Local Contrast Research Articles

Hierarchical Perceptual Noise Injection for Social Media Fingerprint Privacy Protection.

Billions of people share images from their daily lives on social media every day. However, their biometric information (e.g., fingerprints) could be easily stolen from these images. The threat of fingerprint leakage from social media has created a strong desire to anonymize shared images while maintaining image quality, since fingerprints act as a lifelong individual biometric password. To guard the fingerprint leakage, adversarial attack that involves adding imperceptible perturbations to fingerprint images have emerged as a feasible solution. However, existing works of this kind are either weak in black-box transferability or cause the images to have an unnatural appearance. Motivated by the visual perception hierarchy (i.e., high-level perception exploits model-shared semantics that transfer well across models while low-level perception extracts primitive stimuli that result in high visual sensitivity when a suspicious stimulus is provided), we propose FingerSafe, a hierarchical perceptual protective noise injection framework to address the above mentioned problems. For black-box transferability, we inject protective noises into the fingerprint orientation field to perturb the model-shared high-level semantics (i.e., fingerprint ridges). Considering visual naturalness, we suppress the low-level local contrast stimulus by regularizing the response of the Lateral Geniculate Nucleus. Our proposed FingerSafe is the first to provide feasible fingerprint protection in both digital (up to 94.12%) and realistic scenarios (Twitter and Facebook, up to 68.75%). Our code can be found at https://github.com/nlsde-safety-team/FingerSafe.

Bisection Window and Homogeneity Principle- Based Local Contrast Measure for Infrared Sea–Sky Line Detection

Bisection Window and Homogeneity Principle- Based Local Contrast Measure for Infrared Sea–Sky Line Detection

Quantitative assessment of the visual quality of digital images based on the laws of human visual perception

The existing methods of quantitative assessment of the visual quality of digital images are studied. Among the main shortcomings of the studied methods, the following can be singled out. Most of them require a reference image, do not include all the components that affect visual quality and do not take into consideration the laws of human visual perception. It was decided to develop a method for quantitative assessment of the visual quality of images, which will work without a reference image and will take into account the regularities of human visual perception. There are characterized the main regularities of human visual perception, which are used in the development of the technique. A classification of the researched methods of quantitative assessment of image quality is proposed for structuring their analysis. It was decided to investigate methods of quantitative assessment of quality based on statistical analysis of image pixel intensities. There are described factors affecting the quality of images and methods of their control based on changes in the pixel intensity distribution histogram. A generalized expression of quantitative quality assessment based on moments is proposed. A methodology for quantitative assessment of the visual quality of images has been developed, which does not require a reference image and is based on the laws of human visual perception. This method was tested on an image processed by local contrast enhancement and low-pass filtering. The test results showed that the visual perception of image quality coincides with the quantitative assessment of its quality. It is possible to use the proposed method with some modifications to determine the quality of color images. Moreover, a potential avenue for advancing the proposed method involves adapting it for evaluating images afflicted by distortions induced by noise presence.

Infrared small target detection based on ring local contrast measure with edge suppression

Infrared small target detection based on ring local contrast measure with edge suppression

Infrared Small Target Detection Based on Local Contrast Measure With a Flexible Window

Infrared Small Target Detection Based on Local Contrast Measure With a Flexible Window

Gray Level Homogeneity Analysis: A Novel Approach

In this article, we propose a method that helps us to analyze the homogeneity of gray levels locally by calculating a coefficient for each pixel based on the nature of neighboring pixels. This principle of encoding pixels according to their adjacent neighbors is described the nature of the distribution of gray levels within the image and measures their degree of homogeneity locally. This allows us to detect the different regions of the image and their contours based on the coefficient of homogeneity of the gray levels. In addition, this allows us to exploit these homogeneity coefficients to restructure regions of the image, extract and enhance the image contours while reducing the noise present in the image. This homogeneity study principle has several functions in the study and analysis of image texture, as do other methods of homogeneity assessment, such as the local contrast descriptor (LCD) and the co-occurrence matrix.

