Gaussian Mixture Model Algorithm Research Articles

Improved automated parallel implementation of GMM background subtraction on a multicore digital signal processor

<p><span>Scene segmentation is an essential step in a wide range of video processing applications, for instance, object recognition and tracking. The Gaussian mixture model (GMM) for background subtraction (BS) has gained widespread usage in scene segmentation, despite its known computational intensity. To tackle this challenge, we propose a practical solution to accelerate processing through a parallel implementation on an embedded multicore platform. In this paper, we present an improved automated parallel implementation of the GMM algorithm using the Orphan directive provided by open multiprocessing (OpenMP). Experimental assessments conducted on the eight cores of the C6678 digital signal processor (DSP) demonstrate significant advancements in parallel efficiency, particularly when handling high-resolution frames, including high-definition (HD) and full-HD resolutions. The achieved parallel efficiency surpasses the results obtained with classical OpenMP scheduling modes, encompassing dynamic, static, and guided approaches. Specifically, the parallel efficiency reaches approximately 82% for full-HD resolution frames and, 99.3% for low-resolution frames, respectively.</span></p>

Development of a method for predicting hazardous ship trajectories under uncertainty of navigator actions

The object of the research is the automation processes in maritime navigation to ensure the safety of ship movement by predicting their trajectories in complex aquatic areas, such as narrow passages, straits, and ports. The research applied six key stages to create a comprehensive method for clustering and predicting ship trajectories based on ECDIS data. In the first stage, ship movement trajectories were constructed according to risk categories, using the LCSS and DTW algorithms to compare planned and actual trajectories. This allowed for the accurate identification of course deviations and the determination of potentially dangerous sections of the trajectory. The second stage implemented clustering using the DBSCAN and GMM algorithms. DBSCAN was used to identify the density of points in space, and GMM provided modeling of cluster probabilities, allowing for better risk zone determination. The third stage applied the Douglas-Peucker compression algorithm to reduce the number of points in the trajectories, which preserved key characteristics and optimized data processing. In the fourth stage, ship movement stability was assessed using the Fourier transform, which allowed the detection of high-frequency oscillations that may indicate movement instability caused by changes in course or speed. The fifth stage included fuzzy clustering of trajectories using the Gaussian Mixture Model (GMM), which allowed modeling the probabilities of dangerous trajectories, considering the uncertainty of navigational parameters. At the final stage, a multilayer neural network (MLP) was used to predict future points of ship trajectories. The model accurately predicted the ship's coordinates, enabling timely trajectory adjustments. Experimental results showed that the developed method increased the accuracy of ship trajectory prediction to 72–81 % and also significantly reduced the final error, ensuring effective risk management during complex navigation.

Multi-color space features and data regression neural network-based method for molten salt temperature detection in rare earth electrolytic processes

To address the online monitoring of molten salt temperature in the rare earth electrolysis process, this paper proposes a rare earth molten salt temperature detection model (MSTDM) based on multi-color space features and data regression neural networks. The model initially preprocesses the collected molten salt temperature images using techniques such as the Gaussian Mixture Model (GMM) and median filtering, followed by color space transformation. Subsequently, 19 hue features used for temperature quantification are extracted, including the channel responses of RGB, HSV, Lab, and Gray color spaces. Finally, a novel data regression neural network (CNN-BiLSTM-Attention) is employed to learn the long-term dependencies between molten salt hue features and temperature, predicting future molten salt temperatures. Experimental results indicate that using the GMM algorithm to filter out flame interference, introducing novel temperature prediction features, and constructing the CNN-BiLSTM-Attention network all enhance the performance of MSTDM in molten salt temperature prediction tasks. The model achieves a mean absolute error of 3.16 °C in predicting molten salt temperatures at rare earth smelting sites. This research not only provides an effective solution for online monitoring of molten salt temperatures in the rare earth electrolysis process but also offers a reliable reference for temperature monitoring in other fields.

