Real-time Classification Method Research Articles

FDI Attack Detection at the Edge of Smart Grids Based on Classification of Predicted Residuals

The introduction of information and communication technologies makes network environments increasingly open, leaving smart-grid control systems incredibly vulnerable to malicious attacks. False data injection (FDI) attacks stealthily tamper with measurement data, resulting in erroneous decisions made by the control center that greatly influence the normal operation of the power system. By taking advantage of real-time data acquisition with edge computing, in this article, we propose a scheme based on classification of predicted residuals (CPRs) for the FDI attack detection. The CPR scheme first predicts the acquired measurement data at the edge of the sensing network via developing an accurate prediction model. Followed the novel real-time classification method under the edge devices supporting, it classifies the predicted residuals independent of the false data to enhance the detection accuracy. Through these two steps, the detection rate of FDI attacks is greatly improved. The proposed scheme is validated in a real microgrid testbed. Experimental results show that the CPR scheme performs well in detecting FDI attacks and remains sensitive in injection attack probability and magnitude. The detection scheme even has effectiveness at low injection attack probability and magnitude (5% and 0.018 per thousand, respectively). Furthermore, it also proves that the proposed scheme has applicability in high real-time requirements at the edge of smart grids.

A Nonintrusive and Real-Time Classification Method for Driver’s Gaze Region Using an RGB Camera

Driver distraction has become a leading cause of traffic crashes. Visual distraction has the most direct impact on driving safety among various driver distractions. If the driver’s line of sight deviates from the road in front, there will be a high probability of visual distraction. A nonintrusive and real-time classification method for driver’s gaze region is proposed. A Multi-Task Convolutional Neural Network (MTCNN) face detector is used to collect the driver’s face image, and the driver’s gaze direction can be detected with a full-face appearance-based gaze estimation method. The driver’s gaze region is classified by the model trained through the machine learning algorithms such as Support Vector Machines (SVM), Random Forest (RF), and K-Nearest Neighbors (KNN). The simulated experiment and the real vehicle experiment were conducted to test the method. The results show that it has good performance on gaze region classification and strong robustness to complex environments. The models in this paper are all lightweight networks, which can meet the accuracy and speed requirements for the tasks. The method can be a good help for further exploring the visual distraction state level and exert an influence on the research of driving behavior.

A Wearable Device for Evaluation of Relative Position, Force, and Duration of Fetal Movement for Pregnant Woman Care

Fetal movement (FM) is one of the important indexes of clinical observations on fetal activities (such as the position, duration and relative force of FM), and as demand for the home monitoring of pregnant women grows increasingly, a wearable device for the long-term monitoring of fetal movement has drawn more and more attention. This study integrated multi-point IMU sensing with an innovative real-time classification method to build a device for the long-term monitoring of fetal movement, including the evaluation of the relative position, force and duration. In order to validate the sensitivity and clinical feasibility of the device, this study performed phantom simulation tests and clinical tests on 13 pregnant women. The phantom test results showed that the device had high accuracy (>90.3%) in recognizing 12 FM positions, the relative force had high correlation ( R <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> > 0.98), and the duration of FM had a low error percentage (<; 10%). The clinical test results showed that the number of FMs detected by this method was coincident with the pregnant women's self-perceptions, and the questionnaire result showed that the pregnant women highly accepted this device.

Automatic Fracture Characterization Using Tactile and Proximity Optical Sensing.

This paper demonstrates how tactile and proximity sensing can be used to perform automatic mechanical fractures detection (surface cracks). For this purpose, a custom-designed integrated tactile and proximity sensor has been implemented. With the help of fiber optics, the sensor measures the deformation of its body, when interacting with the physical environment, and the distance to the environment's objects. This sensor slides across different surfaces and records data which are then analyzed to detect and classify fractures and other mechanical features. The proposed method implements machine learning techniques (handcrafted features, and state of the art classification algorithms). An average crack detection accuracy of ~94% and width classification accuracy of ~80% is achieved. Kruskal-Wallis results (p < 0.001) indicate statistically significant differences among results obtained when analysing only integrated deformation measurements, only proximity measurements and both deformation and proximity data. A real-time classification method has been implemented for online classification of explored surfaces. In contrast to previous techniques, which mainly rely on visual modality, the proposed approach based on optical fibers might be more suitable for operation in extreme environments (such as nuclear facilities) where radiation may damage electronic components of commonly employed sensing devices, such as standard force sensors based on strain gauges and video cameras.

