Computer Image Research Articles

Nineteen Eighty-Four in 1974: the surveillance imaginary of the New Mexico Civil Liberties Union Campaign for privacy awareness

ABSTRACT This article maps the surveillance imaginary of a unique public awareness initiative by the New Mexico Civil Liberties Union and Godfrey Reggio’s Institute for Regional Education in 1974 that consisted of two parts: a situationist-inspired campaign using and surrealistically subverting popular media to raise awareness of surveillance, paired with a newspaper supplement documenting detailed intrusions on the privacy and liberty of citizens on a number of fronts, representing an encroaching society of social control. Taken as a whole, the surveillance imaginary of the campaign combined images of eyes, computers, and rats in mazes with detailed discussions of databases, militarized policing, and direct behavioral control justified by stories of rising violence in society to critique both state surveillance and the growing corporatization of the public sphere.

Enhancing Spatial Resolution in Tandem Mass Spectrometry Ion/Ion Reaction Imaging Experiments through Image Fusion.

We have recently developed a charge inversion ion/ion reaction to selectively derivatize phosphatidylserine lipids via gas-phase Schiff base formation. This tandem mass spectrometry (MS/MS) workflow enables the separation and detection of isobaric lipids in imaging mass spectrometry, but the images acquired using this workflow are limited to relatively poor spatial resolutions due to the current time and limit of detection requirements for these ion/ion reaction imaging mass spectrometry experiments. This trade-off between chemical specificity and spatial resolution can be overcome by using computational image fusion, which combines complementary information from multiple images. Herein, we demonstrate a proof-of-concept workflow that fuses a low spatial resolution (i.e., 125 μm) ion/ion reaction product ion image with higher spatial resolution (i.e., 25 μm) ion images from a full scan experiment performed using the same tissue section, which results in a predicted ion/ion reaction product ion image with a 5-fold improvement in spatial resolution. Linear regression, random forest regression, and two-dimensional convolutional neural network (2-D CNN) predictive models were tested for this workflow. Linear regression and 2D CNN models proved optimal for predicted ion/ion images of PS 40:6 and SHexCer d38:1, respectively.

A Multi-Level Interpretable Sleep Stage Scoring System by Infusing Experts' Knowledge Into a Deep Network Architecture.

In recent years, deep learning has shown potential and efficiency in a wide area including computer vision, image and signal processing. Yet, translational challenges remain for user applications due to a lack of interpretability of algorithmic decisions and results. This black box problem is particularly problematic for high-risk applications such as medical-related decision-making. The current study goal was to design an interpretable deep learning system for time series classification of electroencephalogram (EEG) for sleep stage scoring as a step toward designing a transparent system. We have developed an interpretable deep neural network that includes a kernel-based layer guided by a set of principles used for sleep scoring by human experts in the visual analysis of polysomnographic records. A kernel-based convolutional layer was defined and used as the first layer of the system and made available for user interpretation. The trained system and its results were interpreted in four levels from microstructure of EEG signals, such as trained kernels and effect of each kernel on the detected stages, to macrostructures, such as transitions between stages. The proposed system demonstrated greater performance than prior studies and the system learned information consistent with expert knowledge.

MP-Net: A Multi-Center Privacy-Preserving Network for Medical Image Segmentation.

In this paper, we present the Multi-Center Privacy-Preserving Network (MP-Net), a novel framework designed for secure medical image segmentation in multi-center collaborations. Our methodology offers a new approach to multi-center collaborative learning, capable of reducing the volume of data transmission and enhancing data privacy protection. Unlike federated learning, which requires the transmission of model data between the central server and local servers in each round, our method only necessitates a single transfer of encrypted data. The proposed MP-Net comprises a three-layer model, consisting of encryption, segmentation, and decryption networks. We encrypt the image data into ciphertext using an encryption network and introduce an improved U-Net for image ciphertext segmentation. Finally, the segmentation mask is obtained through a decryption network. This architecture enables ciphertext-based image segmentation through computable image encryption. We evaluate the effectiveness of our approach on three datasets, including two cardiac MRI datasets and a CTPA dataset. Our results demonstrate that the MP-Net can securely utilize data from multiple centers to establish a more robust and information-rich segmentation model.

