Digital Image Authentication Research Articles

Copy-move forgery detection of medical images using golden ball optimization

The medical images can be tampered with by attackers with a malevolent goal of hiding or creating multiple copies of lesions, resulting in wrong treatment, false insurance claim, defame public figures, and so on. Such forgery demands authentication of digital images before starting the diagnosis and treatment of the patients. The existing key-point (KP) based forgery detection methods may not uniformly distribute the KPs on the images, thereby making the detection process to fail for images with forgeries in smooth regions. This paper attempts to employ a minimum Eigen-value algorithm that distributes the KPs in the entire image region and apply speeded up robust features and singular value decomposition for obtaining reduced descriptors at identified KPs. Besides, the method applies golden ball-based optimization, inspired by the behaviour of players in team-based sports games, for optimal clustering of the evaluated features. This paper studies the performances of the suggested method on 300 medical images and compares the results with the existing methods. It exhibits that the precision, specificity, sensitivity, and accuracy of the developed method are superior to the existing methods. Though the suggested method outperforms, it can further be improved by combining it with block-based methods.

Comparative Study of Recent CMFD Techniques: A Review

Due to the ease in manipulation of digital images with various existing photo editing tools and softwares the task of authentication and identifying the reliability of digital image has become a major concern. Copy-move forgery detection (CMFD) is therefore known to be one of the key-domains in the latest digital image authentication study. Copy-move forgery is a passive approach in which a portion of an image is copied and then pasted onto the same image, resulting in a tampered image. In this paper, current developments in CMFD have been surveyed and a comparative evaluation of recent CMFD techniques has been done along with its advantages and limitations. Also, a detailed description of relevant copy-move forgery detection datasets is provided which will help the researchers to decide which dataset to choose for a given CMFD approach.

Splicing Image Forgery Detection and Localization Based on Color Edge Inconsistency using Statistical Dispersion Measures

Nowadays, due to the availability of low-cost and high-resolution digital cameras, and the rapid growth of user-friendly and advanced digital image processing tools, challenges for ensuring authenticity of digital images have been raised. Therefore, development of reliable image authenticity verification techniques has high importance in digital life. In this paper, we proposed a blind image splicing detection method based on color distribution in the neighborhood of edge pixels. First, we extracted edge pixels using contourlet transform. Then, to accurately distinguish the authentic edges from tampered ones, Interquartile Range (IQR) criteria are utilized to illustrate the distribution of Cr and Cr histograms of the spliced boundaries in YCbCr color space. Finally, a segmentation method is used to improve the localization performance and to reduce especially the computational time. The effectiveness of the method has been demonstrated by our experimental results obtained using the Columbia Image Splicing Detection Evaluation (CISED) dataset in terms of specificity and accuracy. It is observed that the proposed method outperforms some state-of-the-art methods. The detection accuracy is approximately 97 with 100% specificity.

Color image watermarking for content authentication and self-restoration applications based on a dual-domain approach

Digital image watermarking has been witnessed as a powerful tool that can be used in image ownership identification, copyright protection and content authentication applications. In this paper, a dual-domain watermarking scheme for digital image authentication with self-recovery capabilities is proposed. The proposed scheme has a high ability to recover tampered image regions while pertaining high watermarking imperceptibility and high robustness under various types of attacks. The proposed scheme employs two watermarks: 1) content-dependent, and, 2) content-independent, where the content-dependent watermark contains two-bit sequences for authentication and recovery operations. The authentication bits are generated based on the singular value decomposition (SVD) of each block, while the recovery bits are generated from the average value within selected sub-blocks. The independent watermark is embedded using the Discrete Cosine Transform (DCT), whereas, the dependent watermark is embedded directly into the first and second least significant bits (LSB) of the image pixels. The developed dual-domain watermarking scheme has exhibited high robustness performance against various attacks with high imperceptibility and watermarking capacity as well. For authentication purposes, the proposed scheme incorporated three developed measures to detect the authenticity of the probe image, and, to accurately localize tampered regions. The experimental results have shown that the proposed scheme has a high imperceptibility, compared to other existing schemes, and can effectively distinguish malicious from incidental attacks with accurate tampering localization. Moreover, the proposed scheme exhibits high performance against moderate-to-strong incidental attacks such as noise and low-quality JPEG attacks.

Blind Image Source Camera Identification

Advanced imaging devices that are effective and widely available make it easier to create and control digital photographs. As a consequence it is not easier to validate the authenticity of digital images in the case of legal issues. The proposed method is an effective source examination and identification procedure for an image whose source camera is being investigated, with no prior knowledge of either the source camera or the picture. Even though Photo Response Non-Uniformity (PRNU) has acquired an incredible interest in image forensics the extraction of camera fingerprint with minimum seen content is still a challenging problem. In the proposed method fingerprint is extracted using a weighted combination of two types of de-noising filters, weighted nuclear norm minimization (WNNM) based filter and wavelet filter. A set of images captured by different mobile cameras at different locations are used as the dataset. Each image is distinct and contains different seen details having varying exposures. After the fingerprint extraction process, the fingerprints are clustered using correlation clustering followed by consensus clustering and final refinement based on their origin.

