Spatial Domain Image Research Articles

Stress Classification Using Photoplethysmogram-Based Spatial and Frequency Domain Images

Stress is subjective and is manifested differently from one person to another. Thus, the performance of generic classification models that classify stress status is crude. Building a person-specific model leads to a reliable classification, but it requires the collection of new data to train a new model for every individual and needs periodic upgrades because stress is dynamic. In this paper, a new binary classification (called stressed and non-stressed) approach is proposed for a subject’s stress state in which the inter-beat intervals extracted from a photoplethysomogram (PPG) were transferred to spatial images and then to frequency domain images according to the number of consecutive. Then, the convolution neural network (CNN) was used to train and validate the classification accuracy of the person’s stress state. Three types of classification models were built: person-specific models, generic classification models, and calibrated-generic classification models. The average classification accuracies achieved by person-specific models using spatial images and frequency domain images were 99.9%, 100%, and 99.8%, and 99.68%, 98.97%, and 96.4% for the training, validation, and test, respectively. By combining 20% of the samples collected from test subjects into the training data, the calibrated generic models’ accuracy was improved and outperformed the generic performance across both the spatial and frequency domain images. The average classification accuracy of 99.6%, 99.9%, and 88.1%, and 99.2%, 97.4%, and 87.6% were obtained for the training set, validation set, and test set, respectively, using the calibrated generic classification-based method for the series of inter-beat interval (IBI) spatial and frequency domain images. The main contribution of this study is the use of the frequency domain images that are generated from the spatial domain images of the IBI extracted from the PPG signal to classify the stress state of the individual by building person-specific models and calibrated generic models.

A novel minimal distortion-based edge adaptive image steganography scheme using local complexity

The advantage of spatial domain image steganography techniques is their capacity to embed high payloads of data by directly modifying image pixels. While these techniques have a high-embedding capacity, they often create visual and statistical distortion in smoother regions. Most existing edge steganography techniques divide an image into blocks and insert data by processing the blocks in a linear order, but these method also has multiple drawbacks. First, if the selected block has an insufficient number of edge pixels, it may result in multiple blocks being processed. Second, at high embedding rates, the method creates severe distortion as multiple message bits are hidden in edge pixels and surrounding non-edge pixels without analyzing the statistical dependencies and correlation of pixels, compromising data security. The aim of the proposed method is to construct a Block-wise Edge Adaptive Steganography Scheme (BEASS) using textured regions, particularly edges and surrounding pixels. This scheme dynamically chooses the region to embed messages using a local complexity measure of Standard Deviation. It offers high payload, minimal distortion embedding by hiding three message bits into edge pixels using the minimal Mean Square Error to determine the embedding capacity of neighboring non-edge pixels within the block to preserve the statistical dependencies. The practical merit of this approach was validated and compared with existing algorithms, and experimental results find that the proposed method surpasses IQM tests, achieves a high PSNR of61 $\sim $ 65, proves to be robust against kurtosis and skewness distortion, resists histogram attack, RS steganalysis and high dimensional ensemble classifier at 80% block modifications.

Review Paper on Image Enhancement Using Various Methods

Image enhancement methods were used to modify the images that were accepted as visual perception. Image enhancement methods mainly used for image contrast. Histogram compensation, used in the form of the cumulative distribution function (CDF), is one of the challenging factors in image processing. The goal of the enhancement is to improve the structural look of an image without impairing the input image. The enhancement techniques make it easier to identify key features by removing noise and other artifacts from an image. This article discusses studies to implement image enhancement in specific areas such as satellite imagery, infrared imagery, medical imagery and digital entertainment imagery, as well as information from the image enhancement method used. Keywords: Image enhancement, AHE, Histogram Equalization, CLAHE, spatial domain image, frequency domain method, etc. Cite this Article: Manish Sharma, Ashish Dubey, Prashant Badal. Review Paper on Image Enhancement using Various Methods. International Journal of Image Processing and Pattern Recognition. 2020; 6(1): 27–35p.

Comparative and Analysis of spatial domain and frequency domaindigital image watermarking algorithm

Digital image watermarking belongs to the branch of information hiding discipline, the study of the digital watermarking will play an invaluable role in the integrity of business transactions and the protection of personal privacy.This paper introduce the embedding process of digital watermarking technology and the extraction process of digital watermarking technology. We analyze the the characteristics of digital image watermarking and juxtapose the basic principle of the Lease Significant Bits algorithm and the basic principle of the Quaternion Fourier Transform algorithm. Finally, According to analysis and comparison, we summarize the advantages of the frequency domain method.

