha blending has been discussed. For our research work, we have selected color satellite image which are watermarked with variable values of alpha. Before watermarking, we have filtered the satellite image as well as the cover image using Gauss filtering. As a result of filtering, noise has been eliminated from both the images. The pixel modification has been performed during processing with alpha blending on red, green and blue planes respectively. Using proper value of alpha, original satellite image can be retrieved from the digital watermarked image. We have used conventional quality checking method, PSNR, to analyze the quality of the retrieved image. Applying some post-processing work, this can be used in steganography process of satellite image.

Pengamanan Data Teks Dengan NTRU dan Modulus Function pada Koefisien IHWT Citra Warna

The development of digital information have caused the rise of information technology security to protect text data that contains secret. Steganography is one of many solutions for securing text data by hiding the text data on an image so that another party would not know the existence of such data. Criteria of a good steganography involves imperceptibility, fidelity, robustness dan recovery. One steganographic method is CD (Coefficient Difference), adopted from PVD (Pixel Value Differencing) which does hiding in spatial domain using difference of two pixel values that results in large modification of pixel values, reducing imperceptibility. Modulus function is used to solve such shortcoming in CD by using the modulus function on embedding, minimizing pixel modification during the process, resulting in improved imperceptibility. In this final project, IHWT (Integer Haar Wavelet Transform) are used to keep imperceptibility high. To improve the security, cryptographic method NTRU is applied on the secret message before it is hidden in image. The result showed that the combination of NTRU, IHWT and modulus function yields good imperceptibility, with PSNR value above 40 dB while the stego image resist salt and pepper noise attack of 0,002% and contrast addition of maximum amount one

Medical image cipher using hierarchical diffusion and non-sequential encryption

This paper presents a medical image cipher with hierarchical diffusion and non-sequential encryption mechanism. An improved permutation approach is developed, which can contribute to not only image shuffling but also lightweight pixel modification at the same time. Therefore, diffusion property of the proposed scheme is achieved from both the permutation and diffusion phases, and hierarchical diffusion effect is thus obtained. Besides, non-sequential visiting mechanism is developed to encrypt the plain pixels in secret order; in other words, pixel visiting order is exploited as extra encryption factor in addition to the traditional pixel location and gray value. With hyper-chaotic Lorenz system employed as key stream generator, a complete image encryption scheme is finally constructed. Experimental results and security analyses validate the effectiveness and superiority of the proposed cipher.

Adaptive multi-threshold based de-noising filter for medical image applications

Medical image processing is the emerging research area and many researchers contributed to medical image processing by proposing new techniques for medical image enhancement and abnormality detection. Interpretation of medical images is a challenging problem because of the unavoidable noise produced by the medical imaging devices and interference. In this work, a new framework is proposed for noise detection and reduction. This framework comprises two phases. First phase is the noise detection phase which is performed using the newly proposed adaptive multi-threshold scheme (AMT). In second phase, modification of noisy pixel is done using edge preserving median filter (EPM), which conserves the edge component and controls the blurring effect with preservation of fine details of interior region. The proposed work is tested with benchmark images and few medical images. It produces promising result and the results are compared with existing two-stage noise reduction techniques. Popular performance metrics such PSNR and SSIM are used for evaluation. Quantitative analysis and experimental results demonstrate that the proposed method is more efficient and suitable for medical image pre-processing.

Adaptive multi-threshold based de-noising filter for medical image applications

Medical image processing is the emerging research area and many researchers contributed to medical image processing by proposing new techniques for medical image enhancement and abnormality detection. Interpretation of medical images is a challenging problem because of the unavoidable noise produced by the medical imaging devices and interference. In this work, a new framework is proposed for noise detection and reduction. This framework comprises two phases. First phase is the noise detection phase which is performed using the newly proposed adaptive multi-threshold scheme (AMT). In second phase, modification of noisy pixel is done using edge preserving median filter (EPM), which conserves the edge component and controls the blurring effect with preservation of fine details of interior region. The proposed work is tested with benchmark images and few medical images. It produces promising result and the results are compared with existing two-stage noise reduction techniques. Popular performance metrics such PSNR and SSIM are used for evaluation. Quantitative analysis and experimental results demonstrate that the proposed method is more efficient and suitable for medical image pre-processing.

