This PDF file contains the editorial “Special Section Guest Editorial: Image Quality Assessment” for JEI Vol. 19 Issue 01

We propose a means of objectively comparing the results of digital image inpainting algorithms by analyzing changes in predicted human attention prior to and following application. Artifacting is generalized in two catagories, in-region and out-region, depending on whether or not attention changes are primarily within the edited region or in nearby (contrasting) regions. Human qualitative scores are shown to correlate strongly with numerical scores of in-region and out-region artifacting, including the effectiveness of training supervised classifiers of increasing complexity. Results are shown on two novel human-scored datasets.

Objective image and video quality measures play important roles in numerous image and video processing applications. In this work, we propose a new content-weighted method for full-reference (FR) video quality assessment using a three-component image model. Using the idea that different image regions have different perceptual significance relative to quality, we deploy a model that classifies image local regions according to their image gradient properties, then apply variable weights to structural similarity image index (SSIM) [and peak signal-to-noise ratio (PSNR)] scores according to region. A frame-based video quality assessment algorithm is thereby derived. Experimental results on the Video Quality Experts Group (VQEG) FR-TV Phase 1 test dataset show that the proposed algorithm outperforms existing video quality assessment methods.

Quality metrics quantify by how much some aspect of a measurement deviates from a predefined standard. Measurement quality evaluations of laser range scanner data are used to perform range image registration, merging measurements, and view planning. We develop a scanning method that uses laser range scanner quality metrics to both reduce the time required to obtain a complete range image from a single viewpoint and the number of measurements obtained during the scanning process. This approach requires a laser range scanner capable of varying both the area and sampling density of individual scans, but can be combined with view planning methods to reduce the total time required to obtain a complete surface map of an object. Several new quality metrics are introduced: outlier, resolvability, planarity, integration, return, and enclosed quality metrics. These metrics are used as part of a quality-based merge method, referred to here as a quality-weighted modified Kalman minimum variance (weighted-MKMV) estimation method. Experimental evidence is presented confirming that this approach can significantly reduce the total scanning time. This approach could be particularly useful for rapidly generating CAD models of real-world objects.

Images usually exhibit regions that particularly attract the viewer's attention. These regions are typically referred to as regions of interest (ROI), and the underlying phenomenon in the human visual system is known as visual attention (VA). In the context of image quality, one can expect that distortions occurring in the ROI are perceived as being more annoying compared to distortions in the background. However, VA is seldom taken into account in existing image quality metrics. In this work, we provide a VA framework to extend existing image quality metrics with a simple VA model. The performance of the framework is evaluated on three contemporary image quality metrics. We further consider the context of wireless imaging where a broad range of artifacts can be observed. To facilitate the VA-based metric design, we conduct subjective experiments to both obtain a ground truth for the subjective quality of a set of test images and to identify ROI in the corresponding reference images. A methodology is further discussed to optimize the VA metrics with respect to quality prediction accuracy and generalization ability. It is shown that the quality prediction performance of the three considered metrics can be significantly improved by deploying the proposed framework.

The mainstream approach to image quality assessment has centered around accurately modeling the single most relevant strategy employed by the human visual system (HVS) when judging image quality (e.g., detecting visible differences, and extracting image structure/information). In this work, we suggest that a single strategy may not be sufficient; rather, we advocate that the HVS uses multiple strategies to determine image quality. For images containing near-threshold distortions, the image is most apparent, and thus the HVS attempts to look past the image and look for the distortions (a detection-based strategy). For images containing clearly visible distortions, the distortions are most apparent, and thus the HVS attempts to look past the distortion and look for the image's subject matter (an appearance-based strategy). Here, we present a quality assessment method [most apparent distortion (MAD)], which attempts to explicitly model these two separate strategies. Local luminance and contrast masking are used to estimate detection-based perceived distortion in high-quality images, whereas changes in the local statistics of spatial-frequency components are used to estimate appearance-based perceived distortion in low-quality images. We show that a combination of these two measures can perform well in predicting subjective ratings of image quality.