Comparison of different MRI-based unsupervised segmentation algorithms in predicting sarcopenia

PurposeTo compare the performance of MRI-based Gaussian mixture model (GMM), K-means clustering, and Otsu unsupervised algorithms in predicting sarcopenia and to develop a combined model by integrating clinical indicators. MethodsRetrospective analysis was conducted on clinical and lumbar MRI data from 118 patients diagnosed with sarcopenia and 222 patients without the sarcopenia. All patients were randomly divided into training and validation groups in a 7:3 ratio. Regions of interest (ROI), specifically the paravertebral muscles at the L3/4 intervertebral disc level, were delineated on axial T2-weighted images (T2WI). The Gaussian mixture model (GMM), K-means clustering, and Otsu’s thresholding algorithms were employed to automatically segment muscle and adipose tissues at both the cohort and case levels. Subsequently, the mean signal intensity, volumes, and percentages of these tissues were calculated and compared. Logistic regression analyses were conducted to construct models and identify independent predictors of sarcopenia. An combined model was developed by combining the optimal magnetic resonance imaging (MRI) model and clinical predictors. The performance of the constructed model was assessed using receiver operating characteristic (ROC) curve analysis. ResultsAge, BMI, and serum albumin were identified as independent clinical predictors of sarcopenia. The cohort-level GMM demonstrated the best predictive performance both in the training group (AUC=0.840) and validation group (AUC=0.800), while the predictive performance of the other models was lower than that of the clinical model both in the training and validation groups. After combining the cohort-level GMM with the independent clinical predictors, the AUC of the training and validation groups increased to 0.871 and 0.867, respectively. ConclusionThe cohort-level GMM shows potential in predicting sarcopenia, and the incorporation of independent clinical predictors further increased the performance.

Vision-based mixed color detection of plastic particles.

In the production process of high-end PP-R pipes, mixing different colored raw material particles can result in uneven color in the final product, affecting its appearance quality. in addition, color mixing can reduce the physical properties of the pipes, impacting their durability and safety. To address this issue, we propose a visual, non-destructive inspection solution based on image processing technology. The solution aims to enhance detection efficiency and accuracy by reducing background interference and enabling adaptive adjustments in various environments. Initially, the K-Means image segmentation algorithm is employed to eliminate complex background factors from the original image, significantly improving image segmentation accuracy. Subsequently, the Gaussian mixture model algorithm is utilized to automatically extract the color threshold of the foreground image after background removal, facilitating adaptive algorithm adjustments. Finally, the mean value algorithm is introduced to swiftly and accurately identify plastic particles of different colors using the automatically obtained color thresholds. Experimental results demonstrate that this method can quickly and accurately identify different color particles and effectively support the rejection of impurity particles. Through this approach, the algorithm achieves an average detection accuracy of 99.3%.

The value of multiparametric MRI-based habitat imaging for differentiating uterine sarcomas from atypical leiomyomas: a multicentre study.

To explore the feasibility of multiparametric MRI-based habitat imaging for distinguishing uterine sarcoma (US) from atypical leiomyoma (ALM). This retrospective study included the clinical and preoperative MRI data of 69 patients with US and 225 patients with ALM from three hospitals. At both the individual and cohort levels, the K-means and Gaussian mixture model (GMM) algorithms were utilized to perform habitat imaging on MR images, respectively. Specifically, T2-weighted images (T2WI) and contrast-enhanced T1-weighted images (CE-T1WI) were clustered to generate structural habitats, while apparent diffusion coefficient (ADC) maps and CE-T1WI were clustered to create functional habitats. Parameters of each habitat subregion were extracted to construct distinct habitat models. The integrated models were constructed by combining habitat and clinical independent predictors. Model performance was assessed using the area under the curve (AUC). Abnormal vaginal bleeding, lactate dehydrogenase (LDH), and white blood cell (WBC) counts can serve as clinical independent predictors of US. The GMM-based functional habitat model at the cohort level had the highest mean AUC (0.766) in both the training and validation cohorts, followed by the GMM-based structural habitat model at the cohort level (AUC = 0.760). Within the integrated models, the K-means functional habitat model based on the cohort level achieved the highest mean AUC (0.905) in both the training and validation cohorts. Habitat imaging based on multiparametric MRI has the potential to distinguish US from ALM. The combination of clinical independent predictors with the habitat models can effectively improve the performance.