Random-forest-based real-time contrasts control chart using adaptive breakpoints with symbolic aggregate approximation

For high yield management, process monitoring has become an increasingly important task. The real-time contrasts (RTC) control chart uses the real-time classification method for process monitoring and outperforms the existing real-time control chart. The original RTC control chart identifies the cause of faults using a random forest classifier. However, the random forest provides discrete monitoring statistics that could make the overall performance less efficient. To improve the performance of the RTC control chart, we propose a random-forest-based RTC control chart that uses adaptive breakpoints with symbolic aggregate approximation (ABP-SAX). The monitoring statistics of the RTC control chart indicate the process condition, and the quality of the monitoring statistics is determined by the classification performance of the classifier. Therefore, to improve the classification performance of individual decision trees, we proposed ABP-SAX. The original SAX causes time-information loss and distortions in the data slope and pattern. We prevent these problems using the mean squared error to minimize the difference between the represented and original data. After the applied ABP-SAX representation, the raw data are represented by categorical values that preserve information from the original data, and the represented data improve the performance of the RTC control chart. Therefore, the proposed RTC control chart could detect shifts more quickly and identify the cause of the faults. Our improvements can contribute to high yield management and quick response to abnormalities.

Support vector machine based real-time terrain estimation for tracked robots

This paper describes the use of support vector machine (SVM) classifier for real-time terrain estimation to improve autonomous navigation of tracked mobile robots. Real-time terrain identification and terrain property estimation has been explored previously for application on a wide range of systems from planetary exploration rovers to commercial vehicles like cars and trucks. A majority of the existing methods rely on the use of dedicated sensors including vibration sensors, accelerometers, cameras, LIDAR, etc., which makes them susceptible to the failure modes of each of these sensors. This work proposes a method for real-time classification of different terrain types based on the state evolution of a ground robot, specifically the measured change in the pose of the robot for a known control input. By using a trained SVM to perform terrain estimation based on the collected state evolution data, the proposed method does not require dedicated sensing modalities solely for the terrain estimation. In addition, this method is generally applicable in all conditions where the robot can traverse. The training data was obtained from four different terrain conditions including vinyl flooring, asphalt, artificial turf, and grass–gravel, to train the SVM to perform terrain estimation. The proposed technique is validated using a skid-steer tracked robot over multiple simulated and real terrain transitions cases, where the response to control inputs is significantly affected by terrain characteristics. The results show that the proposed method provides greater than 80% accuracy in all cases, with fast detection of terrain transitions. The paper concludes with a detailed description on the application of real-time terrain estimation in improving autonomous navigation.

A method for real-time classification of insect vectors of mosaic and brown streak disease in cassava plants for future implementation within a low-cost, handheld, in-field multispectral imaging sensor

BackgroundThe paper introduces a multispectral imaging system and data-processing approach for the identification and discrimination of morphologically indistinguishable cryptic species of the destructive crop pest, the whitefly Bemisia tabaci. This investigation and the corresponding system design, was undertaken in two phases under controlled laboratory conditions. The first exploited a prototype benchtop variant of the proposed sensor system to analyse four cryptic species of whitefly reared under similar conditions. The second phase, of the methodology development, employed a commercial high-precision laboratory hyperspectral imager to recover reference data from five cryptic species of whitefly, immobilized through flash freezing, and taken from across four feeding environments.ResultsThe initial results, for the single feeding environment, showed that a correct species classification could be achieved in 85–95% of cases, utilising linear Partial Least Squares approaches. The robustness of the classification approach was then extended both in terms of the automated spatial extraction of the most pertinent insect body parts, to assist with the spectral classification model, as well as the incorporation of a non-linear Support Vector Classifier to maintain the overall classification accuracy at 88–98%, irrespective of the feeding and crop environment.ConclusionThis study demonstrates that through an integration of both the spatial data, associated with the multispectral images being used to separate different regions of the insect, and subsequent spectral analysis of those sub-regions, that B. tabaci viral vectors can be differentiated from other cryptic species, that appear morphologically indistinguishable to a human observer, with an accuracy of up to 98%. The implications for the engineering design for an in-field, handheld, sensor system is discussed with respect to the learning gained from this initial stage of the methodology development.

Water quality of Danube Delta systems: ecological status and prediction using machine-learning algorithms.

Environmental issues have a worldwide impact on water bodies, including the Danube Delta, the largest European wetland. The Water Framework Directive (2000/60/EC) implementation operates toward solving environmental issues from European and national level. As a consequence, the water quality and the biocenosis structure was altered, especially the composition of the macro invertebrate community which is closely related to habitat and substrate heterogeneity. This study aims to assess the ecological status of Southern Branch of the Danube Delta, Saint Gheorghe, using benthic fauna and a computational method as an alternative for monitoring the water quality in real time. The analysis of spatial and temporal variability of unicriterial and multicriterial indices were used to assess the current status of aquatic systems. In addition, chemical status was characterized. Coliform bacteria and several chemical parameters were used to feed machine-learning (ML) algorithms to simulate a real-time classification method. Overall, the assessment of the water bodies indicated a moderate ecological status based on the biological quality elements or a good ecological status based on chemical and ML algorithms criteria.