Broadband infrared binary-pattern metasurface absorbers with micro-genetic algorithm optimization.

Broadband binary-pattern metasurface absorbers are designed and demonstrated in the mid-infrared wavelength range through the micro-genetic algorithm. The tungsten-based metasurface absorbers with the optimized binary-pattern nanostructures exhibit broadband near-perfect absorption due to the multiple plasmonic resonances supported within the unit cell. Furthermore, the influence of minor pixel modifications in the optimized binary-pattern nanostructures on the absorption performance is investigated in the experiment. This Letter presents a promising approach to design and optimize complex optical nanostructures with the desired properties for metamaterial and metasurface applications.

A Large Capacity Histogram-Based Watermarking Algorithm for Three Consecutive Bins

The histogram watermark, which performs watermark embedding by slightly modifying the histogram of the original image, has been a hot research topic in information hiding technology due to the superiority of its pixel modification during the watermark embedding process, which is independent of the pixel position. This property makes the histogram-based watermark strong resistant to geometric attacks, such as cropping attack, crossed attack, rotation attack, etc. In this paper, we propose a large capacity histogram-based robust watermarking algorithm based on three consecutive bins for the first time. In our scheme, we divide the shape of three consecutive bins into eight cases. According to these cases, we embed Information Number 0, 1, 2, 3, 4, 5, 6, or 7, respectively. The embedded information capacity reaches one bit per bin (bpb), and the amount of embedded information is equal to 200% of the previous existing algorithms. Experimental results show that the new algorithm not only has a large capacity of embedding information, but also has strong robustness to geometric attacks, as well as common image processing operations.

Digital steganography model and embedding optimization strategy

The existed digital steganography models and theories are not effective enough to guide the steganography processing. Based on previous studies, this paper proposes a complete digital steganography model based on additive noise. And then, with security analysis from KL divergence, the embedding optimization strategy is given through theoretical derivation needless of any side information: optimizing the embedding modification position and optimizing the embedding modification direction (+1 or − 1). Through theoretical derivation, we also obtain the quantitative relationship between the pixels modification probability and the adjacent pixels difference, and prove that modification by ±1 randomly cannot enhance steganographic security definitely. The research in this paper can provide theoretical guidance for the design of steganography algorithms. Compared with previous studies, the proposed embedding optimization strategy has outstanding advantages of being easy to implement and being effective to improve steganographic security. The experiments by optimizing LSBM and MG algorithms show that the proposed embedding optimization strategy can effectively improve each algorithm’s steganographic security at a relative small payload.

Tailoring reversible data hiding for 3D synthetic images

In this paper, we proposed a reversible data hiding (RDH) method for three-dimensional (3D) synthetic images based on the depth-no-synthesis-error model (D-NOSE). By using the depth-image-based rendering technique, the 3D image can be synthesized from a 2D color image and its corresponding depth map to provide the viewer a depth perception. Recently, a 3D-based RDH method is proposed to modify the depth map while still preserving the perfect 3D image synthesization. However, it has not fully exploited the pixel mapping relations in the synthesization, resulting in an insufficient capacity. Moreover, the optimal 3D visual quality is not discussed. To compensate this, we propose to improve the capacity and the 3D visual quality. The proposed method extends the pixels expansion toward two directions to fully exploit the allowable range, and can adjust the pixel modification to control the capacity. Furthermore, for a given capacity, the optimal baseline can be selected for a better 3D visual quality. Experimental results show that the proposed method outperforms the previous methods in terms of capacity and distortion, and a better 3D visual quality can be obtained as well.

A 750 K Photocharge Linear Full Well in a 3.2 μm HDR Pixel with Complementary Carrier Collection.