Detecting antimony(III) on-site using novel gel-based techniques: Colorimetric diffusive equilibrium in thin films for two-dimensional imaging and surface-enhanced Raman scattering for sensitive quantification

Antimony (Sb) pollution has raised increasing public concerns and its rapid on-site screening is central for the risk assessment. Herein, we proposed two gel-based methods based on colorimetric diffusive equilibrium in thin films (DET) and surface-enhanced Raman scattering (SERS), for two-dimensional imaging and sensitive detection of Sb(III) by revisiting the phenylfluorone (PhF) complexation reaction. PhF was well dispersed in the polyvinyl alcohol (PVA) hydrogel and reacted with Sb(III) in the DET gel to form a strong PhF-Sb(III) complex. The distribution of Sb(III) was easily visualized at a submillimeter resolution using computer imaging densitometry, with a detection limit (LOD) of ∼100 nmol L−1. Field application in the Sb mine area reveals limited dissolved Sb(III) penetrating the redox barrier below the sediment-water interface by 20 mm in rivers and tailing pond sediments. To improve the detection sensitivity and apply the principle to trace Sb quantification, a SERS platform was established by anchoring PhF on the hydrogel-stabilized Ag nanoparticles via C–O–Ag bonding to specifically detect Raman-inactive Sb(III). Benefiting from the high SERS activity of PhF and enrichment ability of hydrogel, Sb(III) was quantified with a LOD of 1.2–10.7 nmol L−1 depending on the sample volume. The coexisting ions at a 100-fold higher concentration than Sb(III) resulted in only 3.3–10.4 % variation in SERS intensity, indicating a negligible interference on the SERS platform. The platform exhibited a RSD of 6.6–13.1 % and acceptable recoveries for various environmental matrices, highlighting its promise in on-site application.

A New Method for Evaluating the Homogeneity within and between Weave Repeats in Plain Fabric Structures Using Computer Image Analysis.

This article introduces a novel, rapid, and non-destructive method for assessing homogeneity within and between weave repeats in fabric structures, termed intra-repeat (IAR) and inter-repeat (IER) evaluation. The method focuses on structural parameters, including inter-thread pores (ITPs) and warp and weft pitches, using computer image analysis. Each parameter is assigned to a module in the repeat weave pattern, facilitating the sorting of modules in the IAR and IER fabric structure arrangement. The method was verified using artificial images and 30 real plain fabrics with varying degrees of warp grouping, employing the author's proprietary software, MagFABRIC version 2.1The general measurable coefficients of intra- and inter-homogeneity were defined and related to the airflow measurements of these fabrics. Multiple regression models of airflow revealed strong dependencies, particularly for F = 10, with the size, shape, and position of ITPs and warp and weft pitches showing significant correlation. These findings underscore the importance of the new homogeneity parameters in textile structure analysis, including both IAR and IER woven fabric structure homogeneity parameters. The research aims to model specialized fabrics (e.g., barrier, filtration, composite fabrics) to address local changes in fabric structure affecting properties such as filtration efficiency, air permeability, and mechanical properties, especially in applications like composites or medical implants.

Landmark Detection Using Machine Learning

As smartphones and mobile data become more prevalent in modern society, the possibilities for them to interact with the physical world also grow exponentially. Technologies such as Oculus Rift and Google Glass are attempting to bridge the gap between the virtual and the physical, and as enhancements in computer speed and image processing are made, the concept of Augmented Reality (AR) becomes more tangible. However, one difficulty with AR is the sheer complexity of image processing and feature recognition. A successful AR system must be able to distinguish among a large number of landmarks and should be able to adapt to the existence of new landmarks. Because of the adaptability requirement, AR algorithms naturally lend themselves to using machine learning. As such, the focus of this project is to develop, refine and document a machine learning algorithm that can distinguish landmarks from images using a database of known landmarks.

Image Segmentation Techniques: An In-Depth Review and Analysis

In the fields of computer vision and image processing, image segmentation is a very important task. The image is partitioned into segments or regions then the visual data can be understood, analyzed and used easily. In this article, a comprehensive review of image segmentation methods is presented. It covers both the strengths and the advantages of some techniques as well as the weaknesses and limitations of other algorithms. The techniques used in image segmentation can be divided into four groups based on thresholding, region, edge and deep learning techniques. The fundamental principles, popularity of algorithms and evaluation metrics are mentioned. Also, the applications domains and future directions with current challenges are presented.