Mobile applications present new image quality challenges. Automotive vision requires reliable capture of scene detail. Photospace measurements have shown that the extremely wide intrascene dynamic range of traffic scenes necessitates wide-dynamic-range (WDR) technology. Multiple-slope complementary metal-oxide semiconductor (CMOS) technology adaptively extends dynamic range by partially resetting the pixel, resulting in a response curve with piecewise linear slopes of progressively increasing compression. As compression and thus dynamic range increase, a trade-off against detail loss is observed. Incremental signal-to-noise ratio (iSNR) has been proposed in ISO/TC42 standards for determining dynamic range, and this work describes how to adapt these to WDR. Measurements and computer simulations reveal that the observed trade-off between WDR extension and the loss of local detail can be explained by a drop in iSNR at each reset point. If a reset is not timed optimally, then iSNR may drop below the detection limit causing an iSNR hole to appear within the dynamic range. Thus iSNR has extended utility: it not only determines the dynamic range limits but also defines dynamic range as the luminance range where detail detection is reliable. It has become the critical criterion when maximizing dynamic range to maintain the minimum necessary level of detection reliability.

The use of color electrophotographic (EP) laser printing systems is growing because of their declining cost. Thus, the print quality of color EP laser printers has become increasingly important. Since text and lines are indispensable to print quality, many studies have proposed methods for measuring these print quality attributes. Toner scatter caused by toner overdevelopment in color EP laser printers can significantly impact print quality. A conventional approach to reduce toner overdevelopment is to restrict the color gamut of printers. However, this can result in undesired color shifts and the introduction of halftone texture. Coring, defined as a process where the colorant level is reduced in the interior of text or characters, is a remedy for these shortcomings. The desired amount of reduction for coring depends on line width and overall nominal colorant level. In previous work, these amounts were chosen on the basis of data on the perception of edge blur obtained from softcopy simulation of the blurring. We describe psychophysical studies to directly establish optimal coring values as a function of line width and nominal colorant level. For each line width and nominal colorant level, this is done by asking human subjects to choose the minimum amount of coring that is necessary to eliminate the perception of toner scatter. We conduct four separate psychophysical studies to address different aspects of this question.

ISO 20462 part 3 standardized the hardcopy quality ruler and a softcopy quality ruler based on a binary sort approach involving paired comparisons. The new softcopy ruler method described here utilizes a slider bar to match the quality of the ruler to that of the test image, which is found to substantially reduce the time required per assessment (30 to 15.5 s), with only a modest loss of precision (standard deviations of 2.5 to 2.9 just noticeable differences). In combination, these metrics implied a 20% improvement in the standard error of the mean achievable in a fixed amount of judging time. Ruler images calibrated against the standard quality scale of ISO 20462 are generated for 21 scenes, at 31 quality levels each, achieved through variation of sharpness, while other attributes are held near their preferred positions. The images are bundled with documentation and a MATLAB source code for a graphical user interface that administers softcopy ruler experiments, and these materials are donated to the International Imaging Industry Association for distribution. In conjunction with a specified large flat panel display, these materials should enable users to conduct softcopy quality ruler experiments with minimum effort, and should reduce the barriers to performing calibrated psychophysical measurements.

The capacity of a printing system to accurately reproduce details has an impact on the quality of printed images. The ability of a system to reproduce details is captured in its modulation transfer function (MTF). In the first part of this work, we compare three existing methods to measure the MTF of a printing system. After a thorough investigation, we select the method from Jang and Allebach and propose to modify it. We demonstrate that our proposed modification improves the measurement precision and the simplicity of implementation. Then we discuss the advantages and drawbacks of the different methods depending on the intended usage of the MTF and why Jang and Allebach's method best matches our needs. In the second part, we propose to improve the quality of printed images by compensating for the MTF of the printing system. The MTF is adaptively compensated in the Fourier domain, depending both on frequency and local mean values. Results of a category judgment experiment show significant improvement as the printed MTF-compensated images obtain the best scores.

For wide-dynamic-range cameras that achieve a logarithmic response using a load transistor operating in weak inversion, fixed pattern noise is a particular problem. The effects of fixed pattern noise can be reduced using one of a variety of methods that compensate for different forms of fixed pattern noise. In choosing between these methods, it becomes important to know the level of fixed pattern noise that is tolerable for the application in question. Similarly, when deciding on the number of bits that should be used when digitizing the image, it is important to know the level of quantization noise that is tolerable to maintain color image quality. In this work, a method of determining the tolerable levels of fixed pattern and quantization noise is proposed. This leads to the conclusion that for one type of wide-dynamic-range pixel, a relatively simple fixed pattern noise correction procedure gives acceptable results for applications of medium complexity. A more generally applicable result is that if quantization levels equivalent to a contrast change of less than 2% in logarithmic imagers are obtainable, then excellent color image rendition can be achieved.

