Feature Extraction Algorithm Research Articles

Multi-modal adaptive feature extraction for early-stage weak fault diagnosis in bearings

We present a novel multi-modal adaptive feature extraction algorithm considering both time-domain and frequency-domain modalities (AFETF), coupled with a Bidirectional Long Short-Term Memory (Bi-LSTM) network based on the Grey Wolf Optimizer (GWO) for early-stage weak fault diagnosis in bearings. Singular Value Decomposition (SVD) was employed for noise reduction, while Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) was utilized for signal decomposition, facilitating further signal processing. AFETF algorithm proposed in this paper was employed to extract weak fault features. The adaptive diagnostic process was further enhanced using Bi-LSTM network optimized with GWO, ensuring objectivity in the hyperparameter optimization. The proposed method was validated for datasets containing weak faults with a 0.2 mm crack and strong faults with a 0.4 mm crack, demonstrating its effectiveness in early-stage fault detection.

Research on the Generative Mechanism of Intelligent Reconstruction for the Integration of Traditional Paper-cutting Art Symbols and Modern Pattern Designs

Abstract In the realm of Chinese traditional arts, paper-cutting symbols embody deep-seated aesthetic and cultural values. Within the framework of artificial intelligence, the fusion of these conventional symbols with contemporary pattern design through intelligent reconstruction processes promotes innovation in modern design. This paper introduces a novel pattern feature extraction algorithm specifically designed for identifying and extracting the essence of traditional paper-cutting art. Utilizing Deep Convolutional Generative Adversarial Networks (DCGAN), an automated method for generating woven patterns is developed, integrating conventional motifs with modern design elements. Additionally, a segmentation model based on DeepLabV3+ semantics is constructed to facilitate this integration. The study focuses on the fusion of traditional paper-cutting symbols with fabric patterns, providing a new direction for textile design research. The empirical evaluation involved seven experimenters who predominantly favored patterns that blended abstract and traditional styles, with the average style rating exceeding 3.2. Moreover, the color performance analysis of the reconstructed fabric patterns across five regions showed color difference values greater than 4, indicating superior color fidelity. These findings underscore the potential of combining AI-driven techniques with traditional art forms to enhance and revolutionize modern design practices.

Image feature extraction algorithm based on visual information

Abstract Vision is the main sensory organ for human beings to contact and understand the objective world. The results of various statistical data show that more than 60% of all ways for human beings to obtain external information are through the visual system. Vision is of great significance for human beings to obtain all kinds of information needed for survival, which is the most important sense of human beings. The rapid growth of computer technology, image processing, pattern recognition, and other disciplines have been widely applied. Traditional image processing algorithms have some limitations when dealing with complex images. To solve these problems, some scholars have proposed various new methods. Most of these methods are based on statistical models or artificial neural networks. Although they meet the requirements of modern computer vision systems for feature extraction algorithms with high accuracy, high speed, and low complexity, these algorithms still have many shortcomings. For example, many researchers have used different methods for feature extraction and segmentation to get better segmentation results. Scale-invariant feature transform (SIFT) is a description used in the field of image processing. This description has scale invariance and can detect key points in the image. It is a local feature descriptor. A sparse coding algorithm is an unsupervised learning method, which is used to find a set of “super complete” basis vectors to represent sample data more efficiently. Therefore, combining SIFT and sparse coding, this article proposed an image feature extraction algorithm based on visual information to extract image features. The results showed that the feature extraction time of X algorithm for different targets was within 0.5 s when the other conditions were the same. The feature matching time was within 1 s, and the correct matching rate was more than 90%. The feature extraction time of Y algorithm for different targets was within 2 s. The feature matching time was within 3 s, and the correct matching rate was between 80 and 90%, indicating that the recognition effect of X algorithm was better than that of Y algorithm. It indicates the positive relationship between visual information and image feature extraction algorithm.

