Binary Thresholding Research Articles

Full-field deformation measurement and cracks detection in speckle scene using the deep learning-aided digital image correlation method

An enhanced full-field deformation and crack measurement method for oblique optical-axis conditions was proposed. First, the thickness of the calibration plate causes an error in the homography conversion (H matrix) and the final displacement results. To solve this problem, a Perspective-n-Point (PnP) based thickness compensation method for H calibration was proposed. Second, for better measurement of the full-field deformation, the measured surface is mapped with random speckles and markers, and the complex background and covering problems create obstacles for crack skeleton detection. To overcome these interferences, a strain-guided two-phase detection method based on the deep learning model was innovatively designed. Third, the traditional crack width calculation method is susceptible to the binarization threshold and the connection to the speckles, and a crack width calculation method using the displacement of control points was suggested. During a bending test, the above methods were applied to a basalt fiber-reinforced polymer (BFRP) beam. The enhancement in the displacement measurement accuracy and the robustness of crack detection, particularly the potential of detecting hidden cracks, were verified by comparing the performance of the proposed method to that of traditional contact sensors.

Image Recognition and Parameter Analysis of Concrete Vibration State Based on Support Vector Machine.

In this paper, the directional gradient histogram technology is employed to extract the features of concrete image samples captured under different vibration states. The support vector machine (SVM) is utilized to learn the relationship between image features and vibration state. The machine learning results are subsequently used to assess the feasibility of the concrete vibration state. Simultaneously, the influence mechanism of the calculation parameters of the directional gradient histogram on the recognition accuracy is analyzed. The results demonstrate the feasibility of using the directional gradient histogram-SVM technology to identify the vibration state of concrete. The recognition accuracy initially increases and then decreases as the block size of the directional gradient, or the number of statistical intervals increases. The recognition accuracy also decreases linearly with the increase of the binarization threshold. By using sample images with a resolution of 1024 pixels x 1024 pixels and optimizing the feature extraction parameters, a recognition accuracy of 100% can be attained.

CHARACTERIZATION OF CLINICAL DATA FOR IN MODERATE OSTEOARTHRITIS WITH SUPPORT VECTOR MACHINES AND REGULATORY NETWORK MODELS

Knee osteoarthritis (KOA) diagnosis is based on symptoms, assessed through questionnaires such as the WOMAC. However, the inconsistency of pain recording and the discrepancy between joint phenotype and symptoms highlight the need for objective biomarkers in KOA diagnosis. To this end, we study relationships among clinical and molecular data in a cohort of women (n=51) with Kellgren-Lawrence grade 2–3 KOA through Support Vector Machine (SVM) and a regulation network model (RNM). Clinical descriptors (i.e., pain catastrophism (CA); depression (DE); functionality (FU); joint pain (JP); rigidity (RI); sensitization (SE); synovitis (SY)) are used to classify patients. A Youden's test is performed for each classifier to determine optimal binarization thresholds for the descriptors. Thresholds are tested against patient stratification according to baseline WOMAC data from the Osteoarthritis Initiative, and the mean accuracy is 0.97. For our cohort, the data used as SVM inputs are KOA descriptors, synovial fluid (SL) proteomic measurements (n=25), and transcription factors (TF) activation obtained from RNM [2] stimulated with the SL measurements. The relative weights after classification reflect input importance. The performance of each classifier is evaluated through AUC-ROC analysis. The best classifier with clinical data is CA (AUC = 0.9), highly influenced by FU and SE, suggesting that kinesophobia is involved in pain perception. With SL input, leptin strongly influences every classifier, suggesting the importance of low-grade inflammation. When TF are used, the mean AUC is limited to 0.608, which can be related to the pleomorphic behaviour of osteoarthritic chondrocytes. Nevertheless, FU has an AUC of 0.7 with strong importance of FOXO downregulation. Though larger and longitudinal cohorts are needed, this unique combination of SVM and RNM shall help to map objectively KOA descriptors.Acknowledgements: Catalan & Spanish governments 2020FI_b00680; STRATO-PID2021126469ob-C21-2, European Commission (MSCA-TN-ETN-2020-Disc4All-955735, ERC-2021-CoG-O-Health-101044828). ICREA Academia.