We aim to quantify the perception of color gradation smoothness using objectively measurable properties. We propose a model to compute the smoothness of hardcopy color-to-color gradations. It is a gradient-based method that can be determined as a function of the 95th percentile of second derivative for the tone-jump estimator and the fifth percentile of first derivative for the tone-clipping estimator. Performance of the model and a previously suggested method were psychophysically appreciated, and their prediction accuracies were compared to each other. Our model showed a stronger Pearson correlation to the corresponding visual data, and the magnitude of the Pearson correlation reached up to 0.87. Its statistical significance was verified through analysis of variance. Color variations of the representative memory colors-blue sky, green grass and Caucasian skin-were rendered as gradational scales and utilized as the test stimuli.

Because of the emergence of e-commerce and developments in print engines designed for economical output of very short runs, there are increased business opportunities and consumer options for print-on-demand books and photobooks. The current state of these printing modes allows for direct uploading of book files via the web, printing on nonoffset printers, and distributing by standard parcel or mail delivery services. The goal of this research is to assess the image quality of print-on-demand books and photobooks produced by various Web-based vendors and to identify correlations between psychophysical results and objective metrics. Six vendors were identified for one-off (single-copy) print-on-demand books, and seven vendors were identified for photobooks. Participants rank ordered overall quality of a subset of individual pages from each book, where the pages included text, photographs, or a combination of the two. Observers also reported overall quality ratings and price estimates for the bound books. Objective metrics of color gamut, color accuracy, accuracy of International Color Consortium profile usage, eye-weighted root mean square L^*, and cascaded modulation transfer acutance were obtained and compared to the observer responses. We introduce some new methods for normalizing data as well as for strengthening the statistical significance of the results. Our approach includes the use of latent mixed-effect models. We found statistically significant correlation with overall image quality and some of the spatial metrics, but correlations between psychophysical results and other objective metrics were weak or nonexistent. Strong correlation was found between psychophysical results of overall quality assessment and estimated price associated with quality. The photobook set of vendors reached higher image-quality ratings than the set of print-on-demand vendors. However, the photobook set had higher image-quality variability.

Ultrasound is a widely used medical imaging methodology because it is safe and relatively inexpensive. However, the quality of the images is affected by the point-spread function of the system, medium effects (such as refraction, reflection, and attenuation), and coherent wave interference or speckle. The present research studies the averaging of images that have been displaced laterally and displays them using an interlaced grid. The main goals are to reduce speckle and improve contrast-to-noise ratio and resolution. The point-spread function of the ultrasound scanner was estimated using a thin nylon thread within a water bath. Then, a set of images of a breast phantom (having lateral displacements smaller than the width of the point spread function) were averaged and interlaced. The results show significant improvements in signal-to-noise ratio and contrast-to-noise ratio.

The dynamic range of digital image sensors is growing faster than the reproducible dynamic range of standard display and print devices. In an assessment, we evaluated eight real-time-capable, high-dynamic-range (HDR) tone-mapping operators, using HDR images restricted to 12 bits per channel, and mapping to 8-bit low-dynamic-range (LDR) output data. In our survey, 51 test persons rated the subjective quality of LDR images of 15 scenes. For each scene, the operators were tested against linear tone mapping. From that survey, we derived three operators that performed better than linear mapping and evaluated their impact on exposure when using tone-mapping in an imaging chain. To mathematically predict the subjective quality of tone-mapped images, we tested five figures of merit against the empirical results of the survey. We propose a quality measure derived from the Naka-Rushton equation. We evaluate the proposed metric against four other image quality metrics, with our novel measure matching the subjective results best.

The evaluation of perceived image quality in color prints is a complex task due to its subjectivity and dimensionality. The perceived quality of an image is influenced by a number of different quality attributes. It is difficult and complicated to evaluate the influence of all attributes on overall image quality, and their influence on other attributes. Because of this difficulty, the most important attributes of a color image should be identified to achieve a more efficient and manageable evaluation of the image's quality. Based on a survey of the existing literature and a psychophysical experiment, we identify and categorize existing image quality attributes to propose a refined selection of meaningful ones for the evaluation of color prints.

An algorithm is described for measuring the subjective, visual impact of 1-D defects (streaks and bands) in prints. A general approach to measurements of spatially localized image defects is described, attempting to directly match three stages of processing by the human observer: formation of the perceived image, identification of individual defects, and pooling of the visual defect magnitudes into an overall assessment. The emphasis of the discussion is on the method of pooling of multiple streaks and band defects. It is demonstrated that the commonly used Minkowski pooling method does not satisfy the basic criteria necessary for this application, and a tent-pole pooling method is defined and analyzed. A complete algorithm for measuring streaks and bands, which uses tent-pole pooling, is described. The performance of the algorithm is demonstrated by comparison to results from new and independently collected subjective ratings.