Exploring Gamma-Ray Burst Diversity: Clustering Analysis of the Emission Characteristics of Fermi- and BATSE-detected Gamma-Ray Bursts

Gamma-ray bursts (GRBs), often attributed to massive star collapse or binary compact object mergers, exhibit diverse emission characteristics hinting at multiple GRB classes based on various factors like progenitors, radiation mechanisms, and central engines. This study employs unsupervised clustering using the nested Gaussian mixture model algorithm to analyze data from Fermi and BATSE, identifying four classes (A–D) based on duration, spectral peak, and spectral index of time-integrated spectra of GRBs. Class proportions are approximately 70%, 10%, 3%, and 17%, respectively, with A and B comprising mostly long GRBs, C mainly short GRBs, and D encompassing both types. The classes are further assessed based on spectral index α, indicating distinct radiation mechanisms: α > −0.67 for photospheric emission, α ≤ −1.5 for fast-cooling synchrotron, and −1.5 < α ≤ −0.67 for slow-cooling synchrotron. Classes B and C align with photospheric emission, while A and D predominantly exhibit synchrotron radiation. Short GRBs are predominantly photospheric emission, whereas long GRBs tend to favor synchrotron emission. Overall, 63% of the total bursts exhibit α profiles indicative of synchrotron emission, with the remaining 37% associated with photospheric emission. Considering the limited data of kilonova and supernova associated with GRBs, classes are examined for progenitor origins, suggesting a hybrid nature for A and D, and collapsar and merger origins for B and C, respectively. This clustering analysis results in four GRB classes, which, upon investigation, reveal the diverse and complex nature of GRBs in terms of their radiation, duration, and progenitor.

Underwater Image Segmentation with the GMM (Gaussian Mixture Model) Algorithm

The purpose of this study was to measure the performance of Gaussian Mixture Model (GMM) technique for underwater image segmentation of seagrass objects based on datasets from autonomous surface vehicles (ASV from the Faculty of Fisheries and Marine Sciences, Bogor Agricultural University. The dataset is 640 x 480 pixel image data to support image segmentation research. There are three categories of underwater imagery: (a) underwater imagery featuring seagrass and seawater backgrounds; (b) underwater imagery featuring seagrasses, clear fish, and seawater backgrounds; and (c) underwater imagery featuring seagrasses, faint fish, and seawater backgrounds. Based on the experimental results, seagrass objects in image type (a) have almost identical colors to each pixel in the underwater image, the GMM model was able to distinguish them from the background and seawater background. The GMM model can distinguish between the background and the seawater background in image type (b), but cannot eliminate fish objects in the image. The segmentation results in image type (c) are not perfect because the GMM model removes seagrass objects that have green pixel color.

Uncovering the progressive failure process of primary coal-rock mass specimens: Insights from energy evolution, acoustic emission crack patterns, and visual characterization

The failure and instability of deep coal-rock mass structures, which are influenced by mining disturbances, are major contributors to disasters like rock bursts. This study involved conducting indoor experiments using uniaxial loading test and acoustic emission (AE) experiments on indigenous primary coal-rock mass (PCRM). We utilize chaotic bifurcation based on strain energy to characterize the progressive damage stages of PCRM, demonstrating the chaotic characteristics of energy evolution during the gradual destruction process. The degree of damage resulting from PCRM failure was evaluated by jointly applying P-wave velocity tomography and AE event localization. Additionally, the classification of crack patterns in PCRM was conducted using the RA-AF correlation analysis method. The segmentation boundaries of crack classification patterns were then redefined using the Gaussian mixture model (GMM) algorithm. The research findings indicate that the chaotic bifurcation model of strain energy classifies stress into four stages: the stable region (Ⅰ), the metastable region (Ⅱ), the bifurcation region (Ⅲ), and the chaotic region (Ⅳ). The presence of a low-velocity zone in P-waves may indicate crack accumulation and damage within the system. Additionally, areas with abnormal P-wave velocities exhibit numerous AE events. The crack classification patterns of the four stages undergo an evolution from predominantly tensile cracks to a mixture of tensile and shear cracks, and the GMM algorithm successfully identifies the optimal separation path for crack classification boundaries. We propose an adaptive kernel density estimation (AKDE) algorithm to quantify the spatial distribution of AE events, thus offering a visual representation of the damage patterns at various stages.