Four-axis dual-band error avoidance attitude measurement algorithm with adaptive active differential interference feature extraction

Infrared radiation attitude measurement technology is a new, non-inertial, satellite independent and fast-responding attitude parameters testing method. However, the random occurrence of infrared radiation during missile flight, such as clouds or solar radiation, can cause significant interference in infrared radiation attitude measurement. To enhance the anti-interference capability of infrared radiation attitude measurement, the adaptive active differential (AAD) interference feature extraction algorithm and the four-axis error avoidance (FEA) attitude calculation algorithm with dual-band infrared radiation (DIR) are proposed. Initially, the long-wave and mid-wave infrared radiation laws of clouds and solar radiation are analyzed to identify their different characteristics in both wavebands. The interference type encountered by the rotating missile during flight is extracted by the designed AAD algorithm. Subsequently, based on the results of AAD algorithm, the adjacent axis long-wave or mid-wave infrared sensor data with less or no interference influence is selected. Finally, the attitude angle information is obtained by using two-axis or four-axis attitude calculation algorithm. The results of the semi-physical experiments indicate that the four-axis dual-band infrared attitude measurement device using the FEA algorithm can achieve RMSE about 1 even under the interference of solar or clouds cover.

A Feasibility Analysis of the Use of ISAR Training Data in Machine Learning-Based SAR ATR

Processing of synthetic aperture radar (SAR) images for automatic target recognition (ATR) is a critical application especially in military surveillance. In particular, numerous machine learning-based SAR ATR methods have been proposed for this task. However, data training and testing stages of all these methods are based on the exploitation of SAR signatures of the target under investigation. Considering the high variability of radar targets, obtaining such signature data is obviously a costly and time consuming process. In this study, therefore, a feasibility analysis of the use of inverse-SAR (ISAR) training data in SAR ATR has been made for the first time. The turntable ISAR and circular SAR images of three different vehicles are used in training and testing is performed by means of SAR images of three similar targets within the publicly available MSTAR dataset. Also, three most prominent machine learning methods, namely KNN, SVM and ANN are used in conjunction with three different feature extraction algorithms namely, GLRLM, GLSZM and GLCM. The obtained results reveal that the GLCM+SVM algorithm pair is the most effective model with 85% accuracy.

Research on Physical Education Management System in Higher Education Institutions in the Context of Deep Learning

Abstract This paper focuses on the management of physical movement sports disciplines and combines relevant algorithms to design an artificial intelligence-based physical education management system framework. In terms of related algorithms, the first is for the research of character recognition algorithms. Considering the intra-class differences between individual movements, the LR-based feature extraction algorithm is proposed. Secondly, for the research of the action evaluation algorithm, the statistical value, DTW parameter action time series difference and correlation coefficient based on the statistical value and DTW parameter are used as the features, combined with SVM-based feature extraction algorithm for perceiving the action differences. Finally, with the help of deep neural networks for movement analysis, we can improve students’ movement techniques during training. In the course schedule supported by the system, the proportion of moderate-intensity exercise time is high, basically, more than 40%, which meets the WHO requirements for high-intensity exercise time in physical education classes. The post-course RPE feedback from students was spread out between 12 and 15, and the highest post-course RPE value was 14.82<unk>1.38, which was evaluated within a reasonable range of intervals. The teaching management system in this paper can meet the requirements of physical training and teaching, and provide quantifiable basis and management tools for the informatization and precise teaching of physical movement subjects.

SSTM: Semantic Segmentation-based Template Matching method for heterogenous images

Global navigation satellite systems (GNSS) have disadvantages such as signal interference, loss, and spoofing. This has led to the development of new methods to complement or replace satellite navigation. Absolute visual localization is one of the main methods of vision-based localization. The goal is to localize the current view of the UAV in a reference satellite map or in a georeferenced image of a previous flight. During nighttime, due to poor lighting conditions, infrared images are used as a substitute for visible light images. However, the significant differences between infrared and visible light images pose challenges in matching them. Therefore, this paper proposes a Semantic Segmentation-based Template Matching (SSTM) method for heterogeneous images. The method proposed in this paper breaks the general idea of using a single feature extraction algorithm for heterogeneous image matching. In this paper, the matching of UAV infrared images with visible light satellite remote sensing images is realized by unifying two different image feature extraction modules of the feature description language and the matching scheme of this feature description language. The advantages of this method are demonstrated through comparisons with other advanced techniques. With the development of semantic segmentation techniques, it is believed that this modular “translation” method can achieve matching between arbitrary sensors.

Using Feature Extraction to Perform Equipment Health Monitoring on Ship-Radiated Noise

The current state of the art in hydroacoustics research employs a variety of feature extraction techniques with the goal of accurately classifying a ship based on its radiated noise. These techniques are capable of accuracy in excess of 95%. A question arises as to whether similar techniques could be applied to a known vessel to identify and monitor individual systems from the ship’s noise. In this paper, the fast orthogonal search algorithm is used as a basis for a feature extraction and classification algorithm. This algorithm is applied to real recordings of ship-radiated noise and is shown to be capable of identifying the running status of a subset of the ship’s systems, providing a proof of concept for the detection and monitoring of a ship’s systems based solely on the ships hydroacoustic noise.