Forensics of visual privacy protection in digital images

Visual privacy protection (VPP) is to protect individual’s privacy information in digital images or videos against being seen, such as portrait etc. Once being protected, the privacy may be imperceptible. Forensics of visual privacy protection is becoming a new challenge. As a main visual privacy protection technology, blur operation is always used in VPP. When using the existing approaches such as blur segmentation to detect the images, natural blur or other solid color regions will be falsely alarmed, resulting in low precision. In this paper, we present a novel metric invisibility degree (IvD) to measure the privacy protected blur degree of each pixel. The proposed IvD is defined by calculating the similarity between the test image and the re-blurred image in joint transformation and spatial domain, which could significantly enhance the difference between the privacy protected blur region and the other regions. Then, an effective method based on IvD is developed to automatically forensic and localize the visual privacy protection regions. Firstly, the IvD in block DCT domain of each pixel is calculated, and a IvD map is obtained. Secondly, using the IvD map, the test image is segmented and followed by morphological operation to decrease the mis-alarm regions. Finally, a spatial texture feature descriptor, including gray statistics, smoothness and information capacity, is developed, based on which the detection result is further refined. Experimental results show that the proposed method can detect the privacy regions accurately.

Synthetic CT generation from CBCT images via deep learning.

Cone-beam computed tomography (CBCT) scanning is used daily or weekly (i.e., on-treatment CBCT) for accurate patient setup in image-guided radiotherapy. However, inaccuracy of CT numbers prevents CBCT from performing advanced tasks such as dose calculation and treatment planning. Motivated by the promising performance of deep learning in medical imaging, we propose a deep U-net-based approach that synthesizes CT-like images with accurate numbers from planning CT, while keeping the same anatomical structure as on-treatment CBCT. We formulated the CT synthesis problem under a deep learning framework, where a deep U-net architecture was used to take advantage of the anatomical structure of on-treatment CBCT and image intensity information of planning CT. U-net was chosen because it exploits both global and local features in the image spatial domain, matching our task to suppress global scattering artifacts and local artifacts such as noise in CBCT. To train the synthetic CT generation U-net (sCTU-net), we include on-treatment CBCT and initial planning CT of 37 patients (30 for training, seven for validation) as the input. Additional replanning CT images acquired on the same day as CBCT after deformable registration are utilized as the corresponding reference. To demonstrate the effectiveness of the proposed sCTU-net, we use another seven independent patient cases (560 slices) for testing. We quantitatively compared the resulting synthetic CT (sCT) with the original CBCT image using deformed same-day pCT images as reference. The averaged accuracy measured by mean absolute error (MAE) between sCT and reference CT (rCT) on testing data is 18.98HU, while MAE between CBCT and rCT is 44.38 HU. The proposed sCTU-net can synthesize CT-quality images with accurate CT numbers from on-treatment CBCT and planning CT. This potentially enables advanced CBCT applications for adaptive treatment planning.

Spatiotemporal Remote-Sensing Image Fusion With Patch-Group Compressed Sensing

Generally, it is difficult to acquire remote sensing data whose resolution is both highly spatial and highly temporal from a single satellite. In this paper, a novel compressed sensing (CS)-based spatiotemporal data fusion (CSBS) method is proposed to synthesize such high-spatiotemporal resolution images. With CSBS, a low-spatial resolution remote senisng image is treated as a sampling of the high-spatial resolution image. The down-sampling in the spatial domain of images is modeled as a CS measurement matrix in CSBS. Moreover, continuity constraints in the temporal domain are also introduced into the CSBS object function for CS reconstruction. To better represent the intrinsic features of the data, images are segmented into many small patches and clustered into several groups via K-means. Dictionary training, measurement matrix identification, and high-resolution prediction are carried out group-by-group. Based on features learned from patch groups, the transformational relationship between spatial-temporal images having different resolutions are easily identified. Compared with previous compressed sensing and dictionary learning methods, CSBS is characterized by: (1) the patch-group stratagem in dictionary learning and measurement matrix learning; (2) the combination of continuity in temporal domain and sparsity in spatial domain. The proposed method is then comprehensively compared with different methods using land-surface reflectance data. Experiment results validate the effectiveness and advancement of CSBS for spatiotemporal data fusion.