MSCT-UNET: multi-scale contrastive transformer within U-shaped network for medical image segmentation

Objective. Automatic mutli-organ segmentation from anotomical images is essential in disease diagnosis and treatment planning. The U-shaped neural network with encoder–decoder has achieved great success in various segmentation tasks. However, a pure convolutional neural network (CNN) is not suitable for modeling long-range relations due to limited receptive fields, and a pure transformer is not good at capturing pixel-level features. Approach. We propose a new hybrid network named MSCT-UNET which fuses CNN features with transformer features at multi-scale and introduces multi-task contrastive learning to improve the segmentation performance. Specifically, the multi-scale low-level features extracted from CNN are further encoded through several transformers to build hierarchical global contexts. Then the cross fusion block fuses the low-level and high-level features in different directions. The deep-fused features are flowed back to the CNN and transformer branch for the next scale fusion. We introduce multi-task contrastive learning including a self-supervised global contrast learning and a supervised local contrast learning into MSCT-UNET. We also make the decoder stronger by using a transformer to better restore the segmentation map. Results. Evaluation results on ACDC, Synapase and BraTS datasets demonstrate the improved performance over other methods compared. Ablation study results prove the effectiveness of our major innovations. Significance. The hybrid encoder of MSCT-UNET can capture multi-scale long-range dependencies and fine-grained detail features at the same time. The cross fusion block can fuse these features deeply. The multi-task contrastive learning of MSCT-UNET can strengthen the representation ability of the encoder and jointly optimize the networks. The source code is publicly available at: https://github.com/msctunet/MSCT_UNET.git.

Normalization by orientation-tuned surround in human V1-V3.

An influential account of neuronal responses in primary visual cortex is the normalized energy model. This model is often implemented as a multi-stage computation. The first stage is linear filtering. The second stage is the extraction of contrast energy, whereby a complex cell computes the squared and summed outputs of a pair of the linear filters in quadrature phase. The third stage is normalization, in which a local population of complex cells mutually inhibit one another. Because the population includes cells tuned to a range of orientations and spatial frequencies, the result is that the responses are effectively normalized by the local stimulus contrast. Here, using evidence from human functional MRI, we show that the classical model fails to account for the relative responses to two classes of stimuli: straight, parallel, band-passed contours (gratings), and curved, band-passed contours (snakes). The snakes elicit fMRI responses that are about twice as large as the gratings, yet a traditional divisive normalization model predicts responses that are about the same. Motivated by these observations and others from the literature, we implement a divisive normalization model in which cells matched in orientation tuning ("tuned normalization") preferentially inhibit each other. We first show that this model accounts for differential responses to these two classes of stimuli. We then show that the model successfully generalizes to other band-pass textures, both in V1 and in extrastriate cortex (V2 and V3). We conclude that even in primary visual cortex, complex features of images such as the degree of heterogeneity, can have large effects on neural responses.

Exploring the restorative benefits of work in smart working structures on vacations in small villages.

Processes of redefinition of work, already in place in the pre-pandemic era, with the advent of COVID-19, have become widely required. A "model of work" that uses new technologies and the development of existing ones to improve workers' performance and satisfaction has emerged. Smart working has changed how people work and, as a result, spaces must also change to support them. The use of prefabricated movable buildings (PMB) could represent an opportunity to create smart (and co-) working spaces in a regenerative contest. Small villages with historical, architectural, and naturalistic elements and slow-life features are potential platforms where vacation and work could merge easily. This paper aims to establish if working in movable offices, like the PMBs, which provide a high level of visual and acoustics interaction with the surroundings, and which is positioned in small villages' squares, can affect workers' sense of restoration and working performance. In a laboratory setting, in the SENS i-Lab, a videowall and a 3D spatial audio system were used to compare, in terms of restorativeness and self-reported cognitive performance, the effects of a high-rise building context of a City Business District (CBD), i.e. the control scenario, and of two typical squares of small Italian villages with historic buildings, without (HIS) and with water elements (HIS-W). The findings of the experimental sessions showed that when working in close contact (visual and auditory) with historical or water elements, where life flows slowly, workers perceived a higher sense of restoration while job performance was unchanged. Implementing smart working policies in small villages that encourage the use of energy-efficient prefabricated movable buildings and that offer a high level of visual and acoustic interaction with the surrounding environment may represent a promising strategy to foster the development of the local economy and contrast the depopulation of small villages, improving the worker well-being and the reducing the impact of their activities of the environment.