Influence of the number of attribute binarization thresholds on the quality of the generated set of rules for the method of logical analysis of data

For efficient operation of machine learning methods, it is necessary to set their parameters up properly. Both the computational complexity and the accuracy of the method for the problem being solved can depend on the selected parameter values. The paper discusses the method of logical analysis of data, which is used to solve classification problems and consists of a number of steps. At each step, it is necessary to set the method parameters up to suit the problem being solved. When setting parameters, one should be guided by a compromise between the accuracy of the method and the computational complexity. With comparable values for classification accuracy of several variants of the method implementation, preference will always be given to the simplest of them in terms of computational complexity. Since the method works only with binary characteristics, at the first stage it is necessary to binarize quantitative characteristics. The binarization procedure is associated with the choice of the “number of binarization thresholds” parameter. This paper proposes an experimental approach to determine the best value of the specified method parameter for the problem being solved.

Methods for detecting and counting nodes in images of crack networks

The article discusses a technique for segmenting a network of cracks in micrographs and identifying the main elements such as a node, the junction of several cracks, and an edge, the body of the crack itself, to build a model of the network as an undirected graph. Crack segmentation was carried out using two methods: using threshold binarization and applying masks that separate nodes from edges based on morphological characteristics, and a combined method using a convolutional neural network to detect nodes. Such methods make it possible to detect nodes and edges automatically, facilitating the construction of a model and opening up new possibilities in theoretical calculations of the resistance of a network of conductors in transparent conductive coatings.

Determining the Presence and Size of Shoulder Lesions in Sows Using Computer Vision.

Shoulder sores predominantly arise in breeding sows and often result in untimely culling. Reported prevalence rates vary significantly, spanning between 5% and 50% depending upon the type of crate flooring inside a farm, the animal's body condition, or an existing injury that causes lameness. These lesions represent not only a welfare concern but also have an economic impact due to the labor needed for treatment and medication. The objective of this study was to evaluate the use of computer vision techniques in detecting and determining the size of shoulder lesions. A Microsoft Kinect V2 camera captured the top-down depth and RGB images of sows in farrowing crates. The RGB images were collected at a resolution of 1920 × 1080. To ensure the best view of the lesions, images were selected with sows lying on their right and left sides with all legs extended. A total of 824 RGB images from 70 sows with lesions at various stages of development were identified and annotated. Three deep learning-based object detection models, YOLOv5, YOLOv8, and Faster-RCNN, pre-trained with the COCO and ImageNet datasets, were implemented to localize the lesion area. YOLOv5 was the best predictor as it was able to detect lesions with an mAP@0.5 of 0.92. To estimate the lesion area, lesion pixel segmentation was carried out on the localized region using traditional image processing techniques like Otsu's binarization and adaptive thresholding alongside DL-based segmentation models based on U-Net architecture. In conclusion, this study demonstrates the potential of computer vision techniques in effectively detecting and assessing the size of shoulder lesions in breeding sows, providing a promising avenue for improving sow welfare and reducing economic losses.

Analysis and improvement of methods and means for eliminating distortions in image and video signals

The main purpose of this article was to study existing image processing methods, binarization and noise reduction, and to develop an improved adaptive thresholding algorithm. The analysis of methods and tools for eliminating distortions in image and video signals is an urgent task in numerous fields, including medicine, computer vision, and document science. The article discusses in detail the existing binarization and noise reduction methods, highlighting their advantages and limitations. However, the main achievement is the development and implementation of an improved adaptive thresholding algorithm. This algorithm considers the specific features of the image and automatically adapts the binarization threshold for better processing quality. It is a significant contribution to the field of image processing and can be used in various fields, including medical diagnostics and visual object detection in images.

Image‐based crop row detection utilizing the Hough transform and DBSCAN clustering analysis

AbstractMore accurate methods for crop row detection benefit intelligent operation of agricultural machinery, especially avoiding mishandling or crushing crops. For achieving such a target, a traditional method combining the ExGR exponents, Otsu algorithm, Canny method, Hough transform and DBSCAN clustering analysis is proposed so that centerlines of crop rows can be detected effectively without manual intervention. Specifically, ExGR exponents are first adopted to gray green plants. The threshold of binarization will be further obtained by the Otsu algorithm. Further adopting the edge detection algorithm (Canny), edges of crops can be determined. Finally, combining the Hough transform and DBSCAN clustering analysis, the crop row detection is effectively available. Utilizing these methods, numerical simulation and their comparisons with existing methods are also achieved. For example, the Canny algorithm is relatively accurate than the Suzuki algorithm as well as their combinations with a geometric center extraction method if the density of weed is high. Compared with the K‐means clustering method, the DBSCAN algorithm is more suitable to characterize crop rows optimally in more complex conditions. It is validated from experiments that the combination of Canny algorithm, Hough transform and DBSCAN clustering is better than other mentioned traditional methods.