A SUPERPOINT NEURAL NETWORK IMPLEMENTATION FOR ACCURATE FEATURE EXTRACTION IN UNSTRUCTURED ENVIRONMENTS

Abstract. Feature extraction plays a crucial role in visual localization, SLAM (Simultaneous Localization and Mapping) and autonomous navigation, by enabling the extraction and tracking of distinctive visual features for both mapping and localization tasks. However, most of the studies, investigate the efficiency and performance of the algorithms in urban, vegetated or indoor environments and not in unstructured environments which suffers by poor information in visual cues where a feature extraction algorithm or architecture could base on. In this study, an investigation of SuperPoint architecture’s efficiency in keypoint detection and description applied to unstructured and planetary scenes was conducted, producing three different models: (a) an original SuperPoint model trained from scratch, (b) an original fine-tuned SuperPoint model, (c) an optimized SuperPoint model, trained from scratch with the same parametarization as the corresponding original model. For the training process, a dataset of 48 000 images was utilized representing unstructured scenes from Earth, Moon and Mars while a benchmark dataset was used aiming to evaluate the model in illumination and viewpoint changes. The experimentation proved that the optimized SuperPoint model provides superior performance using repeatability and homography estimation metrics, compared with the original SuperPoint models, and handcrafted keypoint detectors and descriptors.

A New Algorithm for Speech Feature Extraction Using Polynomial Chirplet Transform

A New Algorithm for Speech Feature Extraction Using Polynomial Chirplet Transform

Classification Insights into Brain MRI Classification: Techniques, Interpretability, and Future

This research paper comprehensively analyses various techniques for classifying brain MRI images. Through an extensive literature survey, the study explores and evaluates the effectiveness of different methodologies, ranging from traditional machine learning approaches to advanced deep learning models, in the context of brain tumour classification. The review critically examines feature extraction, selection methods, and classification algorithms, focusing on their performance in enhancing diagnostic accuracy. The synthesis of existing literature provides valuable insights into the current landscape of brain MRI classification techniques, shedding light on their strengths, limitations, and potential areas for future research.

Quantitative analysis of energy-dispersive X-ray fluorescence spectroscopy based on machine learning and a generative data enhancement technique

This paper proposes a data enhancement technique to generate expanded datasets for machine learning by developing an X-ray fluorescence spectra simulator based on the physical process. The simulator consists of several modules, including the excitation source, the interaction process, and the detection system. The spectra generated by the simulator are subject to dimension reduction through feature selection and feature extraction algorithms, and then serve as the input for the XGBoost (extreme gradient boosting) model. Six elements of metal samples with various content ranges were selected as the research target. The results showed that for simulated data, the R2 value for elements with concentrations ranging from 0% to 100% is greater than 95%, and for elements with concentrations of <0.3%, the R2 value is greater than 85%. The experimental data were predicted by the model trained by the simulated spectra. Therefore, this approach provides reliable results for practical application and can supply additional datasets to obtain reasonable prediction results for machine learning with inadequate reference materials.

Decision support for personalized therapy in implantable medical devices: A digital twin approach

Implantable Medical Devices (IMDs) offer timely therapeutic interventions for life-threatening conditions without disrupting patients’ daily activities. Given the substantial variability in individual health conditions, customizing device therapies to suit specific patients and their evolving needs is crucial. Although current customization relies on modifying device parameter settings, no universally accepted clinical guidelines exist for this form of customization, particularly for patients with complex or rare conditions. To address this gap, we introduce a decision support framework that leverages expert system methodologies for personalized therapy in IMDs. This comprehensive expert system encompasses a dynamic digital twin of the patient, built using physiological data obtained from the IMD and additional wearable devices. The digital twin captures disease mechanisms and patient-specific physiological parameters, while a periodically trained Reinforcement Learning (RL) agent and specialized algorithms for the ongoing maintenance of the digital twin complete the expert system. Together, these elements facilitate periodic, informed therapy customization based on up-to-date patient conditions. Our approach includes novel algorithms for feature extraction, physiological modeling, and the maintenance of the digital twin. We substantiate the efficacy and safety of our digital-twin-based decision support framework through a case study focused on Implantable Cardioverter Defibrillators (ICDs). Experimental results on virtual patients confirm that our integrated framework surpasses traditional customization strategies in both therapeutic effectiveness and patient safety considerations.