SYSTEM DESIGN FOR CALCUALTING THE NUMBER AND DENSITY OF MOTORCYCLES IN PARKING AREA BASED ON BACKGROUND SUBTRACTION METHOD

There are still issues with a number of students parking their vehicles improperly in the parking spaces provided. This causes suboptimal parking capacity due to a lack of information about parking capacity. In recent years, at a certain library, vehicle detection has been implemented using a Gaussian mixture model algorithm using a Raspberry Pi. However, this library does not provide information about the density status of the parking area. Therefore, an information system was created to determine the level of density of the parking area based on the ratio of vehicles entering and exiting compared to the maximum capacity of the parking area. The system uses the Gaussian mixture model algorithm with the machine learning method of background subtraction MOG2, which can calculate the number of vehicles based on the difference between objects and the background of objects, using test data in the form of videos recorded using a camera positioned horizontally to the entrance and exit lanes of the parking area. This research resulted in an accuracy of 89.7% for Video1TA, precision of 93.2%, a crowded parking area density level, and a value of 129.12. Video2TA had a value of 101.08, precision of 100%, and accuracy of 90%, while Video3TA had an accuracy of 35%, precision of 56.7%, and a value of 49.48. The density levels of videos 2 and 3 are the same, indicating that the parking area is still empty. The test results show that the value can affect the system in detecting an object.

Improved GMMKNN Hybrid Recommendation Algorithm

This paper proposes a hybrid recommendation algorithm that combines the advantages of Gaussian Mixture Model (GMM) and K-Nearest Neighbors (KNN) algorithms. The algorithm first applies GMM to cluster the training data, grouping users with similar interests using clustering techniques. It then utilizes the KNN algorithm for prediction. During the KNN recommendation process for a target user, the algorithm searches for neighboring users with similar interests and features within the same cluster. This significantly reduces the search scope of the nearest neighbors, thanks to the assistance of the GMM algorithm. The algorithm then selects the K most similar neighbors from the target user’s cluster as the candidate pool for personalized recommendations. Additionally, the algorithm improves the weight calculation method by incorporating a Gaussian kernel function for weight estimation. Experimental results on the dataset demonstrate that the proposed improved algorithm effectively enhances the accuracy of prediction results.

Does the Too-Big-To-Fail status affect depositor’s discipline: a Gaussian mixture model algorithm approach

PurposeThis study empirically demonstrates a contradiction between pillar 3 of Basel norms III and the designation of Systemically Important Banks (SIBs), also known as Too Big to Fail (TBTF). The objective of this study is threefold, which has been approached in a phased manner. The first is to determine the systemic importance of the banks under study; second, to examine if market discipline exists at different levels of systemic importance of banks and lastly, to examine if the strength of market discipline varies at different levels of systemic importance.Design/methodology/approachThis study is based on all the public and private sector banks operating in the Indian banking sector. The Gaussian Mixture Model algorithm has been utilized to classify banks into distinct levels of systemic importance. Thereafter, market discipline has been observed by analyzing depositors' sentiments toward banks' risk (CAMEL indicators). The analysis has been performed by employing the system Generalized Method of Moments (GMM) to estimate models with different dependent variables.FindingsThe findings affirm the existence of market discipline across all levels of systemic importance. However, the strength of market discipline varies with the systemic importance of the banks, with weak market discipline being a negative externality of the SIBs designation.Originality/valueBy employing the Gaussian Mixture Model algorithm to develop a framework for categorizing banks on the basis of their systemic importance, this study is the first to go beyond the conventional method as outlined by the Reserve Bank of India (RBI).