Fusion of Handcrafted and Deep Features for Forgery Detection in Digital Images

Content authentication of digital images has captured the attention of forensic experts and security researchers due to a multi-fold increase in the dissemination of multimedia data through the open and vulnerable Internet. Shrewd attackers successfully devise novel ways to challenge state-of-art forensic tools used for forgery detection in digital images. Feature engineering approaches have yielded up to 97% accuracy on benchmarked datasets. Deep learning approaches have shown promising results in various image classification problems but cannot find hidden patterns in digital images, which can reliably detect image forgeries. State-of-art accuracy of deep learning approaches for forgery detection is up to 98% on benchmarked datasets. The objective of the proposed approach is to further escalate the detection accuracy, pushing it near 100%. In this paper, a synergy of handcrafted features based on color characteristics and deep features using the image’s luminance channel is employed to mine patterns responsible for accurate forgery detection. In the first Stream, 648-D Markov-based features are computed from the quaternion discrete cosine transform of the image. In the second Stream, the luminance channel of YCbCr colorspace is used to extract the Local Binary Pattern of the image. Further, local binary feature maps are fed to the pre-trained ResNet-18 model to obtain a 512-D feature vector called ’ResFeats’ from the last layer of the model’s convolutional base portion. The handcrafted features from Stream I and ResFeats from Stream II are combined to form an 1160-D feature vector. Further, classification is performed using a shallow neural network, and the method is tested on CASIA v1 and CASIA v2 datasets. The accuracy of the proposed fusion-based approach is 99.3% on benchmark datasets.

Classification and evaluation of digital forensic tools

Digital forensic tools (DFTs) are used to detect the authenticity of digital images. Different DFTs have been developed to detect the forgery like (i) forensic focused operating system, (ii) computer forensics, (iii) memory forensics, (iv) mobile device forensics, and (v) software forensics tools (SFTs). These tools are dedicated to detect the forged images depending on the type of the applications. Based on our review, we found that in literature of the DFTs less attention is given to the evaluation and analysis of the forensic tools. Among various DFTs, we choose SFTs because it is concerned with the detection of the forged digital images. Therefore,the purpose of this study is to classify the different DFTs and evaluate the software forensic tools (SFTs) based on the different features which are present in the SFTs. In our work, we evaluate the following five SFTs, i.e.,“FotoForensics”, “JPEGsnoop”, “Ghiro”, “Forensically”, and “Izitru”, based on different features so that new research directions can be identified for the development of the SFTs.

Efficient Encryption System for Numerical Image Safe Transmission

In this paper, we propose an efficient cryptosystem for digital image encryption and authentication. The cryptosystem is a hybrid scheme that uses symmetric and asymmetric approaches. The first one is used to encrypt the host image by utilizing a chaos-based key generator. The second one is used to encrypt the initial secret key and the owner’s signature that permit authentication. The algorithm is evaluated and validated by its application on several types of standard images and tools such as the statistical analysis, the key, and the performance analysis. The results indicate that the proposed cryptosystem provides high performance and enhanced security. The NIST 800-22 is used for testing the pseudorandom numbers generation (PRNG). The obtained simulation results are better than those cited in the recent works in terms of execution time and security level and low computational complexity.

A comprehensive survey of detecting tampered images and localization of the tampered region

Digital image fraudulent has become a very dangerous problem because of the rapidly growing media editing software. The information on the digital image can be easily replaced by fake information. Several cybercrime activities are attempted by hacking the digital images. Due to the presence of various kinds of media editing software, the digital image authenticity is very much significant. Nowadays, the detection of the manipulated images and localization of the manipulated regions are important issues. Many efforts have been made over the past decade to detect the tampered images and localization of the tampered regions with high accuracy based on some specially designed mechanisms. This paper presents a detailed survey of existing approaches for detecting tampered images and localization of the tampered regions.