Expansion microscopy for visualization of protein clusters in cultured cells and brain tissues

Expansion microscopy for visualization of protein clusters in cultured cells and brain tissues

Robust multifunctional watermarking system based zero-watermarking and polar harmonic Fourier moments

The pseudo-randomness exhibited by chaotic systems has proven valuable not only in plaintext information encryption of images with varying dimensions but also as a key signal for protecting image copyright information. Image copyright protection is a crucial part of image information protection. Currently, digital watermarking technology is predominantly employed for copyright protection of digital images. However, most of the existing watermarking technologies primarily focus on the improvement of watermarking performance, neglecting the security of the watermarking system itself. Drawing inspiration from the Latin square rule, this paper proposes a novel hyperchaotic mapping modulation framework, which is employed to construct a new 3D hyperchaotic mapping and enhance the security of the fragile watermark and the zero-watermark generation process. Additionally, it is integrated into the classical image encryption framework of scrambling-diffusion to safeguard meaningful copyright information. Moreover, a functional watermarking system is designed based on the RSA asymmetric encryption algorithm, least significant bit (LSB), and polar harmonic Fourier moments (PHFMs). The proposed system enables identity verification, tampering location, and copyright authentication simultaneously. During the generation of zero watermarks, the binarization threshold of the PHFMs information matrix is determined by the method of maximum variance. The experimental results demonstrate that this scheme achieves a high level of security, better robustness to both conventional and geometric attacks, and precise localization for different types of tampering attacks.

Analyzing Pore Evolution Characteristics in Cementitious Materials Using a Plane Distribution Model

This research aims to analyze the distribution and evolution of pores within the planar structure of cement-based materials. Utilizing digital imaging methods, a model for pore plane distribution was established, and the evolutionary patterns of both total pore numbers and varying pore sizes in cement-based materials were investigated. The research introduced an innovative experimental method for analyzing pore distribution within cement-based planar structures. Additionally, a hybrid method was proposed, combining automated image binarization thresholding with manual comparative analysis, thereby enhancing the feasibility of comparative research. Pores were categorized into four distinct sizes: tiny pores (5–200 μm), small pores (200–500 μm), medium pores (500–1000 μm), and large pores (>1000 μm). Areas with apertures <5 μm were classified as dense areas. The findings indicated that the overall number of pores in cement-based materials increased due to the influence of styrene butadiene latex additives. However, at a 15% dosage, the rate of pore formation reached an inflection point, confirming that various factors, such as styrene butadiene latex, air bubbles, and the cement-based material itself, collectively influenced pore formation. The research also demonstrated that styrene butadiene latex affected the four categorized pore sizes differently. Importantly, a higher latex dosage did not necessarily lead to a proportional increase in pore content. Pore content was influenced by multiple factors and exhibited different distribution patterns. The number of micropores, although relatively small, gradually increased with higher latex dosages, while small and medium pores generally showed an upward trend. At a 10% latex dosage, both small and medium pores reached a turning point in their rate of increase. Large pores also exhibited a general increase, peaking at a latex dosage of 10%. It was confirmed that both the total pore volume and the content of micropores were critical factors in determining the mechanical properties of cementitious materials. Higher porosity and micropore content generally weakened mechanical performance. However, at a small latex dosage, there was an improvement in flexural strength. When the latex dosage reached 15%, the total pore and micropore content declined, resulting in a balanced increase in flexural strength and a mitigated decline in compressive strength. This study offers valuable insights into the evolution of total pore volume and the content of pores of various sizes, providing a theoretical basis for the meticulous selection of additive types and dosages from a microscopic perspective.