Automatic classification of woven fabric structure based on texture feature and PNN

In today’s textile industry, the classification of woven fabrics is usually manual which requires considerable human efforts and a long time. With the rapid development of computer vision, the automatic and efficient methods for woven fabric classification are desperately needed. This paper proposes an automatic and real-time classification method to analyze three woven fabrics: plain, twill and satin weave. The methodology involves two approaches to extract texture features, that is, gray-level co-occurrence matrix (GLCM) and Gabor wavelet. Then, principal component analysis (PCA) is utilized to deal with the texture feature vectors to gain minimize redundancy and maximize principal component feature vectors. Finally, in the classification phase, probabilistic neural network (PNN) is applied to classify three basic woven fabrics. With strong realtime, fault-tolerance and non-linear classification capability, PNN can be a promising tool for classification of woven fabrics. The experimental results show that PNN classifier with faster training speed can classify woven fabrics accurately and efficiently. Besides, compared with GLCM method and Gabor wavelet method, the fusion of the two feature vectors obtains the best classification result (95 %).

A new feature extraction method for odour classification

A new method for real time classification of volatile chemical substance traces is presented. The method is based on electrochemical signals of an array of semiconductor gas sensors. In these sensor signals characteristic patterns of different substances are hidden. There are non-linear correlative relationships between the measured sensor signals and the chemical substances which are treated using two methods derived from statistical learning theory (Support Vector Machine – SVM, Maximum Likelihood Estimation – MLE) for the detection of the substance characteristics in the sensor signals. A key criterion for the presented pattern recognition is a newly developed type of features, which is specially adapted to the low frequency signals of semiconductor sensors. The presented features are based on the evaluation of the range of the transient response in the sensor signals in the frequency domain. To derive the new features, both real measurement data and synthetic generated signals were used. In the experiments the focus was set on the creation of reproducible sensor signals to get characteristic signal patterns. Synthetic signals were derived from a Gaussian Plume Model. With the new features, training data sets were calculated using the classification methods SVM and MLE. With these training data sets new sensor measurements may be assigned to the substances which are to be sought. The advantage of the presented method is that no feature reduction is needed and no loss of information occurs in the learning process. The classification results based on the new features have been compared with the classification based on a conventional method for feature extraction. It was proved that the recognition rate of the substances used with the new feature type is higher. The substance classification is primarily limited by the sensitivity of the semiconductor sensors, because sufficiently large sensor signals must have been provided to obtain appropriate substance patterns. At the present stage of development the method presented is suitable for the classification of substance groups, such as nitro aromatics or alcohols, but not for specific substances.

A real-time classification method of power quality disturbances

This paper describes a real-time classification method of power quality (PQ) disturbances. With an acceptable computation burden, both the elementary parameters of the power signal and the types of the disturbances in the power signal are obtained easily. The proposed method addresses the selection of discriminative features for detection and classification of PQ disturbances. Five distinguished time-frequency statistical features of PQ disturbances are extracted using RMS (root-mean-square) method and discrete Fourier transform (DFT). Using a rule-based decision tree (RBDT), the nine types of PQ disturbances can be recognized easily and there is no need to use other complicated classifiers. Finally, the proposed method is tested using the simulated waveforms. And some preliminary experimental results of the accuracy characterization of an initial development instrument are reported. The simulation and application results validate the accuracy and efficiency of the proposed method.

Classifying organic crystals via in-process image analysis and the use of monitoring charts to follow polymorphic and morphological changes

The polymorphic and morphological forms, and growth rates of different crystal facets are important parameters in pharmaceutical crystallisation processes that call for techniques for their real-time monitoring. In this paper, we describe a method for real-time classification of crystal polymorphic and morphological forms based on on-line imaging and image analysis. The focus of discussion is on deriving quantitative shape descriptors from images obtained on-line in real-time, automatic classification of crystals, the estimation of ratios of different polymorphs and new monitoring charts showing the relative average growth rates of facets of crystals. The work presented is based on experimental results of crystallisation in a batch reactor of ( L)-glutamic acid that has two polymorphs, the prismatic alpha and needle-shaped beta forms.

A NEW APPROACH FOR CLASSIFICATION OF GENERIC AUDIO DATA

The existing audio retrieval systems fall into one of two categories: single-domain systems that can accept data of only a single type (e.g. speech) or multiple-domain systems that offer content-based retrieval for multiple types of audio data. Since a single-domain system has limited applications, a multiple-domain system will be more useful. However, different types of audio data will have different properties, this will make a multiple-domain system harder to be developed. If we can classify audio information in advance, the above problems can be solved. In this paper, we will propose a real-time classification method to classify audio signals into several basic audio types such as pure speech, music, song, speech with music background, and speech with environmental noise background. In order to make the proposed method robust for a variety of audio sources, we use Bayesian decision function for multivariable Gaussian distribution instead of manually adjusting a threshold for each discriminator. The proposed approach can be applied to content-based audio/video retrieval. In the experiment, the efficiency and effectiveness of this method are shown by an accuracy rate of more than 96% for general audio data classification.