Feature Vector Creation Using Hierarchical Data Structure for Spatial Domain Image Retrieval

Building efficient Image Features with optimal dimension and high discriminative power is important in processing the digital information hidden inside the images. Image Retrieval systems use the feature vectors to retrieve similar images based on feature vector similarity. Image Feature Vector generation is important step in CBIR. Most of the CBIR use Linear Data structure for Feature Vector. In this paper the Hierarchical data Structure: Binary Tree is used for Feature vector creation in Image Retrieval Systems. Images in Spatial Domain are considered for Feature Vector Generation. Low-level Feature: Color Descriptors are used as Image Contents, to represents the image in feature vector form. Performance of the image retrieval is tested for Color Images. It has been observed that, along with the Dimensionally reduction, the proposed approach of Feature Vector generation, improves the discriminative power of Feature Vector. For RGB colorspaces the cascaded statistical features, when stores in hierarchical data structure, this provides the facility to perform the calculations in parallel manner also capable to keep discriminative power comparable w.r.t size of feature vector.

Stereoscopic Video Quality Assessment Using Oriented Local Gravitational Force Statistics

We develop a new no-reference (NR) stereoscopic video quality assessment model that adopts oriented local gravitational force (OLGF) statistics in the space-time domain. The OLGF is a novel extension of an existing local gravitational force descriptor and includes two new components: relative local gravitational force magnitude and relative local gravitational force orientation. The resulting algorithm, called Stereoscopic Video Integrity Predictor using OLGF Statistics (SVIPOS), first uses our previous work to generate a cyclopean image and a product image in spatial domain from a stereoscopic video sequence pair, and calculates a frame difference image in temporal domain from a left video sequence. Then the OLGF model is deployed to compute various local gravitational force responses on the cyclopean image, the product image and the frame difference image. Finally, a support vector regression is used to map statistical features extracted from the computed gravitational force responses to stereoscopic video quality predications. SVIPOS is thoroughly tested on three publicly stereoscopic video quality databases. The experimental results show that SVIPOS outperforms state-of-the-art stereoscopic video quality methods.

Medical Image Magnification Based on Original and Estimated Pixel Selection Models

Background: The issue of medial image resolution enhancement is one of the most important topics for medical imaging that helps improve the performance of many post-processing aspects like classification and segmentation towards medical diagnosis.Objective: Our aim in this paper is to evaluate different types of pixel selection models in terms of pixel originality in medical image reconstruction problems. A previous investigation showed that selecting far original pixels has highly better performance than using near unoriginal/estimated pixels while magnifying some benchmarks in digital image processing.Material and Methods: In our technical study, we apply two classical interpolators, cubic convolution (CC) and bi-linear (BL), in order to reconstruct medical images in spatial domain. In addition to the interpolators, we use some geometrical image transforms for creating the reconstruction models.Results: The results clearly demonstrate that despite the absolute preference of the original pixel selection model in the first research, we cannot see this preference in medical dataset in which the results of BL interpolator for both tested models (original and estimated pixel selection models) are approximately the same as each other and for CC interpolator, we only see a relatively better preference for the original pixel selection model.Conclusion: The current research reveals the fact that selection models are not a general factor in reconstruction problems, and the structure of the basic interpolators is also a main factor which affects the final results. In other words, some interpolators in medical dataset can be affected by the selection models, while, some cannot.

Probabilistic Analysis of Targeted Attacks Using Transform-Domain Adversarial Examples

In the past decade, Deep Neural Networks (DNNs) have achieved breakthrough collaborations in developing smart intelligent systems within the field of computer vision, natural language processing, autonomous systems, etc. Recent research has revealed that stability of such smart systems is at greater risk when they come across to adversarial perturbations. Although, these perturbations may not be perceivable in nature when seen from naked eye, yet, they are capable enough to fool state-of-the-art DNN classifiers. Till now, much of the previous work related to fool such classifiers focuses on generating adversaries that directly change pixel values of an image in spatial-domain. In this paper, we propose a novel transform-domain imperceptible attack methodology “TDIAM” to generate adversaries based on image steganography-approach using a “single carefully selected targeted watermark”. We use three different frequency-domain approaches, i.e., Discrete Wavelet Transform (DWT), Discrete Cosine Transform (DCT) and Fast Fourier Transform (FFT) to craft perturbations in selective frequency component which makes it robust and it requires less computational time as it is a non-gradient approach. We present our case study on MNIST handwritten digits dataset. Our results demonstrate that the generated perturbation vector successfully fool simple Convolutional Neural Network (CNN), LeNet-5 and AlexNet architectures by increasing probability of adversarial examples for the targeted class (to which the targeted watermark belongs) in both “black-box” and “white-box” adversarial attacks. The results have shown that among these three perturbation approaches, DWT based perturbation shown promising results by effectively fooling DNNs while ensuring the high imperceptibility as well.