Biomass Burning Emissions of Black Carbon over the Maritime Continent and ENSO Variability

Abstract Fire emissions from the Maritime Continent (MC) over the western tropical Pacific are strongly influenced by El Niño–Southern Oscillation (ENSO), posing various climate effects to the Earth system. In this study, we show that the historical biomass burning emissions of black carbon (BCbb) aerosol in the dry season from the MC are strengthened in El Niño years due to the dry conditions. The eastern Pacific type of El Niño exerts a stronger modulation in BCbb emissions over the MC region than the central Pacific type of El Niño. Based on simulations using the fully coupled Community Earth System Model (CESM), the impacts of increased BCbb emissions on ENSO variability and frequency are also investigated in this study. With BCbb emissions from the MC scaled up by a factor of 10, which enables the identification of climate response from the internal variability, the increased BCbb heats the local atmosphere and changes land–sea thermal contrast, which suppresses the westward transport of the eastern Pacific surface water. It leads to an increase in sea surface temperature in the eastern tropical Pacific, which further enhances ENSO variability and increases the frequency of extreme El Niño and La Niña events. This study highlights the potential role of BCbb emissions on extreme ENSO frequency, and this role may be increasingly important in the warming future with higher wildfire risks.

Volume of contrast‐based Substantia Nigra segmentation is decreased in dementia patients compared to healthy controls

AbstractBackgroundThe substantia nigra (SN) is susceptible to neurodegeneration in Alzheimer’s disease. Automatic segmentation of the SN using MRI permits in vivo morphological assessment of the SN in cohort studies. Here we evaluate the multi‐brain segmentation (MB, Brudfors, 2020) to learn a probabilistic tissue prior of the SN using neuromelanin‐sensitive MRI while systematically studying effects of (A) SN tissue contrast; and (B) volumetric differences between healthy subjects and memory patients.MethodWe studied 50 3T single‐site T1w FLASH MRI from cognitively healthy older adults from the DZNEDELCODE cohort and a second ageing cohort in Magdeburg and 25 multi‐site diagnosed memory patients (MCI or AD). We quantified local SN signal contrast and image noise as a quality metric (obtained from CAT/SPM toolbox). We learnt tissue priors for each 10, 20, 30 and 50 older adults with the highest SN tissue contrast. For each tissue prior, we obtained model‐based SN tissue segments and calculated (integrated) SN volume. We compared SN volume between tissue prior segmentations and correlated it to age and quality. We coregistered all patients to the N = 50 tissue prior and compared volumetric differences between healthy adults and patients using voxel‐based morphometry (VBM).ResultThe template‐space N = 50 SN prior (in red) was visually larger than all other priors (Fig 1A). Patients visually had a lower SN volume than healthy from N = 50 (Fig 1B). The N = 50 native space SN tissue prior volume was significantly higher than all other tissue prior volumes (p>0.01, Fig 2A). SN volume correlated with quality for N = 50 (p = 0.02, Fig 2B). The N = 50 sample SN contrast (Fig 2C) did not differ between patients and healthy adults. Patients had a decreased volume in a cluster overlapping with the SN (N = 268, pFDR corrected = 0.02 (Fig 2D).ConclusionHere we demonstrate that model‐based learning of a probabilistic SN tissue prior using intensity‐based generative approach to segmentation is feasable in a heterogenous cohort. Additionally, using this prior, we showed that while there is no difference of the SN contrast, there is a significant decrease in SN volume consistent with previous analyses (Krohn et al 2022).

Multimodal Medical Image Fusion Utilizing Two-scale Image Decomposition via Saliency Detection.

Modern medical imaging modalities used by clinicians have many applications in the diagnosis of complicated diseases. These imaging technologies reveal the internal anatomy and physiology of the body. The fundamental idea behind medical image fusion is to increase the image's global and local contrast, enhance the visual impact, and change its format so that it is better suited for computer processing or human viewing while preventing noise magnification and accomplishing excellent real-time performance. The top goal is to combine data from various modal images (CT/MRI and MR-T1/MR-T2) into a solitary image that, to the greatest degree possible, retains the key characteristics (prominent features) of the source images. The clinical accuracy of medical issues is compromised because innumerable classical fusion methods struggle to conserve all the prominent features of the original images. Furthermore, complex implementation, high computation time, and more memory requirements are key problems of transform domain methods. With the purpose of solving these problems, this research suggests a fusion framework for multimodal medical images that makes use of a multi-scale edge-preserving filter and visual saliency detection. The source images are decomposed using a two-scale edge-preserving filter into base and detail layers. Base layers are combined using the addition fusion rule, while detail layers are fused using weight maps constructed using the maximum symmetric surround saliency detection algorithm. The resultant image constructed by the presumed method has improved objective evaluation metrics than other classical methods, as well as unhindered edge contour, more global contrast, and no ringing effect or artifacts. The methodology offers a dominant and symbiotic arsenal of clinical symptomatic, therapeutic, and biomedical research competencies that have the prospective to considerably strengthen medical practice and biological understanding.