HDMA-CGAN: Advancing Image Style Transfer with Deep Learning

The integration of artificial intelligence (AI) and deep learning heralds a transformative era in pattern recognition and computer vision, notably in image style transfer. We introduce the hierarchical dynamic multi-attention cycle generative adversarial network (HDMA-CGAN), an innovative deep learning architecture poised to redefine image style transfer capabilities. HDMA-CGAN employs a novel multi-attention mechanism and color optimization strategies, enabling precise style replication with improved fidelity and vibrancy. Our model surpasses existing benchmarks in image quality, validated by leading metrics such as peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), and Fréchet inception distance (FID). Although HDMA-CGAN advances the state of the art, it necessitates high computational resources and faces challenges with very high-resolution images. Future work could explore optimizing the model’s efficiency for real-time applications and extending its application to video content. This work enhances the tools available for visual content creation and digital media enhancement, leveraging advanced pattern recognition and AI techniques to significantly impact computer vision and image processing.

A W-Shaped Self-Supervised Computational Ghost Imaging Restoration Method for Occluded Targets.

We developed a novel method based on self-supervised learning to improve the ghost imaging of occluded objects. In particular, we introduced a W-shaped neural network to preprocess the input image and enhance the overall quality and efficiency of the reconstruction method. We verified the superiority of our W-shaped self-supervised computational ghost imaging (WSCGI) method through numerical simulations and experimental validations. Our results underscore the potential of self-supervised learning in advancing ghost imaging.

The Renin-Angiotensin-Aldosterone System Regulates Sarcoidosis Granulomatous Inflammation.

Rationale: Sarcoidosis is a granulomatous disorder of unclear cause notable for abnormal elevation of blood and tissue ACE1 (angiotensin converting enzyme 1) levels and activity. ACE1 regulates the renin-angiotensin-aldosterone system (RAAS), the terminal product of which is aldosterone, which selectively engages mineralocorticoid receptors to promote inflammation. Objectives: We sought to determine whether the RAAS promotes sarcoidosis granuloma formation and related inflammatory responses. Methods: Using an established ex vivo model, we first determined whether aldosterone was produced by sarcoidosis granulomas and verified the presence of CYP11B2, the enzyme required for its production. We then evaluated the effects of selective inhibitors of ACE1 (captopril), angiotensin type 1 receptor (losartan), and mineralocorticoid receptors (spironolactone, eplerenone) on granuloma formation, reflected by computer image analysis-generated granuloma area, and selected cytokines incriminated in sarcoidosis pathogenesis. Measurements and Main Results: Aldosterone was spontaneously produced by sarcoidosis peripheral blood mononuclear cells, and both intra- and extracellular levels steadily increased during granuloma formation. In parallel, peripheral blood mononuclear cells were shown to express more CYP11B2 during granuloma formation. Significant inhibition of sarcoidosis granulomas and related cytokines (TNFα, IL-1β, IFNγ, IL-10) was observed in response to pretreatments with captopril, losartan, spironolactone, or eplerenone, comparable to that of prednisone. Conclusions: The RAAS is intact in sarcoidosis granulomas and contributes significantly to early granuloma formation and to related inflammatory mediator responses, with important implications for clinical management.

Evaluation of Parasight All-in-One system for the automated enumeration of helminth ova in canine and feline feces

Background Digital imaging combined with deep-learning-based computational image analysis is a growing area in medical diagnostics, including parasitology, where a number of automated analytical devices have been developed and are available for use in clinical practice.MethodsThe performance of Parasight All-in-One (AIO), a second-generation device, was evaluated by comparing it to a well-accepted research method (mini-FLOTAC) and to another commercially available test (Imagyst). Fifty-nine canine and feline infected fecal specimens were quantitatively analyzed by all three methods. Since some samples were positive for more than one parasite, the dataset consisted of 48 specimens positive for Ancylostoma spp., 13 for Toxocara spp. and 23 for Trichuris spp.ResultsThe magnitude of Parasight AIO counts correlated well with those of mini-FLOTAC but not with those of Imagyst. Parasight AIO counted approximately 3.5-fold more ova of Ancylostoma spp. and Trichuris spp. and 4.6-fold more ova of Toxocara spp. than the mini-FLOTAC, and counted 27.9-, 17.1- and 10.2-fold more of these same ova than Imagyst, respectively. These differences translated into differences between the test sensitivities at low egg count levels (< 50 eggs/g), with Parasight AIO > mini-FLOTAC > Imagyst. At higher egg counts Parasight AIO and mini-FLOTAC performed with comparable precision (which was significantly higher that than Imagyst), whereas at lower counts (> 30 eggs/g) Parasight was more precise than both mini-FLOTAC and Imagyst, while the latter two methods did not significantly differ from each other.ConclusionsIn general, Parasight AIO analyses were both more precise and sensitive than mini-FLOTAC and Imagyst and quantitatively correlated well with mini-FLOTAC. While Parasight AIO produced lower raw counts in eggs-per-gram than mini-FLOTAC, these could be corrected using the data generated from these correlations.Graphical