High Quality Color Image Compression for Discrete Transform Domain Downward Conversion Block Based Image Coding

Text and image data are important elements for information processing almost in all the computer applications. Uncompressed image or text data require high transmission bandwidth and significant storage capacity. Designing and compression scheme is more critical with the recent growth of computer applications. Among the various spatial domain image compression techniques, multi-level Block partition Coding (ML-BTC) is one of the best methods which has the least computational complexity. The parameters such as Peak Signal to Noise Ratio (PSNR) and Mean Square Error (MSE) are measured and it is found that the implemented methods of BTC are superior to the traditional BTC. This paves the way for a nearly error free and compressed transmission of the images through the communication channel.

CHOG Based EFD for Geometric Shape Retrieval of Images for Cloth and Object Invariant Gait Recognition

Gait refers to person identification based on the observation of human walking style. One of the prominent hurdles in gait recognition is, the challenges posed by change in apparels like clothes and object held by the subject. The paper explores the feature extraction techniques like CHOG and Elliptical Fourier Descriptors in spatial and frequency domain respectively to mitigate this negative impact on gait recognition. The CHOG behavioural feature extraction technique is used to capture the effective distribution of local gradient on gait sequence images. Further the Elliptical Fourier Descriptor (EFD) is found in frequency domain to access the geometric characteristics of a spatial domain image. The work is carried out on 36 subjects with 5 different apparels and 3 different objects each with 6 gait cycles from standard dataset CASIA SET – B and CMU - MoBo. SVM classifier is used to effectively discriminate the gait cycle patterns using optimal hyper plane. The results obtained have given an improvement of 7% to 24% increase in gait recognition over earlier techniques like GEI, CDA, LDA, ENTROPY, static and dynamic region splitting.

A robust way of steganography by using blocks of an image in spatial domain

Steganography is a way to convey secret communication, with rapid electronic communication and high demand of using the internet, steganography has become a wide field of research and discussion. In this paper a new approach for hiding information in cover image proposed in spatial domain, the proposed approach divides the host image into blocks of size (8x8) pixels and message bits are embeds into the pixels of a cover image. The 64-pixel values of each block converted to be represented in binary system and compared with corresponding secret data bits for finding the matching and hold 6-pixels. The search process performed by comparing each secret data bit (8-bits) with created binary plane at the cover image, if matching is found the last row of the created binary plane which is (LSB) is modified to indicate the location of the matched bits sequence “which is the secret data” and number of the row, if matching is not found in all 7th rows the secret sequence is copied in to the corresponding 8th row location.The payload of this technique is 6 pixels’ message (48-bits) in each block. In the experiments secret messages are randomly embedded into different images. The quality of the stego-image from which the original text message is extracted is not affected at all. For validation of the presented mechanism, the capacity, the circuit complexity, and the measurement of distortion against steganalysis is evaluated using the peak-signal-to-noise ratio (PSNR) are analyzed.

A Novel Adaptive PET/CT Image Fusion Algorithm

Background: Lung cancer is one of the common malignant tumors. The successful diagnosis of lung cancer depends on the accuracy of the image obtained from medical imaging modalities. Objective: The fusion of CT and PET is combining the complimentary and redundant information both images and can increase the ease of perception. Since the existing fusion method sare not perfect enough, and the fusion effect remains to be improved, the paper proposes a novel method called adaptive PET/CT fusion for lung cancer in Piella framework. Methods: This algorithm firstly adopted the DTCWT to decompose the PET and CT images into different components, respectively. In accordance with the characteristics of low-frequency and high-frequency components and the features of PET and CT image, 5 membership functions are used as a combination method so as to determine the fusion weight for low-frequency components. In order to fuse different high-frequency components, we select the energy difference of decomposition coefficients as the match measure, and the local energy as the activity measure; in addition, the decision factor is also determined for the high-frequency components. Results: The proposed method is compared with some of the pixel-level spatial domain image fusion algorithms. The experimental results show that our proposed algorithm is feasible and effective. Conclusion: Our proposed algorithm can better retain and protrude the lesions edge information and the texture information of lesions in the image fusion.

Payload Scaling for Adaptive Steganography: An Empirical Study

Payload-scaling laws of imperfect steganography inform the steganographer about how the size of secret payload should grow with cover size for constant statistical detectability. In this letter, we carry out an empirical study for the case when the steganographer and the steganalyst operate at a game-theoretic equilibrium. We first explore the possibility to leverage a generalization of the square root law to content-adaptive steganography due to Ker. Since this result does not appear to be tight enough for realistic cover sizes, we instead work with a detectability limited sender in image sources with a forced model as well as real images in both spatial and JPEG domain. The scaling is observed in practice when the images are carefully cropped to preserve the distribution of costs across scales.