We present an innovative approach to the objective quality evaluation that could be computed using the mean difference between the original and tested images in different wavelet subbands. Discrete wavelet transform (DWT) subband decomposition properties are similar to human visual system characteristics facilitating integration of DWT into image-quality evaluation. DWT decomposition is done with multiresolution analysis of a signal that allows us to decompose a signal into approximation and detail subbands. DWT coefficients were computed using reverse biorthogonal spline wavelet filter banks. Wavelet coefficients are used to compute new image-quality measure (IQM). IQM is defined as perceptual weighted difference between coefficients of original and degraded image.

Image and video quality assessment (QA) is a critical issue in image and video processing applications. General full-reference (FR) QA criteria such as peak signal-to-noise ratio (PSNR) and mean squared error (MSE) do not accord well with human subjective assessment. Some QA indices that consider human visual sensitivity, such as mean structural similarity (MSSIM) with structural sensitivity, visual information fidelity (VIF) with statistical sensitivity, etc., were proposed in view of the differences between reference and distortion frames on a pixel or local level. However, they ignore the role of human visual attention (HVA). Recently, some new strategies with HVA have been proposed, but the methods extracting the visual attention are too complex for real-time realization. We take advantage of the phase spectrum of quaternion Fourier transform (PQFT), a very fast algorithm we previously proposed, to extract saliency maps of color images or videos. Then we propose saliency-based methods for both image QA (IQA) and video QA (VQA) by adding weights related to saliency features to these original IQA or VQA criteria. Experimental results show that our saliency-based strategy can approach more closely to human subjective assessment compared with these original IQA or VQA methods and does not take more time because of the fast PQFT algorithm.

We propose a novel unsupervised multiresolution adaptive and progressive gradient-based color-image segmentation algorithm (MAPGSEG) that takes advantage of gradient information in an adaptive and progressive framework. The proposed methodology is initiated with a dyadic wavelet decomposition scheme of an arbitrary input image accompanied by a vector gradient calculation of its color-converted counterpart in the 1976 Commission Internationale de l'Eclairage (CIE) L*a*b* color space. The resultant gradient map is used to automatically and adaptively generate thresholds to segregate regions of varying gradient densities at different resolution levels of the input image pyramid. At each level, the classification obtained by a progressively thresholded growth procedure is integrated with an entropy-based texture model by using a unique region-merging procedure to obtain an interim segmentation. A confidence map and nonlinear spatial filtering techniques are combined, and regions of high confidence are passed from one resolution level to another until the final segmentation at the highest (original) resolution is achieved. A performance evaluation of our results on several hundred images with a recently proposed metric called the normalized probabilistic Rand index demonstrates that the proposed work computationally outperforms published segmentation techniques with superior quality.

The topic of aerial image registration attracts considerable interest within the imaging research community due to its significance for several applications, including change detection, sensor fusion, and topographic mapping. Our interest is focused on finding the optimal transformation between two aerial images that depict the same visual scene in the presence of pronounced spatial, temporal, and sensor variations. We first introduce a stochastic edge estimation process suitable for geometric shape-based registration, which we also compare to intensity-based registration. Furthermore, we propose an objective function that weights the L2 distances of the edge estimates by the feature points' energy, which we denote by sum of normalized squared differences and compare to standard objective functions, such as mutual information and the sum of absolute centered differences. In the optimization stage, we employ a genetic algorithm scheme in a multiscale image representation scheme to enhance the registration accuracy and reduce the computational load. Our experimental tests, measuring registration accuracy, rate of convergence, and statistical properties of registration errors, suggest that the proposed edge-based representation and objective function in conjunction with genetic algorithm optimization are capable of addressing several forms of imaging variations and producing encouraging registration results.

We propose a multithreshold progressive reconstruction method. The image is encoded three times using Joint Photographic Experts Group (JPEG): first with a low-quality factor, then with a medium-quality factor, and last with a high-quality factor. Huffman coding is employed to encode the difference between the important image and the high-quality JPEG decompressed image. The three JPEG codes and the Huffman code are shared, respectively, according to four prespecified thresholds. The n-generated equally important shadows can be stored or transmitted using n channels in parallel. Cooperation among these generated shadows can progressively reconstruct the important image. The reconstructed image is loss-free when the number of collected shadows reaches the largest threshold. Each shadow is very compact and so can be hidden successfully in the JPEG codes of cover images to reduce the probability of being attacked when transmitted in an unfriendly environment. Comparisons with other image sharing methods are made. The contributions, such as easiness to apply to scalable Moving Picture Experts Group (MPEG) video transmission or resistance to differential attack, are also included.