A passive approach for the detection of splicing forgery in digital images

With the technology progress, a plethora of freely accessible software has questioned the authenticity of digital images. This field is continuously creating challenges for researchers to ascertain the integrity of images. Hence, there is a need to improve the performance of forgery detection algorithms from time to time. This paper is focused on the detection of splicing forgery because it is one of the most frequently used image manipulation techniques. In the proposed scheme, Markov features in both Discrete Wavelet Transform (DWT) and Local Binary Pattern (LBP) domains are extracted and combined for the detection of image splicing. Three-level DWT is applied to the source image by the means of discrete Haar wavelet. The image is split in to high and low-frequency sub-bands after applying one level DWT. Furthermore, low-frequency sub-band is decomposed twice to obtain three-level DWT, which leads to more information and less amount of noise. The efficacy of the proposed scheme has been appraised on six benchmark datasets i.e. CASIA v2.0, DVMM, IFS-TC, CASIA v1.0, Columbia, and DSO-1. Moreover, the SVM classifier is trained to classify the images as tampered or authentic. The effectiveness of the proposed scheme is evaluated based on various performance parameters such as accuracy, sensitivity, specificity, and informedness. The proposed results show improved accuracy i.e. 99.69%, 99.76%, 97.80%, 98.61%, 96.90%, and 92.50% on CASIA v1.0, CASIA v2.0, DVMM, Columbia, IFS-TC, and DSO-1, respectively, in comparison to other existing approaches.

Detection of Computer Graphics Using Attention-Based Dual-Branch Convolutional Neural Network from Fused Color Components.

With the development of 3D rendering techniques, people can create photorealistic computer graphics (CG) easily with the advanced software, which is of great benefit to the video game and film industries. On the other hand, the abuse of CGs has threatened the integrity and authenticity of digital images. In the last decade, several detection methods of CGs have been proposed successfully. However, existing methods cannot provide reliable detection results for CGs with the small patch size and post-processing operations. To overcome the above-mentioned limitation, we proposed an attention-based dual-branch convolutional neural network (AD-CNN) to extract robust representations from fused color components. In pre-processing, raw RGB components and their blurred version with Gaussian low-pass filter are stacked together in channel-wise as the input for the AD-CNN, which aims to help the network learn more generalized patterns. The proposed AD-CNN starts with a dual-branch structure where two branches work in parallel and have the identical shallow CNN architecture, except that the first convolutional layer in each branch has various kernel sizes to exploit low-level forensics traces in multi-scale. The output features from each branch are jointly optimized by the attention-based fusion module which can assign the asymmetric weights to different branches automatically. Finally, the fused feature is fed into the following fully-connected layers to obtain final detection results. Comparative and self-analysis experiments have demonstrated the better detection capability and robustness of the proposed detection compared with other state-of-the-art methods under various experimental settings, especially for image patch with the small size and post-processing operations.

Fragile watermarking for image authentication using BRINT and ELM

Privacy and security concerns regarding digital information have been increasing with advancements in multimedia and network technologies. Digital images are an integral component of online-distributed content, which plays a crucial role in forming public perceptions and opinions. They also play a major role in sensitive areas, such as law and defense. Any attempt to tamper with them is, therefore, a serious issue. One common approach to this problem is fragile image watermarking, which is used to ensure the authenticity of digital content. In this paper, a new fragile watermarking method for the authentication of digital images is proposed based on a binary rotation invariant and noise tolerant (BRINT) local texture descriptor and an extreme learning machine (ELM). BRINT is used to generate and retrieve the watermark in both the embedding and extraction procedures. In parallel, ELM is used in both procedures to learn and recover any tampered areas. The experimental results showed that the proposed scheme does not degrade image quality, allows for tamper detection, and has a recovery ability comparable with state-of-the-art fragile and semi-fragile watermarking schemes. Moreover, the proposed scheme has been validated as a fragile watermarking method with the potential to detect and locate modifications in digital images, such as copy-paste forgery, JPEG compression, and noise addition. This method is useful in sensitive fields, like defense, law, and journalism, in which decision-making based on digital visual information is necessary, to ensure that an image is authentic and has not been tampered with.

Proving Authenticity and Integrity of Digital Images using Image Forensics

Proving Authenticity and Integrity of Digital Images using Image Forensics

A Survey of Deep Learning-Based Source Image Forensics.

Image source forensics is widely considered as one of the most effective ways to verify in a blind way digital image authenticity and integrity. In the last few years, many researchers have applied data-driven approaches to this task, inspired by the excellent performance obtained by those techniques on computer vision problems. In this survey, we present the most important data-driven algorithms that deal with the problem of image source forensics. To make order in this vast field, we have divided the area in five sub-topics: source camera identification, recaptured image forensic, computer graphics (CG) image forensic, GAN-generated image detection, and source social network identification. Moreover, we have included the works on anti-forensics and counter anti-forensics. For each of these tasks, we have highlighted advantages and limitations of the methods currently proposed in this promising and rich research field.