Anisotropic regularization for sparsely sampled and noise-robust Fourier ptychography

Fourier ptychography (FP) is a powerful computational imaging technique that provides super-resolution and quantitative phase imaging capabilities by scanning samples in Fourier space with angle-varying illuminations. However, the image reconstruction in FP is inherently ill-posed, particularly when the measurements are noisy and have insufficient data redundancy in the Fourier space. To improve FP reconstruction in high-throughput imaging scenarios, we propose a regularized FP reconstruction algorithm utilizing anisotropic total variation (TV) and Tikhonov regularizations for the object and pupil functions, respectively. To solve this regularized FP problem, we formulate a reconstruction algorithm using the alternating direction method of multipliers and show that our approach successfully recovers high-quality images with sparsely sampled and/or noisy measurements. The results are quantitatively and qualitatively compared against various FP reconstruction algorithms to analyze the effect of regularization under harsh imaging conditions. In particular, we demonstrate the effectiveness of our method on the real experimental FP microscopy images, where the TV regularizer effectively suppresses the measurement noise while maintaining the edge information in the biological specimen and helps retrieve the correct amplitude and phase images even under insufficient sampling.

Can deep neural networks work with amplitude and phase input of defocused images?

Deep neural network (DNN) models, particularly convolutional neural networks (CNNs), have demonstrated remarkable performance in biomedical image classification due to their ability to automatically learn features from large datasets. One common challenge in the preparation of large, microscopic datasets for DNN tasks is sample defocusing, potentially impairing the model performance. To handle defocusing, computational imaging, or specifically quantitative phase imaging (QPI), performs digital refocusing by using both the phase and the amplitude of the complex optical field. This leads us to investigate whether feeding the complex field into DNN would potentially address the defocusing problem as in-focus information is implicitly encoded in the complex field. In this paper, we assess the feasibility of employing neural networks to directly process full amplitude and phase data from a defocus plane without digital refocusing. Our specific focus lies in understanding the tolerance for defocus in image classification neural networks when amplitude and phase are taken as inputs. To accomplish this, we used Fourier ptychography microscopy (FPM) to acquire in-focus phase and amplitude images of two distinct object types – normal red blood cells and echinocytes. We then digitally propagate the complex field to generate progressively defocused images of the samples to serve as training and testing datasets for image classification neural networks. While the digitally defocused images contain sufficient information to recover the original in-focus images, we observed that current standard implementations of deep learning models are unable to effectively utilize the defocused field to distinguish between the two cell types. We conclude that the physical-model-based digital refocusing capability of QPI remains indispensable for overcoming defocusing issues in current standard DNN models.

Computational image analysis of distortion, sharpness, and depth of field in a next-generation hybrid exoscopic and microsurgical operative platform.

The development of surgical microscope-associated cameras has given rise to a new operating style embodied by hybrid microsurgical and exoscopic operative systems. These platforms utilize specialized camera systems to visualize cranial neuroanatomy at various depths. Our study aims to understand how different camera settings in a novel hybrid exoscope system influence image quality in the context of neurosurgical procedures. We built an image database using captured cadaveric dissection images obtained with a prototype version of a hybrid (microsurgical/exoscopic) operative platform. We performed comprehensive 4K-resolution image capture using 76 camera settings across three magnification levels and two working distances. Computer algorithms such as structural similarity (SSIM) and mean squared error (MSE) were used to measure image distortion across different camera settings. We utilized a Laplacian filter to compute the overall sharpness of the acquired images. Additionally, a monocular depth estimation deep learning model was used to examine the image's capability to visualize the depth of deeper structures accurately. A total of 1,368 high-resolution pictures were captured. The SSIM index ranged from 0.63 to 0.85. The MSE was nearly zero for all image batches. It was determined that the exoscope could accurately detect both the sharpness and depth based on the Laplacian filter and depth maps, respectively. Our findings demonstrate that users can utilize the full range of camera settings available on the exoscope, including adjustments to aperture, color saturation, contrast, sharpness, and brilliance, without introducing significant image distortions relative to the standard mode. The evolution of the camera incorporated into a surgical microscope enables exoscopic visualization during cranial base surgery. Our result should encourage surgeons to take full advantage of the exoscope's extensive range of camera settings to match their personal preferences or specific clinical requirements of the surgical scenario. This places the exoscope as an invaluable asset in contemporary surgical practice, merging high-definition imaging with ergonomic design and adaptable operability.