Adaptive incremental stippling for sample distribution in spatially adaptive PIV image analysis

Adaptive sampling strategies in PIV have been shown to efficiently combine the need for limited user-dependence with increased performances in terms of spatial resolution and computational effort, thus rendering such approaches of great interest. The allocation of correlation windows across the spatial image domain is dependent on the interpretation of an underlying objective function, and the distribution of windows accordingly. It is important that such allocation is computationally efficient, robust to changing objective functions and conditions, and conducive to high quality sampling. In this paper, an alternative sample distribution method, based on adaptive incremental stippling, is presented and shown to combine the speed of PDF-based methods with the quality of ‘ideal’ spring-force methods. Case-dependent parameter tuning is no longer necessary, thus improving robustness. In addition, an algorithm to adaptively size initial correlation windows is proposed to further minimise user dependence.

Seamless and non-repetitive 4D texture variation synthesis and real-time rendering for measured optical material behavior

We show how to overcome the single weakness of an existing fully automatic system for acquisition of spatially varying optical material behavior of real object surfaces. While the expression of spatially varying material behavior with spherical dependence on incoming light as a 4D texture (an ABTF material model) allows flexible mapping onto arbitrary 3D geometry, with photo-realistic rendering and interaction in real time, this very method of texture-like representation exposes it to common problems of texturing, striking in two disadvantages. Firstly, non-seamless textures create visible artifacts at boundaries. Secondly, even a perfectly seamless texture causes repetition artifacts due to their organised placement in large numbers over a 3D surface. We have solved both problems through our novel texture synthesis method that generates a set of seamless texture variations randomly distributed over the surface at shading time. When compared to regular 2D textures, the inter-dimensional coherence of the 4D ABTF material model poses entirely new challenges to texture synthesis, which includes maintaining the consistency of material behavior throughout the 4D space spanned by the spatial image domain and the angular illumination hemisphere. In addition, we tackle the increased memory consumption caused by the numerous variations through a fitting scheme specifically designed to reconstruct the most prominent effects captured in the material model.

A multi-scale data fusion framework for bone age assessment with convolutional neural networks

Bone age assessment (BAA) has various clinical applications such as diagnosis of endocrine disorders and prediction of final adult height for adolescents. Recent studies indicate that deep learning techniques have great potential in developing automated BAA methods with significant advantages over the conventional methods based on handcrafted features. In this paper, we propose a multi-scale data fusion framework for bone age assessment with X-ray images based on non-subsampled contourlet transform (NSCT) and convolutional neural networks (CNNs). Unlike the existing CNN-based BAA methods that adopt the original spatial domain image as network input directly, we pre-extract a rich set of features for the input image by performing NSCT to obtain its multi-scale and multi-direction representations. This feature pre-extraction strategy could be beneficial to network training as the number of annotated examples in the problem of BAA is typically quite limited. The obtained NSCT coefficient maps at each scale are fed into a convolutional network individually and the information from different scales are then merged to achieve the final prediction. Specifically, two CNN models with different data fusion strategies are presented for BAA: a regression model with feature-level fusion and a classification model with decision-level fusion. Experiments on the public BAA dataset Digital Hand Atlas demonstrate that the proposed method can obtain promising results and outperform many state-of-the-art BAA methods. In particular, the proposed approaches exhibit obvious advantages over the corresponding spatial domain approaches (generally with an improvement of more than 0.1 years on the mean absolute error), showing great potential in the future study of this field.

Progressive and hierarchical share-in-share scheme over cloud

Recently sharing of secret information (image) before dissemination is gaining popularity as a means of reducing the vulnerability against the wide attacking surface on cloud computing. An integrated scheme comprising of secret image sharing (SIS) and data hiding is suggested here that offers progressive quality access control of the decoded images. This scheme is a (t,k,n)-hierarchical SIS technique that uses affine Boolean classification and compressed sensing (CS) in the share generation process. Prediction error representation of the secret image in spatial domain is explored to make the signal more sparse leading to CS reconstruction from less number of measurements. In the decoding process, an improved quality of the secret image is achieved with the involvement of ‘t’ essential stego shares (ESSs) and ‘k-t’ non-essential stego shares (NSSs) over ‘t’ ESSs and ‘n-k’ general shares (GSs). A large set of results illustrate the merits of the proposed method in terms of hierarchical access structure, flexible shadow size, progressive quality access on the decoded image and robustness against varieties of data manipulation.