A Digitalization Algorithm Based on the Voltage Waveform of the Multifunction Vehicle Bus

The MVB is a kind of widely used vehicle-level bus, which is crucial for the normal operation of trains. However, the MVB bus contains device node terminals and the topology is complex, which makes fault location difficult. The voltage waveform of the MVB physical layer can reflect the electrical characteristics and fault characteristics of a network in real time and has the value of assisting the diagnosis of MVB network faults. Based on the characteristics of the MVB physical-layer voltage waveform, this paper studies the optimal sampling rate and feature-extraction algorithm of the MVB voltage waveform and the combination of the data of the data-link layer to assign accurate timestamps to waveform data, and finally presents a design for an MVB dual-mode data-acquisition platform with multiple sampling rates/test conditions. Experimental results show that the proposed algorithm can accurately extract waveform feature information under the optimal sampling rate of 62.5 MHz, which makes the collection of MVB voltage waveform data more reliable and practical.

Machine learning approaches for cardiovascular hypertension stage estimation using photoplethysmography and clinical features.

Cardiovascular diseases (CVDs) are a leading cause of death worldwide, with hypertension emerging as a significant risk factor. Early detection and treatment of hypertension can significantly reduce the risk of developing CVDs and related complications. This work proposes a novel approach employing features extracted from the acceleration photoplethysmography (APG) waveform, alongside clinical parameters, to estimate different stages of hypertension. The current study used a publicly available dataset and a novel feature extraction algorithm to extract APG waveform features. Three distinct supervised machine learning algorithms were employed in the classification task, namely: Decision Tree (DT), Linear Discriminant Analysis (LDA), and Linear Support Vector Machine (LSVM). Results indicate that the DT model achieved exceptional training accuracy of 100% during cross-validation and maintained a high accuracy of 96.87% on the test dataset. The LDA model demonstrated competitive performance, yielding 85.02% accuracy during cross-validation and 84.37% on the test dataset. Meanwhile, the LSVM model exhibited robust accuracy, achieving 88.77% during cross-validation and 93.75% on the test dataset. These findings underscore the potential of APG analysis as a valuable tool for clinicians in estimating hypertension stages, supporting the need for early detection and intervention. This investigation not only advances hypertension risk assessment but also advocates for enhanced cardiovascular healthcare outcomes.

A color extraction algorithm by segmentation

The segmentation and extraction on color features can provide useful information for many different application domains. However, most of the existing image processing algorithms on feature extraction are gray image-based and consider only one-dimensional parameters. In order to carry out a fast and accurate color feature extraction, this paper proposes a color extraction algorithm by segmentation that is called a color extraction algorithm This algorithm is compared under different color distribution situations, and the extraction effect on color is also shown by the combination of the segmentation and feature extraction algorithms. Experimental results show that such segmentation algorithm has some advantages for color segmentation. In the fuzzy color image preprocessing, this paper gives the location method of region of interest. Moreover, compared with other existing extraction algorithms, the presented segmentation extraction algorithm in this paper not only has higher accuracy, shorter extraction time and stronger anti-interference ability, but also has better effect on more divergent color edge. Experimental evaluation of the proposed color extraction algorithm demonstrates dominance over existing algorithms for feature extraction. These researches in this paper provide a new way of thinking for color feature extraction by segmentation, which has an important theoretical references and practical significance.

Predicting Groundwater Level using Temporal Attention Enhanced Graph Neural Network

A revolutionary approach for groundwater management is essential. To predicting the water levels using for integrating IoT sensors, cloud computing, and advanced data analysis methods. IoT sensors are employed for real-time measurement of groundwater levels, creating a robust dataset. The paper focuses on predicting future groundwater levels, a crucial aspect of sustainable resource management. To enhance predictive accuracy, a preprocessing algorithm, such as Min-Max normalization, is introduced to clean and normalize the collected data, ensuring its reliability. Additionally, a feature extraction algorithm, such as Principal Component Analysis (PCA), is implemented to identify relevant patterns and trends within the dataset, enhancing the efficiency of subsequent analysis. A novel classification algorithm, Spatial-Temporal Graph Convolutional Network, is introduced, enabling the identification of potential groundwater recharge areas. This classification algorithm leverages historical data and extracted features to categorize regions based on their suitability for groundwater renewal. Finally, this research uses a Temporal attention-enhanced graph Neural Network machine learning algorithm to predict groundwater levels in the next few years. This algorithm utilizes the preprocessed data and extracted features, identifying intricate patterns and trends in historical data to generate precise predictions for groundwater levels in the upcoming years.