Digital Image Watermarking Techniques: A Review

Digital image authentication is an extremely significant concern for the digital revolution, as it is easy to tamper with any image. In the last few decades, it has been an urgent concern for researchers to ensure the authenticity of digital images. Based on the desired applications, several suitable watermarking techniques have been developed to mitigate this concern. However, it is tough to achieve a watermarking system that is simultaneously robust and secure. This paper gives details of standard watermarking system frameworks and lists some standard requirements that are used in designing watermarking techniques for several distinct applications. The current trends of digital image watermarking techniques are also reviewed in order to find the state-of-the-art methods and their limitations. Some conventional attacks are discussed, and future research directions are given.

Comparative performance study of classification models for image-splicing detection

Abstract Recently, images have been manipulated for malicious activities, rather than to enhance their quality. This allows malicious users to make changes to images to create forged versions using digital processing tools. Therefore, the authenticity of digital images has become an important research area because humans cannot observe image forgery processes. The objective of this study was to detect spliced images using three classification techniques—support vector machine (SVM), naive Bayes, and K nearest neighbors (KNN)—to identify the most suitable one. Their classification quality was evaluated using accuracy, sensitivity, and specificity as performance measures. The experimental results showed that the KNN classifier achieved the highest accuracy and sensitivity among the three classifiers. However, naive Bayes achieved the highest specificity among the classifiers.

Passive Image Forgery Detection Based on the Demosaicing Algorithm and JPEG Compression

Multimedia files play an important role in everyday life. Today, the majority of the population owns state-of-the-art cameras integrated into their mobile devices. Technological development not only facilitates the generation of multimedia content, but also the intentional manipulation of it, and this is where forensic techniques of detecting manipulation on images and videos take on great importance. Although historically there has been confidence in the integrity of images, the advance of technology has begun to erode this confidence. This work proposes a digital image authentication method based on the quadratic mean error of the Color Filter Array interpolation pattern estimated from the analysed image. For the evaluation of the proposed method, experiments were carried out with public databases of forged images that are widely developed for research purposes. The results of the experiments demonstrate the efficiency of the proposed method.

USING THE METHOD OF STEGANOGRAPHIC NOISE ANALYSIS TO IDENTIFY SIGNS OF DIGITAL PHOTO EDITING

In today’s world, digital technology has almost replaced analog, and in the materials of pre-trial and judicial proceedings used physical evidence in the form of electronic documents or their images, photographs, videos, scanned copies, having a digital nature of education. To date, the question of determining the authenticity and research installation of digital image pressing issues fototechnika examination.
 The main purpose of the study using the method of steganographic analysis of image noise is to determine the authenticity of digital photos, identify signs of changes in the «content» of such images.
 The possibilities of digital image research methods using the method of steganographic noise analysis to identify signs of editing digital photos (images) to improve the efficiency, quality and evidential significance of photographic examinations are considered. The proposed new methodical approach in solving problems for the study of the installation of digital images, or when establishing their veracity (authenticity).
 Developed under the methodology of the study of signs installation of digital images based on the analysis of entropy noise software allows you to conduct research to identify signs of installation on the basis of the analysis of entropy noise digital images by maximization of complex criterion that takes into account the entropy and the variance of the image. It can be used both on local computers, and with use of the powerful server equipment in the mode of remote access through the web interface, including, with use of low-power mobile terminals of the user (smartphones, tablets and so forth) that will considerably facilitate work of experts on research of installation of digital images or at establishment of their reliability (authenticity).

Robust Copy-Paste Detection Algorithm using SIFT for Digital Image Forensics

Forensics of images verifies the authenticity of digital images. Because of the easier availability of software, manipulation of images has become quicker and easier. Image composition, copy-paste, multiple cloning, splicing, etc have become a common practice. The paper proposes a robust algorithm for the detection of duplicity using Scale Invariant Feature Transform (SIFT) approach. Copied location of an image is occasionally pasted in another place of the identical image or in another image, which creates difficulties in detecting the copy-paste region and identifying the located region as well as creates inefficiency in the accuracy of forgeries. We developed an approach that shows improved techniques through runtime optimizations and compared various parameters with existing methodologies in order to obtain highly correlated image areas for detecting the manipulated regions. The proposed approach can detect copy-paste forgeries effectively with high accuracy, reliability, and inconsistencies regardless of the test scenario.

Robust Perceptual Image Hashing using SIFT and SVD

With the advancement in technology, digital data such as image, video, etc. can be easily manipulated. Image hashing is a method that can be used for authentication and identification of digital images. In this communication, a robust image hashing technique is proposed using scale invariant feature transform (SIFT), singular value decomposition (SVD) and Zernike moment. Zernike moment is used to restore the image against the rotation attack. Potential points are generated from the image by using SIFT. Block processing of equal size is performed on the input image. Key points within the same block are used to generate hashing values. The experimental outcome shows that the proposed technique can withstand different types of attacks. Receiver operating characteristic curve comparison specifies that our method outperforms other existing methods under consideration.