Impact of ISP Tuning on Object Detection.

In advanced driver assistance systems (ADAS) or autonomous vehicle research, acquiring semantic information about the surrounding environment generally relies heavily on camera-based object detection. Image signal processors (ISPs) in cameras are generally tuned for human perception. In most cases, ISP parameters are selected subjectively and the resulting image differs depending on the individual who tuned it. While the installation of cameras on cars started as a means of providing a view of the vehicle's environment to the driver, cameras are increasingly becoming part of safety-critical object detection systems for ADAS. Deep learning-based object detection has become prominent, but the effect of varying the ISP parameters has an unknown performance impact. In this study, we analyze the performance of 14 popular object detection models in the context of changes in the ISP parameters. We consider eight ISP blocks: demosaicing, gamma, denoising, edge enhancement, local tone mapping, saturation, contrast, and hue angle. We investigate two raw datasets, PASCALRAW and a custom raw dataset collected from an advanced driver assistance system (ADAS) perspective. We found that varying from a default ISP degrades the object detection performance and that the models differ in sensitivity to varying ISP parameters. Finally, we propose a novel methodology that increases object detection model robustness via ISP variation data augmentation.

Radiomics Texture Analysis of Bone Marrow Alterations in MRI Knee Examinations.

Accurate diagnosis and timely intervention are key to addressing common knee conditions effectively. In this work, we aim to identify textural changes in knee lesions based on bone marrow edema (BME), injury (INJ), and osteoarthritis (OST). One hundred and twenty-one MRI knee examinations were selected. Cases were divided into three groups based on radiological findings: forty-one in the BME, thirty-seven in the INJ, and forty-three in the OST groups. From each ROI, eighty-one radiomic descriptors were calculated, encoding texture information. The results suggested differences in the texture characteristics of regions of interest (ROIs) extracted from PD-FSE and STIR sequences. We observed that the ROIs associated with BME exhibited greater local contrast and a wider range of structural diversity compared to the ROIs corresponding to OST. When it comes to STIR sequences, the ROIs related to BME showed higher uniformity in terms of both signal intensity and the variability of local structures compared to the INJ ROIs. A combined radiomic descriptor managed to achieve a high separation ability, with AUC of 0.93 ± 0.02 in the test set. Radiomics analysis may provide a non-invasive and quantitative means to assess the spatial distribution and heterogeneity of bone marrow edema, aiding in its early detection and characterization.

The Case of Literary Journalism: Rethinking Fictionality, Narrativity, and Imagination

ABSTRACT This article discusses the genre of literary journalism/reportage against a background of earlier assumptions on fictionality. At a local level in nonfiction, fictionality can be expressed through invented stories and scenarios that create a contrast to the global, nonfictive context. However, fictionality can also be expressed through stylistic devices that traditionally have been associated with narrative fiction. A local contrast may appear, but only if the genre in itself is not narrative. If the focus is on the nonfictional and narrative genre of literary journalism/reportage, there will be no contrast. Here, the rhetoric will work just like in narrative fiction and should be considered to be part of the features of narrativity. Furthermore, the concept imagination should be perceived in close relation to Monika Fludernik’s understanding of narrative as experience. The conclusion is a call to partly rethink existing connections between fictionality, narrativity, and imagination in order to better understand the narrative nature of reportage.

Robust infrared small target detection with multi-feature fusion

The performance of the infrared search and track system depends on the robust of the infrared small target detection. In this paper, a novel small target detection with multi-feature fusion is proposed. At first, the salient proposal region is obtained with phase spectrum Fourier transform (PFT). Then, geometric moment is utilized to design an adaptive steerable edges awareness factor, which effectively eliminates the strong edge clutters. Afterwards, we adopt a three-layers local contrast structure and the difference-ration form is introduced to enhance the target and suppress the complex backgrounds. Finally, the adaptive threshold is applied to extract the target in the saliency map. Experiments are conducted on four real IR image sequences, which demonstrate that the proposed method outperforms the comparing methods with various complex backgrounds.