A crosstalk-free multiple-image encryption scheme based on computational ghost imaging with binarized detection

Crosstalk noise is a main problem limiting the performance of multiple-image encryption (MIE) scheme. In this work, we proposed a crosstalk-free MIE scheme based on computational ghost imaging (CGI) with binarized detection. In the encryption process, the plaintext images are encrypted into intensity sequences by the CGI system and quantified into two levels to obtain binary ciphertext sequences, which does not cause severe degradation in decrypted image quality compared to traditional CGI. Then, for the binary ciphertext sequences, we can combine them into a decimal grayscale ciphertext. To enhance security, a pixel bit layer scrambling (PBLS) algorithm is designed to scramble the grayscale ciphertext to obtain the final ciphertext. In the decryption process, anyone of the plaintext images can be decrypted without being affected by other plaintext images after performing inverse PBLS algorithm on the ciphertext and extracting the binary ciphertext sequence. The effectiveness, robustness, encryption capacity and security of the proposed scheme are demonstrated by numerical simulations and theoretical analysis.

Sparse-angle optical projection tomography based on multi-layer sparsity and deep image priors

Optical projection tomography (OPT) is a computational imaging technique to acquire the volumetric images of biological samples ranging from millimeters to centimeters. For in-vivo OPT, it is essential to minimize the inspection time to reduce the adverse impacts on organisms, including the anesthetic side effect and phototoxicity. It can be achieved by projecting the samples from equally spaced sparse angles, but this method will induce radial artifacts in the reconstructed tomographic images. This paper develops a high-quality reconstruction method for sparse-angle OPT by jointly exploiting the multi-layer sparsity prior and deep image prior (DIP) on the volumetric images. The DIP module works in an unsupervised manner without requirement on a training dataset. This method can also address the inter-layer correlation within the samples, and process multi-layer images in parallel to improve the reconstruction accuracy and efficiency. Simulations and experiments demonstrate the superiority of the proposed method over some widely used reconstruction algorithms for sparse-angle OPT.

Enhancement of Three-Dimensional Computational Integral Imaging via Post-Processing with Visibility Coefficient Estimation

Enhancement of Three-Dimensional Computational Integral Imaging via Post-Processing with Visibility Coefficient Estimation

An approach-based machine learning and automated thermal images to predict the dark-cutting incidence in cattle management of healthcare supply chain

The healthcare supply chain is a network made up of various systems, processes, and elements that function and interact seamlessly to offer healthcare services and products. Food safety is an essential component of the healthcare supply chain. In the cattle industry, the healthcare supply chain contributes positively to the global economy through providing high-quality products such as milk and meat. Stress in cattle is one of main factor that cause low quality meat called “dark meat”. Numerous studies have been conducted on the development of different non-invasive methods based on Infrared Thermography Technology (IRT) to enhance the meat quality by detecting stress in cattle pre-slaughtering. These studies have the following issues: lack of automating in detecting body temperature of cattle and ignoring detecting stress with prediction dark meat incidence. The present study endeavors a new fully automated system for detecting stress, and dark meat incidence, incorporating the following new approaches: Multiview face detecting, Automatic eye localisation for detecting body temperature automatically employing computer vision and image processing, respectively. Furthermore, machine learning algorithms like Support Vector Machine (SVM), Logistic Regression (LR), Naïve Bayes (NB), and Decision Tree (DT) have been developed specifically for stress detection and the prediction of dark meat. To develop automated system, two forms of data were collected: statistical temperature and infrared thermal images. Infrared thermal images are used to develop Multiview face detection and Automatic eye localisation. Temperature data used to develop the machine learning model. Results reveal that Multiview face detection better than the current methods in term of Precision 0.99, Recall 0.91, F-score 0.95 with high True positive rate 0.90 and zero False-positive rate. Automatic eye localisation has high accuracy, with the following values: sensitivity 0.9780, precision 0.7212, F measure of 0.8024, and misclassification 0.0455. Lastly, results elaborate that the decision tree model can attain a notable level of accuracy in terms of specificity, recall, F-measure, and Area Under the Curve (AUC), all at an optimal rate of 98%.

Comparison of Velocity Profiles in Stented Carotid Artery Bifurcation Between Computational Fluid Dynamics and Particle Image Velocimetry Measurements

Comparison of Velocity Profiles in Stented Carotid Artery Bifurcation Between Computational Fluid Dynamics and Particle Image Velocimetry Measurements