Crosswalk Detection from Satellite Imagery for Pedestrian Network Completion

Pedestrian network creation and completion for various tasks in transportation planning engineering generally requires laborious manual labeling of sidewalks and crosswalks. In this study, we employed the deep learning-based object detection model YOLO v5 for detecting crosswalks using Google Maps satellite imagery data of four North American cities. We also tested for two additional scenarios involving two image-processing techniques to exploit the special properties of crosswalks. Observations showed that the original images performed better than with binary thresholding and nearly the same even when noisy regions outside the right-of-way were removed. We proposed an algorithm to assign the detected crosswalks to intersection legs. We demonstrated the effectiveness of this technique for Washington, D.C. and Los Angeles, CA, showing classification accuracy rates of 71% and 89%, respectively. We also showed the influence of increasing distance threshold, a tolerance radius of prediction accuracy, in degrading the classification performance. This method, when combined with existing methods for sidewalk detection from aerial and street view images, could be reasonably used to help complete urban pedestrian networks in cities where high-resolution satellite images are available.

Classifying dynamic motor imagery with the locals-balanced extreme learning machine

The dynamic motor imagery (dMI) provides additional benefits as compared to the traditional motor imagery (MI) in training and neurorehabilitation field. The objective of this work is to develop a brain-computer interface (BCI) for dMI electroencephalograph (EEG). We propose a novel method by combining synchronization likelihood (SL) based functional brain network (FBN) and modified extreme learning machine (ELM) to interpret EEG. The proposed method 1) uses L1-norm and adaptive threshold instead of L2-norm and empirical threshold in SL method; 2) improves the procedure of FBN calculations (namely, SL matrix binarization) by combining threshold and MST methods; 3) detects two defects in standard ELM in fusion, and proposes the locals-balanced ELM (LBELM); 4) employs the special version of leave-one-out cross validation (LOO) approach improved the computational complexity of optimal threshold for binarization, the most effective fusion of two complementary features, and the optimal regularization parameters for regularized LBELM. As compared with the empirical threshold, the recognition rate of the optimal threshold increases by 1.39–11.11 %. Similarly, as compared with the regularized ELM, the recognition rate of LBELM increases by 2.78–9.73 %.

Low consumption automatic discovery protocol for DDS-based large-scale distributed parallel computing

DDS (Data Distribution Service) is an efficient communication specification for distributed parallel computing. However, as the scale of computation expands, high network load and memory consumption consistently limit its performance. This paper proposes a low consumption automatic discovery protocol to improve DDS in large-scale distributed parallel computing. Firstly, an improved Bloom Filter called TBF (Threshold Bloom Filter) is presented to compress the data topic. Then it is combined with the SDP(Simple Discovery Protocol) to reduce the consumption of the automatic discovery process in DDS. On this basis, data publication and subscription between the distributed computing nodes are matched using binarization threshold θ and decision threshold T , which can be obtained through iterative optimization algorithms. Experiment results show that the SDPTBF can guarantee higher transmission accuracy while reducing network load and memory consumption, and therefore improve the performance of DDS-based large-scale distributed parallel computing.

A Meteor Detection Algorithm for GWAC System

Compared with the international meteor surveillance systems, the ground wide angle camera (GWAC) system exhibits characteristics such as images with the resolution of 4K × 4K and single-site observation. These characteristics present challenges for meteor detection in the GWAC system. Consequently, this paper proposes a new meteor detection algorithm for the GWAC system on the base of the solely mini-GWAC system data algorithm. The new algorithm consists of the following key steps: (1) to compare differences between adjacent frames, applying block-based image binarization thresholds, and incorporating median filtering to reduce noise; (2) to adopt the probabilistic Hough transform (PHT) to identify moving objects and cluster them based on the origin moment of the line segments, while assessing the credibility of clustering; (3) to introduce the so-called maximum disappearance frame for moving objects in the tracking algorithm, enhancing the ability to track multi-frame moving objects. The utilization of the line segment inclination angle of the moving object as the direction of movement facilitates the tracking of multiple moving objects, thereby reducing the probability of mistakenly selecting single-frame moving objects; (4) to leverage the light curves of single-frame moving objects to select meteors to enhance the accuracy of meteor detection. Comparative experiments demonstrate that our proposed algorithm processes each frame image in just 0.39 s, achieving an accuracy of 89.8% in the dataset of 5856 adjacent frames. The experimental results indicate that the algorithm achieved an accuracy of 90.27% when applied in the meteor detection of the image data captured by the GWAC system from Dec. 10th to 19th in 2019 and 2021, obtaining excellent detection results.