Detail-Preserving Multi-exposure Fusion for DR Images of Turbine Blades with Local Contrast Analysis and Exposure Intensity

Detail-Preserving Multi-exposure Fusion for DR Images of Turbine Blades with Local Contrast Analysis and Exposure Intensity

Nanoscale chemical characterization of secondary protein structure of F-Actin using mid-infrared photoinduced force microscopy (PiF-IR)

The recently developed photoinduced force microscopy for mid-infrared (PiF-IR) offers high spectral resolution in combination with surface sensitivity and a spatial resolution in the range of a few nanometers. Although PiF-IR has primarily been applied to polymer materials, this technology presents significant potential for the chemical characterization of cellular structures approaching single-molecule sensitivity. We applied PiF-IR to differently polymerized F-Actin samples finding general agreement with FTIR spectra from the same samples. Single PiF-IR spectra of F-Actin show variations in the amide I band spectral region, which is related to secondary protein structure. Local variations are also seen in PiF-IR hyperspectra in this region. Such high sensitivity is a necessary requirement for discriminating Actin organization into bundles and other networks in cells and tissue. We applied PiF-IR to mouse liver tissue ex vivo. Single-frequency PiF-IR scans at three different IR frequencies show significant variations in local contrast. However, the presence of other proteins and the unique spatial resolution of PiF-IR pose a challenge to interpreting and validating such data. Careful design of model systems and further theoretical understanding of PiF-IR data far from bulk averages are needed to fully unfold the potential of PiF-IR for high-resolution chemical investigation in the Life Sciences.

Hyperspectral anomaly detection based on local contrast estimation and sub-block background estimation

Traditional hyperspectral anomaly detectors used spatial and spectral information to extract information about the anomalies. However, most methods ignored spatial structural information of the hyperspectral image (HSI), resulting in poor background suppression and a high risk of false alarms. To solve this problem, we proposed a novel hyperspectral anomaly detection method based on local contrast estimation and subblock background estimation. First, a novel multiscale grid block window module (MSGBW) was proposed to utilize the spatial structural information of the HSI. Second, to improve the difference between anomalies and background, the local contrast was defined based on the MSGBW to obtain the contrast score. Third, to reduce the contamination of background features by anomalies, a subblock-based background estimation method was proposed to obtain clean background data. The clean background was used to obtain the max-eigen score by the Mahalanobis distance and eigen decomposition. Finally, the anomaly detection result was obtained by fusing the contrast score and the max-eigen score. In our experiments, four real-world datasets were adopted to evaluate the performance of the proposed method. The experimental results demonstrated that the proposed method obtained excellent performance in both qualitative and quantitative evaluations compared with the other seven state-of-the-art methods.

A Low-Delay Dynamic Range Compression and Contrast Enhancement Algorithm Based on an Uncooled Infrared Sensor with Local Optimal Contrast

Real-time compression of images with a high dynamic range into those with a low dynamic range while preserving the maximum amount of detail is still a critical technology in infrared image processing. We propose a dynamic range compression and enhancement algorithm for infrared images with local optimal contrast (DRCE-LOC). The algorithm has four steps. The first involves blocking the original image to determine the optimal stretching coefficient by using the information of the local block. In the second, the algorithm combines the original image with a low-pass filter to create the background and detailed layers, compressing the background layer with a dynamic range of adaptive gain, and enhancing the detailed layer for the visual characteristics of the human eye. Third, the original image was used as input, the compressed background layer was used as a brightness-guided image, and the local optimal stretching coefficient was used for dynamic range compression. Fourth, an 8-bit image was created (from typical 14-bit input) by merging the enhanced details and the compressed background. Implemented on FPGA, it used 2.2554 Mb of Block RAM, five dividers, and a root calculator with a total image delay of 0.018 s. The study analyzed mainstream algorithms in various scenarios (rich scenes, small targets, and indoor scenes), confirming the proposed algorithm’s superiority in real-time processing, resource utilization, preservation of the image’s details, and visual effects.