UAV-Based Classification of Intercropped Forage Cactus: A Comparison of RGB and Multispectral Sample Spaces Using Machine Learning in an Irrigated Area

Precision agriculture requires accurate methods for classifying crops and soil cover in agricultural production areas. The study aims to evaluate three machine learning-based classifiers to identify intercropped forage cactus cultivation in irrigated areas using Unmanned Aerial Vehicles (UAV). It conducted a comparative analysis between multispectral and visible Red-Green-Blue (RGB) sampling, followed by the efficiency analysis of Gaussian Mixture Model (GMM), K-Nearest Neighbors (KNN), and Random Forest (RF) algorithms. The classification targets included exposed soil, mulching soil cover, developed and undeveloped forage cactus, moringa, and gliricidia in the Brazilian semiarid. The results indicated that the KNN and RF algorithms outperformed other methods, showing no significant differences according to the kappa index for both Multispectral and RGB sample spaces. In contrast, the GMM showed lower performance, with kappa index values of 0.82 and 0.78, compared to RF 0.86 and 0.82, and KNN 0.86 and 0.82. The KNN and RF algorithms performed well, with individual accuracy rates above 85% for both sample spaces. Overall, the KNN algorithm demonstrated superiority for the RGB sample space, whereas the RF algorithm excelled for the multispectral sample space. Even with the better performance of multispectral images, machine learning algorithms applied to RGB samples produced promising results for crop classification.

Image enhancement and microstructure characterization of energy dispersive X-ray spectroscopy images of blended cement pastes

Energy dispersive X-ray spectroscopy (EDS) image analysis has been widely used to study the microstructure of cementitious materials, whereas in some cases the efficiency of the characterization suffered from the extremely long time for image acquisition. In this study, we proposed an EDS image analysis methodology including image enhancement, phase extraction, and characterization of reaction products. The guided filter and Logarithm correction were combined and employed to improve the resolvability of low-quality EDS images. Gaussian Mixture Model (GMM) algorithm was applied to segment the unreacted phases and two different groups of reaction products from the enhanced EDS images. Moreover, the spatial distribution and chemical composition of reaction products were characterized using the lineal-path function and simple lineal iterative clustering (SLIC) algorithm. Results showed that low-quality EDS images after enhancing could achieve the comparable performance with high-quality EDS images in terms of the phase proportions. This study offers a practical and effective methodology for enhancing and analyzing the low-quality EDS images, giving insights into the phase assemblage and microstructure evolution of cementitious materials.

Abnormality Determination of Spermatozoa Motility Using Gaussian Mixture Model and Matching-based Algorithm

Sperm analysis is an initial step in the examination conducted to identify infertility cases in humans. One aspect of sperm analysis involves observing the movement of spermatozoa and determining whether it is normal or abnormal. Normal spermatozoa movement is characterized by progressive motion at an average speed of 20 µm/second, while abnormal movement includes slow or non-motile spermatozoa. Traditional methods can be employed to assess the normality or abnormality of sperm movement, but they have drawbacks such as time-consuming procedures and diverse results depending on the expertise of the examiner. On the other hand, utilizing Computer-Assisted Sperm Analysis (CASA) equipment provides consistent results, albeit at a relatively high cost. Therefore, this research proposes an alternative method for determining sperm movement abnormalities using the Gaussian Mixture Model (GMM) for background subtraction and a matching-based algorithm to track and analyze the formed trajectories, distinguishing between normal and abnormal sperm movement. Human spermatozoa in real-time are used, and their movements are recorded in video format using a bright field microscope. The testing results for determining sperm movement abnormalities based on the GMM method and matching-based algorithm were successful, particularly in videos recorded at 50 fps recording speed, 20 minutes of liquefaction time, and 40x microscope lens magnification. This condition exhibited the highest average accuracy, with a tracking accuracy of 77.3% and an average accuracy for determining sperm motility abnormalities of 87.7%. Therefore, the combined tracking of sperm movement based on the GMM method and matching-based algorithm can be utilized to identify abnormalities in the movement of human spermatozoa.

Predictive Maintenance System for Wafer Transfer Robot Using Gaussian Mixture Model and Mean-Shift Clustering

Equipment maintenance is the technology of monitoring the condition of equipment and predicting the error of the equipment. As an interest in equipment maintenance increases, many studies are being conducted on methods for minimizing economic losses due to equipment failure or enormous maintenance costs, especially in the semiconductor manufacturing field. In this paper, a new predictive maintenance (PdM) method is proposed to predict the error of the wafer transfer robot and to classify the significance level of the error. The experimental data of the wafer transfer robot were collected using acceleration sensor. To analyse the data, pre-processing was performed using fast Fourier transformation (FFT). The error levels were defined by classifying the number of error areas into 1, 2 and 3 regions. The Gaussian mixture model (GMM) algorithm was used to classify normal and abnormal states and mean-shift (MS) algorithm was used to classify the error level for the data classified as abnormal state. The proposed method can detect errors and classify the error level accurately in the data where normal data and error data were mixed.The advantage of the proposed method is that it can accurately distinguish errors from normal data by applying GMM with higher accuracy that previous studies and classify the level of errors using MS.