Dried shiitake mushroom grade recognition using D-VGG network and machine vision.

Grading dried shiitake mushrooms is an indispensable production step, as there are large quality differences between different grades, which affect the product's price and marketability. Dried shiitake mushroom samples have irregular shapes, small morphological differences between different grades of the same species, and they may occur in mixed grades, which causes challenges to the automatic grade recognition using machine vision. In this study, a comprehensive method to solve this problem is provided, including image acquisition, preprocessing, dataset creation, and grade recognition. The osprey optimization algorithm (OOA) is used to improve the computational efficiency of Otsu's threshold binarization and obtain complete mushroom contours samples efficiently. Then, a method for dried shiitake mushroom grade recognition based on the improved VGG network (D-VGG) is proposed. The method uses the VGG16 network as the base framework, optimizes the convolutional layer of the network, and uses a global average pooling layer instead of a fully connected layer to reduce the risk of model overfitting. In addition, a residual module and batch normalization are introduced to enhance the learning effect of texture details, accelerate the convergence of the model, and improve the stability of the training process. An improved channel attention network is proposed to enhance the feature weights of different channels and improve the grading performance of the model. The experimental results show that the improved network model (D-VGG) can recognize different dried shiitake mushroom grades with high accuracy and recognition efficiency, achieving a final grading accuracy of 96.21%, with only 46.77 ms required to process a single image. The dried shiitake mushroom grade recognition method proposed in this study provides a new implementation approach for the dried shiitake mushroom quality grading process, as well as a reference for real-time grade recognition of other agricultural products.

Alteration of perivascular reflectivity on optical coherence tomography of branched retinal vein obstruction

This study aimed to evaluate perivascular reflectivity in patients with branched retinal vascular obstruction (BRVO) using en-face optical coherence tomography (OCT). The study retrospectively analyzed 45 patients with recurrent BRVO, 30 with indolent BRVO, and 45 age- and sex-matched controls. Using a 3.0 × 3.0-mm deep capillary plexus slab on macular scans, OCT angiography (OCTA) and structural en-face OCT scans were divided into four quadrants. Obstructive quadrants of OCTA scans were binarized using a threshold value of mean + 2 standard deviation. The selected area of high signal strength (HSS) was applied to the structural en-face OCT scans, and the corrected mean perivascular reflectivity was calculated as the mean reflectivity on the HSS area/overall en-face OCT mean reflectivity. The same procedure was performed in the quadrants of the matched controls. Regression analysis was conducted on several factors possibly associated with corrected perivascular reflectivity. The perivascular reflectivity in the obstructive BRVO quadrant was significantly higher than in the indolent BRVO and control quadrants (P = 0.009, P = 0.003). Both univariate and multivariate regression analyses showed a significant correlation between the average number of intravitreal injections (anti-vascular endothelial growth factor or dexamethasone implant) per year and refractive errors and image binarization threshold and perivascular reflectivity (P = 0.011, 0.013, < 0.001/univariate; 0.007, 0.041, 0.005/multivariate, respectively). En-face OCT scans of the deep capillary plexus slab revealed higher perivascular reflectivity in recurrent BRVO eyes than in indolent BRVO and control eyes. The results also indicate a remarkable correlation between perivascular reflectivity and the average number of intravitreal injections, suggesting a link to recurrence rates.

Development of Trio Optimal Feature Extraction Model for Attention-Based Adaptive Weighted RNN-Based Lung and Colon Cancer Detection Framework Using Histopathological Images