A field validation of a laser detection and ranging (LADAR) system was conducted by the U.S. Army Engineer Research and Development Center (ERDC), Vicksburg, Mississippi. The LADAR system, a commercial-off-the-shelf (COTS) LADAR system custom-modified by Autonomous Solutions, Inc. (ASI), was tested for accuracy in measuring terrain geometry. A verification method was developed to compare the LADAR dataset to a ground-truth dataset that consisted of total station measurements. Three control points were measured and used to align the two datasets. The influence of slopes, surface materials, light, fog, and dust were investigated. The study revealed that slopes only affected measurements when the terrain was obscured from the LADAR system, and ambient light conditions did not significantly affect the LADAR measurements. The accuracy of the LADAR system, which was equipped with fog correction, was adversely affected by particles suspended in air, such as fog or dust. Also, in some cases the material type had an effect on the accuracy of the LADAR measurements.

Mosaic images are captured by a single charge-coupled device/complementary metal-oxide-semiconductor (CCD/CMOS) sensor with the Bayer color filter array. We present a new quality-effective zooming algorithm for mosaic images. First, based on adaptive heterogeneity projection masks and Sobel- and luminance-estimation-based masks, more accurate gradient information is extracted from the mosaic image directly. According to the extracted gradient information, the mosaic green (G) channel is first zoomed. To reduce color artifacts, instead of directly moving the original red (R) value to its right position and the blue (B) value to its lower position, the color difference interpolation is utilized to expand the G-R and G-B color difference values. Finally, the zoomed mosaic R and B channels can be constructed using the zoomed G channel and the two expanded color difference values; afterward, the zoomed mosaic image is obtained. Based on 24 popular test mosaic images, experimental results demonstrate that the proposed zooming algorithm has more than 1.79 dB quality improvement when compared with two previous zooming algorithms, one by Battiato et al. (2002) and the other by Lukac et al. (2005).

A cascade of filtering windows is implemented iteratively for removing random-valued impulse noise in heavily corrupted images. This method is based on the peer group concept (PGC), so a pixel is considered as noise-free if and only if for each window size, there exists a peer group of certain threshold cardinality for it. Otherwise, the pixel is considered as noisy. In the restoration process, the corrupted pixels are restored by taking the mean value of the remaining good pixels in the filtering window. Extensive simulations demonstrate that the proposed method produces competitive results at low noise rates, but at high noise rates, it outperforms other state-of-the-art methods. This approach efficiently suppresses the impulse noise, shows a low computational complexity, and has an equal effect on both color and gray-level images.

Recently watermarking algorithms of digital images based on singular value decomposition (SVD) have been proposed. Most SVD-based watermarking techniques use singular values as the embedding watermark information. These SVD-based techniques are advantageous for watermarking images since slight changes in the singular values do not significantly affect the image quality. However, it has been reported that these SVD-based watermarking algorithms, as implemented, suffer from a very high probability of false positive detections of watermarks. The false positive detection vulnerability of two recent SVD-based watermarking schemes is exposed. Two solutions are proposed to mitigate this vulnerability without changing the design principle of these watermarking algorithms.

We propose a practical context-based adaptive image resolution upconversion algorithm. The basic idea is to use a low-resolution (LR) image patch as a context in which the missing high-resolution (HR) pixels are estimated. The context is quantized into classes and for each class an adaptive linear filter is designed using a training set. The training set incorporates the prior knowledge of the point spread function, edges, textures, smooth shades, etc. into the upconversion filter design. For low complexity, two 1-D context-based adaptive interpolators are used to generate the estimates of the missing pixels in two perpendicular directions. The two directional estimates are fused by linear minimum mean-squares weighting to obtain a more robust estimate. Upon the recovery of the missing HR pixels, an efficient spatial deconvolution is proposed to deblur the observed LR image. Also, an iterative upconversion step is performed to further improve the upconverted image. Experimental results show that the proposed context-based adaptive resolution upconverter performs better than the existing methods in both peak SNR and visual quality.