Bibliometric Study : Rainfall Classification - Prediction using Machine Learning Methods

This Research aims to review the machine learning methods used for classifying or predicting rainfall, using various features from existing data. The study used a bibliometric approach to search for metadata related to rainfall classification and prediction studies using machine learning keywords in Scopus journals. There found 94 metadata articles stored in a Comma Separated Values (CSV) file. The data in this article used published articles from 2014 to 2023 with relevant topics. The study provides information on the latest machine learning methods used for classifying or predicting rainfall. The findings of the study include an increase in published articles by 221.43% from 2018 to 2022. The article titled "An extensive evaluation of seven machine learning methods for rainfall prediction in weather derivatives" by Cramer S., Kampouridis M, Freitas A.A, and Alexandridis A.K received the highest citation count of 129. The study also classified 15 keywords into 3 clusters, with common and fewer keywords. India emerged as the country with the most publications on classifying or predictiong rainfall, and the subject areas of computer science and engineering dominated the distribution of articles. Developing the use of deep learning methods and adding feature extraction algorithms in selecting features used to model data can improve the efficiency and accuracy of the rainfall classification - prediction process. The development of research data using radar images with the type of image processing research can also be maximised for research related to classification - prediction of rainfall using machine learning methods.

Forensic Characterization of Mobile Phone Cases Using Infrared Spectroscopy with Deep Learning Methods

The fast and nondestructive identification of mobile phone cases is a crucial task for law enforcement personnel in the field of forensic science. Attenuated total reflection-Fourier transform infrared spectroscopy (ATR-FTIR) has been employed in conjunction with feature extraction and deep learning algorithms to develop a method for distinguishing mobile phone cases. As preprocessing steps, multivariate scatter correction and standard normal variate were performed. Feature variables were extracted using the competitive adaptive reweighted sampling and shuffled frog leaping algorithm. Several supervised pattern recognition methods, namely ResNet50, VGG16, and multilayer perceptron (MLP), were utilized for constructing classifiers. The results indicated that the model’s performance based on multivariate scattering correction (MSC) with competitive adaptive re-weighted sampling (CARS) surpassed that of other approaches. Notably, the ResNet50 model exhibited superior discriminative ability compared to the VGG16 and MLP models, achieving 100% accuracy in the differentiation of all samples. This designed method represents a potentially rapid and nondestructive approach for identifying mobile phone cases in forensic science.

Sliding principal component and dynamic reward reinforcement learning based IIoT attack detection

The Internet of Things (IoT) involves the gathering of all those devices that connect to the Internet with the purpose of collecting and sharing data. The application of IoT in the different sectors, including health, industry has also picked up the threads to augment over the past few years. The IoT and, by integrity, the IIoT, are found to be highly susceptible to different types of threats and attacks owing to the networks nature that in turn leads to even poor outcomes (i.e., increasing error rate). Hence, it is critical to design attack detection systems that can provide the security of IIoT networks. To overcome this research work of IIoT attack detection in large amount of evolutions is failed to determine the certain attacks resulting in a minimum detection performance, reinforcement learning-based attack detection method called sliding principal component and dynamic reward reinforcement learning (SPC–DRRL) for detecting various IIoT network attacks is introduced. In the first stage of this research methodology, preprocessing of raw TON_IoT dataset is performed by employing min–max normalization scaling function to obtain normalized values with same scale. Next, with the processed sample data as output, to extract data from multi-sources (i.e., different service profiles from the dataset), a robust log likelihood sliding principal component-based feature extraction algorithm is applied with an arbitrary size sliding window to extract computationally-efficient features. Finally, dynamic reward reinforcement learning-based IIoT attack detection model is presented to control the error rate involved in the design. Here, with the design of dynamic reward function and introducing incident repository that not only generates the reward function in an arbitrary fashion but also stores the action results in the incident repository for the next training, therefore reducing the attack detection error rate. Moreover, an IIoT attack detection system based on SPC–DRRL is constructed. Finally, we verify the algorithm on the ToN_IoT dataset of University of New South Wales Australia. The experimental results show that the IIoT attack detection time and overhead along with the error rate are reduced considerably with higher accuracy than that of traditional reinforcement learning methods.