Mainly driven by automotive applications, there is an increasing interest in image sensors combining a high dynamic range (HDR) and immunity to the flicker issue. The native HDR pixel concept based on a parallel electron and hole collection for, respectively, a low signal level and a high signal level is particularly well-suited for this performance challenge. The theoretical performance of this pixel is modeled and compared to alternative HDR pixel architectures. This concept is proven with the fabrication of a 3.2 μm pixel in a back-side illuminated (BSI) process including capacitive deep trench isolation (CDTI). The electron-based image uses a standard 4T architecture with a pinned diode and provides state-of-the-art low-light performance, which is not altered by the pixel modifications introduced for the hole collection. The hole-based image reaches 750 kh+ linear storage capability thanks to a 73 fF CDTI capacitor. Both images are taken from the same integration window, so the HDR reconstruction is not only immune to the flicker issue but also to motion artifacts.

Geometrically Invariant Image Watermarking Using Histogram Adjustment

In this article, a blind image watermarking scheme, which is a robust against common image processing and geometric attacks is proposed by adopting the concept of histogram-based embedding. The average filter is employed to low-pass pre-filter the host image. The watermark bits are embedded into the histogram of the low-frequency component and the template bits are embedded in the high-frequency residual. The embedding is performed by adjusting the value of two consecutive histogram bins. Furthermore, a post-quantization is employed after the embedding round to improve robustness. All pixel modifications incurred are based on the human visual system (HVS) characteristics. As a result, a good tradeoff between robustness and imperceptibility is achieved. Experimental results reported the satisfactory performance of the proposed scheme with respect to both common image processing and geometric attacks.

Pairwise IPVO-based reversible data hiding

Recently, Peng et al. proposed a reversible data hiding method based on improved pixel-value-ordering (PVO) and prediction-error expansion. In this paper, a novel method is proposed by extending Peng et al.’s work. In our method, three largest (or smallest) pixels in a block are utilized to generate two differences, and a new pixel modification strategy is proposed so that the PVO remains unchanged after data embedding. Taking three largest pixels for example, we utilize the third largest pixel to predict the second largest one, and meanwhile use the second largest one to predict the maximum. In this way, two differences are obtained. They are modified jointly so as to be embedded with log 23 bits instead of 2 bits in the traditional RDH methods. The advantage of doing so is to exclude situations where PVO is changed. Moreover, two embedding layers are utilized together to further decrease the embedding distortion. Extensive experiments verify that the proposed method outperforms Peng et al. ’s and some other state-of-the-art works.

A Robust Chaos-Based Image Cryptosystem with an Improved Key Generator and Plain Image Sensitivity Mechanism

In this paper, we propose an effective gray image cryptosystem containing Arnold cat map for pixel permutation and an improved Logistic map for the generation of encryption keys to be used for pixel modification. Firstly, a new chaotic map is designed to show better performance than the standard one in terms of key space range, complexity and uniformity. Generated secret key is not only sensitive to the control parameters and initial condition of the improved map but also strongly depend on the plain image characteristic which provides an effective resistance against statistical and differential attacks. Additionally, to get higher encryption strength of the cryptosystem, both confusion and diffusion processes are performed with different keys in every iterations. Theoretical analysis and simulation results confirm that the proposed algorithm has superior security and effectively encrypts and decrypts the gray images as well.

Reversible data hiding based on flexible block-partition and adaptive block-modification strategy

A novel reversible data hiding (RDH) method based on flexible block-partition and adaptive pixel-modification strategy is proposed in this paper, which is implemented from the following two aspects. One aims at partitioning flexibly a smooth block into non-overlapped sub-blocks of arbitrary size according to a local complexity measurement. After partition, since each resulting sub-block can be treated as an independent embedding unit, this block can be embedded with more data bits. The other is that the different pixel modification method is utilized for blocks (or sub-blocks) of different levels. Specifically, for a block, if it is not suitable for further division, only the maximum and minimum are modified at the same time so as to keep the distortion as low as possible. If it is divided into smaller sub-blocks, two pixel-modification schemes are designed for sub-blocks of size 1×3 and other sizes, respectively. One is that a 1×3 block can be embedded with 2 data bits by only modifying the maximum of this block. In this way, the embedding distortion is further decreased. The other is that in order to exploit better redundancy, two largest and two smallest pixels are modified simultaneously so that at most 4 data bits are embedded into a sub-block. Extensive experiments verify that the proposed method outperforms Peng et al.’s, Wang et al.’s, Li et al.’s, Sachnev et al.’s and Hong et al.’s works.