We redefine the unusual event detection problem from a different point of view. Several fundamental event features are investigated and adopted. These features are redescribed in a uniform model. Thus, using this model, supervised/unsupervised unusual event detection algorithms can be designed to fit various situations. Trajectory is treated as the most important feature. To more accurately measure the similarity of different moving object trajectories, a novel distance measurement, the sectional contextual edit distance (SCED), is developed. In the SCED, cost functions are designed according to contextual information and trajectories are segmented into subsections automatically, based on the relevant contexts. Velocity and orientation are also taken into account in cost functions to build an integrated distance similarity measurement. Experimental results demonstrate better performance using the newly proposed similarity measurement while being compared with the existing methods, and some cases of the unusual event detection problem are also demonstrated.

A study of dark current in digital imagers in digital single-lens reflex (DSLR) and compact consumer-grade digital cameras is presented. Dark current is shown to vary with temperature, exposure time, and ISO setting. Further, dark current is shown to increase in successive images during a series of images. DSLR and compact consumer cameras are often designed such that they are contained within a densely packed camera body, and therefore the digital imagers within the camera frame are prone to heat generated by the sensor as well as nearby elements within the camera body. It is the scope of this work to characterize the dark current in such cameras and to show that the dark current, in part due to heat generated by the camera itself, can be corrected by using hot pixels on the imager. This method generates computed dark frames based on the dark current indicator value of the hottest pixels on the chip. We compare this method to standard methods of dark current correction.

We present a modification of the new edge-directed interpolation method that eliminates the prediction error accumulation problem by adopting a modified training window structure, and extending the covariance matching into multiple directions to suppress the covariance mismatch problem. Simulation results show that the proposed method achieves remarkable subjective performance in preserving the edge smoothness and sharpness among other methods in the literaturé. It also demonstrates consistent objective performance among a variety of images.

RC5 is a well-known block cipher notable for its simplicity, speed, suitability for hardware and software implementation, and its low memory requirement. However, RC5 can be broken with various attacks, revealing weak security. This is mainly due to the poor diffusion operation that allows the cipher to be attacked with less complexity than that of the exhaustive search. In this work, we modify the RC5 to give high level security, and adopt it for image encryption by adjusting the structure of both the encryption routine and the key schedule. We generate round keys based on chaos. Then, we strengthen the diffusion operation by making heavy use of rotations. The security analysis of the modified cipher against several attacks is explored from a strict cryptographic viewpoint. Experimental results demonstrate that the modified RC5 provides an efficient and secure way for image encryption.

Image classification has become an important topic in multimedia processing. Recently, neural network-based methods have been proposed to solve the classification problem. Among them, the radial basis function neural network (RBFNN) is the most popular architecture, because it has good learning and approximation capabilities. However, traditional RBFNNs are sensitive to center initialization. To obtain appropriate centers, it needs to find significant features for further RBF clustering. In addition, the training procedure of a traditional RBFNN is time consuming. Therefore, in this work, a combination of a self-organizing map (SOM) and learning vector quantization (LVQ) neural networks is proposed to select more appropriate centers for an RBFNN, and a modular RBF neural network (MRBFNN) is proposed to improve the classification rate and to speed up the training time. Experimental results show that the proposed MRBFNN has better performance than those of the traditional RBFNN, the discrete wavelength transform (DWT)-based method, the tree structured wavelet (TWS), the discrete wavelet frame (DWF), the rotated wavelet filter (RWF), and the wavelet neural network based on adaptive norm entropy (WNN-ANE) methods.

Spatial LSB±1 steganography changes smooth characteristics between adjoining pixels of the raw image. We present a novel steganalysis method for LSB±1 steganography based on feature vectors derived from the co-occurrence matrix in the spatial domain. We investigate how LSB±1 steganography affects the bit planes of an image and show that it changes more least significant bit (LSB) planes of it. The co-occurrence matrix is derived from an image in which some of its most significant bit planes are clipped. By this preprocessing, in addition to reducing the dimensions of the feature vector, the effects of embedding were also preserved. We compute the co-occurrence matrix in different directions and with different dependency and use the elements of the resulting co-occurrence matrix as features. This method is sensitive to the data embedding process. We use a Fisher linear discrimination (FLD) classifier and test our algorithm on different databases and embedding rates. We compare our scheme with the current LSB±1 steganalysis methods. It is shown that the proposed scheme outperforms the state-of-the-art methods in detecting the LSB±1 steganographic method for grayscale images.

This PDF file contains the editorial “The Essential Guide to Video Processing” for JEI Vol. 19 Issue 01