Due to the combination of genetic diseases as well as a variety of biomedical abnormalities, the fatal disease named cancer is caused. Colon and lung cancer are regarded as the two leading diseases for disability and death. The most significant component for demonstrating the best course of action is the histopathological identification of such malignancies. So, in order to minimize the mortality rate caused by cancer, there is a need for early detection of the aliment on both fronts accordingly. In this case, both the deep and machine learning techniques have been utilized to speed up the detection process of cancer which may also help the researchers to study a huge amount of patients over a short period and less loss. Hence, it is highly essential to design a new lung and colon detection model based on deep learning approaches. Initially, a different set of histopathological images is collected from benchmark resources to perform effective analysis. Then, to attain the first set of features, the collected image is offered to the dilated net for attaining deep image features with the help of the Visual Geometry Group (VGG16) and Residual Neural Network (ResNet). Further, the second set of features is attained by the below process. Here, the collected image is given to pre-processing phase and the image is pre–pre-processed with the help of Contrast-limited Adaptive Histogram Equalization (CLAHE) and filter technique. Then, the pre-processed image is offered to the segmentation phase with the help of adaptive binary thresholding and offered to a dilated network that holds VGG16 and ResNet and attained the second set of features. The parameters of adaptive binary thresholding are tuned with the help of a developed hybrid approach called Sand Cat swarm JAya Optimization (SC-JAO) via Sand Cat swarm Optimization (SCO) and JAYA (SC-JAO). Finally, the third set of features is attained by offering the image to pre-processing phase. Then, the pre-processed image is offered to the segmentation phase and the image is a segmented phase and features are tuned by developed SC-JAO. Further, the segmented features are offered to attain the textural features like Gray-Level Co-Occurrence Matrix (GLCM) and Local Weber Pattern (LWP) and attained the third set of features. Then, the attained three different sets of features are given to the optimal weighted feature phase, where the parameters are optimized by the SC-JAO algorithm and then given to the disease prediction phase. Here, disease prediction is made with the help of Attention-based Adaptive Weighted Recurrent Neural Networks (AAW-RNN), and their parameters are tuned by developed SC-JAO. Thus, the developed model achieved an effective lung and colon detection rate over conventional approaches over multiple experimental analyses.

Spatially explicit accuracy assessment of deep learning-based, fine-resolution built-up land data in the United States.

Geospatial datasets derived from remote sensing data by means of machine learning methods are often based on probabilistic outputs of abstract nature, which are difficult to translate into interpretable measures. For example, the Global Human Settlement Layer GHS-BUILT-S2 product reports the probability of the presence of built-up areas in 2018 in a global 10 m × 10 m grid. However, practitioners typically require interpretable measures such as binary surfaces indicating the presence or absence of built-up areas or estimates of sub-pixel built-up surface fractions. Herein, we assess the relationship between the built-up probability in GHS-BUILT-S2 and reference built-up surface fractions derived from a highly reliable reference database for several regions in the United States. Furthermore, we identify a binarization threshold using an agreement maximization method that creates binary built-up land data from these built-up probabilities. These binary surfaces are input to a spatially explicit, scale-sensitive accuracy assessment which includes the use of a novel, visual-analytical tool which we call focal precision-recall signature plots. Our analysis reveals that a threshold of 0.5 applied to GHS-BUILT-S2 maximizes the agreement with binarized built-up land data derived from the reference built-up area fraction. We find high levels of accuracy (i.e., county-level F-1 scores of almost 0.8 on average) in the derived built-up areas, and consistently high accuracy along the rural-urban gradient in our study area. These results reveal considerable accuracy improvements in human settlement models based on Sentinel-2 data and deep learning, as compared to earlier, Landsat-based versions of the Global Human Settlement Layer.

Fourier Single-Pixel Imaging Based on Online Modulation Pattern Binarization

Down-sampling Fourier single-pixel imaging is typically achieved by truncating the Fourier spectrum, where exclusively the low-frequency Fourier coefficients are extracted while discarding the high-frequency components. However, the truncation of the Fourier spectrum can lead to an undesired ringing effect in the reconstructed result. Moreover, the original Fourier single-pixel imaging necessitated grayscale Fourier basis patterns for illumination. This requirement limits imaging speed because digital micromirror devices (DMDs) generate grayscale patterns at a lower refresh rate. In order to solve the above problem, a fast and high-quality Fourier single-pixel imaging reconstruction method is proposed in the paper. In the method, the threshold binarization of the Fourier base pattern is performed online to improve the DMD refresh rate, and the reconstruction quality of Fourier single-pixel imaging at a low-sampling rate is improved by generating an adversarial network. This method enables fast reconstruction of target images with higher quality despite low-sampling rates. Compared with conventional Fourier single-pixel imaging, numerical simulation and experimentation demonstrate the effectiveness of the proposed method. Notably, this method is particularly significant for fast Fourier single-pixel imaging applications.