Performance evaluation of modified Gaussian mixture model algorithm for image segmentation of mammograms

Performance evaluation of modified Gaussian mixture model algorithm for image segmentation of mammograms

An Accurate Dynamic Model Identification Method of an Industrial Robot Based on Double-Encoder Compensation

Aiming at the challenges to accurately simulate complex friction models, link dynamics, and part uncertainty for high-precision robot-based manufacturing considering mechanical deformation and resonance, this study proposes a high-precision dynamic identification method with a double encoder. Considering the influence of the dynamic model of the manipulator on its control accuracy, a three-iterative global parameter identification method based on the least square method and GMM (Gaussian Mixture Model) under the optimized excitation trajectory is proposed. Firstly, a bidirectional friction model is constructed to avoid using residual torque to reduce the identification accuracy. Secondly, the condition number of the block regression matrix is used as the optimization objective. Finally, the joint torque is theoretically identified with the weighted least squares method. A nonlinear model distinguishing between high and low speeds was established to fit the nonlinear friction of the robot. By converting the position and velocity of the motor-side encoder to the linkage side using the deceleration ratio, the deformation quantity could be calculated based on the discrepancy between theoretical and actual values. The GMM algorithm is used to compensate the uncertainty torque that was caused by model inaccuracy. The effectiveness of the proposed method is verified by a simulation and experiment on a 6-DoF industrial robot. Results prove that the proposed method can enhance the online torque estimation performance by up to 20%.

A comparative study of Gaussian mixture algorithm and K-means algorithm for efficient energy clustering in MWSN

Wireless sensor networks (WSNs) represent one of the key technologies in internet of things (IoTs) networks. Since WSNs have finite energy sources, there is ongoing research work to develop new strategies for minimizing power consumption or enhancing traditional techniques. In this paper, a novel Gaussian mixture models (GMMs) algorithm is proposed for mobile wireless sensor networks (MWSNs) for energy saving. Performance evaluation of the clustering process with the GMM algorithm shows a remarkable energy saving in the network of up to 92%. In addition, a comparison with another clustering strategy that uses the K-means algorithm has been made, and the developed method has outperformed K-means with superior performance, saving energy of up to 92% at 4,500 rounds.

Speech Emotion Recognition using Extreme Machine Learning

Detecting Emotion from Spoken Words (SER) is the task of detecting the underlying emotion in spoken language. It is a challenging task, as emotions are subjective and highly contextual. Machine learning algorithms have been widely used for SER, and one such algorithm is the Gaussian Mixture Model (GMM) algorithm. The GMM algorithm is a statistical model that represents the probability distribution of a random variable as a sum of Gaussian distributions. It has been widely used for speech recognition and classification tasks. In this article, we offer a method for SER using Extreme Machine Learning (EML) with the GMM algorithm. EML is a type of machine learning that uses randomization to achieve high accuracy at a low computational cost. It has been effectively utilised in various classification tasks. For the planned approach includes two steps: feature extraction and emotion classification. Cepstral Coefficients of Melody Frequency (MFCCs) are used in order to extract features. MFCCs are commonly used for speech processing and represent the spectral envelope of the speech signal. The GMM algorithm is used for emotion classification. The input features are modelled as a mixture of Gaussians, and the emotion is classified based on the likelihood of the input features belonging to each Gaussian. Measurements were taken of the suggested method on the The Berlin Database of Emotional Speech (EMO-DB) and achieved an accuracy of 74.33%. In conclusion, the proposed approach to SER using EML and the GMM algorithm shows promising results. It is a computationally efficient and effective approach to SER and can be used in various applications, such as speech-based emotion detection for virtual assistants, call centre analytics, and emotional analysis in psychotherapy.