Real-time Pattern Recognition Research Articles

Pattern recognition of urban traffic accidents: Application of cluster analysis and support vector machine in accident pattern recognition

Urban traffic accidents impose a significant threat to public safety because of its frequent occurrence and potential for severe injuries and fatalities. Hence, an effective analysis of accident patterns is crucial for designing accident prevention strategies. Recent advancement in data analytics have provided opportunities to improve the pattern of urban traffic accidents. However, the existing works face several challenges in adapting the complex dynamics, and heterogeneity of the accident data. To overcome these challenges, we proposed an innovative solution by combining the K-means clustering and Support Vector Machine to precisely predict the traffic accident patterns. By leveraging the efficiencies of clustering technique and machine learning, this work intends to identify the intricate patterns within the traffic database. Initially, a traffic accident database was collected and fed into the system. The collected database was pre-processed to improve and standardize the raw dataset. Further, cluster analysis is employed to identify distinct patterns within the dataset and group similar accidents into clusters. This clustering enables the system to recognize common accident scenarios and identify recent accident trends. Subsequently, a Support Vector Machine is deployed to classify accidents into distinct categories through intensive training with identified clusters. The combination enables the system to understand the complex relationships among diverse accident variables, making it an effective framework for real-time pattern recognition. The proposed strategy is implemented in Python and validated using the publicly available traffic accident database. The experimental results manifest that the proposed method achieved 99.65% accuracy, 99.53% precision, 99.62% recall, and 99.57% f-measure. Finally, the comparison with the existing techniques shows that the developed strategy offers improved accuracy, precision, recall, and f-measure compared to existing ones. shows that the developed strategy offers improved accuracy, precision, recall, and f-measure compared to existing ones.

Gastroenterologist-level detection of gastric precursor lesions and neoplasia with a deep convolutional neural network

Background: Gastric precursor lesions and neoplasia with very delicate changes in the gastric mucosa could be easily missed or misdiagnosed under endoscopy. Here we developed an automatic real-time pattern recognition tool based on convolutional neural networks (CNNs) algorithm to help endoscopists in detection of chronic atrophic gastritis (CAG) and gastric cancer (GC) lesions. Methods: The five-convolution-layer ZF model and thirteen-convolution-layer VGG16 model were combined in our neural network A total of 10,014 CAG and 3724 GC annotated images were used in the network training. Another independent set consisted of 50 CAG, 50 GC and 100 negative controls images were used to evaluate the performance of the final network. Results: In CAG detection, the performance of our model was much better than the average performance of the 77 endoscopists in sensitivity, specificity and accuracy (95% versus 74%, 86% versus 82%, 90% versus 78%, respectively). In GC detection, the performance of our model achieved a slightly higher sensitivity (90% versus 87%), but a lower specificity (50% versus 74%) and accuracy (70% versus 80%) than the average performance of the 89 endoscopists. Conclusion: In conclusion, we provided a CNN based computational tool to improve the detection of CAG and GC under endoscopy, and simplify diagnostic procedures.

New technologies for UAV navigation with real-time pattern recognition

The study is dedicated to the investigation and analysis of real-time navigation technologies for UAVs utilizing image scanning and object recognition. This research employs the DJI Inspire 2 drone alongside MapBotix software, integrating adapted algorithms such as Faster R-CNN and MaskRCNN. During the flight of the DJI Inspire 2 drone over the Banska Bystrica region, information concerning object recognition and change detection within the studied area was collected. A 40-second scanning process documented two alterations, one of which occurred at the 10th second. Subsequent stages of image recognition were executed utilizing scaling techniques, thus ensuring the accuracy of data extracted from captured images. This research, in particular, pioneers the exploration of real-time UAV navigation technologies integrated with image recognition. The application of a customized MaskRCNN algorithm, comprising 9 convolutional layers, 4 max-pooling layers, and 1 detection layer, for analyzing scanned object images signifies a novel approach in this domain.

A Novel BILSTM-CNN Method for Pattern Recognition in Real Time Under Triple Physical Loads in Lower Extremity Exoskeleton

Pattern recognition is of great importance in compliant control of lower extremity exoskeleton. We propose a novel method based on the BILSTM-CNN model for pattern recognition in real time under triple physical loads on different terrains. BILSTM is skilled in dealing with temporal series data while CNN is proficient in coping with spatial series data. Five patterns include level ground walking (LW), stair ascending (SA), stair descending (SD), ramp ascending (RA), and ramp descending (RD). Their accuracies in multigrade loads of 0kg, 20kg, and 40kg reach 98.81%, 98.24%, and 97.04% respectively. Moreover, compared with the universal methods of LSTM, CNN, and BP, the hybrid method has competitive advantage in evaluation indexes of accuracy, precision, recall, and F1 score. In addition to five steady patterns, eight pattern transitions are identified between two neighboring states, such as LW to SA, LW to SD, LW to RA, LW to RD, SA to LW, SD to LW, RA to LW, and RD to LW. For pattern transitions, prediction time (Pre-T) of next pattern in multilevel loads of 0kg, 20kg, and 40kg are 190-620 ms, 180-420 ms, and 50-90 ms respectively prior to the step into that pattern. Pattern transition time (Aug-In) in multilevel loads of 0kg, 20kg, and 40kg are 50-190 ms, 30-310 ms, and 0-280 ms respectively before Pre-T of next pattern. Eventually, the experimental results indicate the proposed method has excellent performance on pattern recognition and pattern transition.

Silicon vertex detector with timing for the Upgrade II of LHCb

LHCb has recently submitted a physics case for an Upgrade II detector to begin operation in 2031. The upcoming upgrade is designed to run at instantaneous luminosities of 1.5×1034cm−2s−1, to accumulate a data sample with a corresponding integrated luminosity of over 300fb−1. The LHCb physics programme relies on an efficient and precise vertex detector (VELO). Compared to Upgrade I, the data output rates, radiation levels and occupancies will be about ten times higher. To cope with the pile-up increase, new techniques to assign b-quark–hadrons to their primary vertex, and to perform the real-time pattern recognition are needed. To solve these problems, a new 4D hybrid pixel detector with enhanced rate and timing capabilities in the ASIC and sensor will be developed. This report will discuss the most promising technologies to be used in the future upgrade for the HL-LHC, with emphasis on the timing precision as a tool for vertexing in the next generation detectors. An initial simulation effort has been made to investigate what would be the required temporal resolution sufficient to mitigate pile-up and identify secondary vertices, which points to at least 20 ps per track. The most recent results from beam tests motivated by time measurements will be presented together with the R&D scenarios for the future upgrade. Improvements in the mechanical design of the Upgrade II VELO will also be needed to allow for periodic module replacement. The design will be further optimised to minimise the material before the first measured point on a track and to achieve a fully integrated module design with thinned sensors and ASICs combined with a lightweight cooling solution.

Real-time Accelerometer-Based Event Detection Through KNN Classification Method

Real-time pattern and activity recognition techniques have experienced a recent surge in the use of advanced statistical techniques for analyzing data. More explicitly, Machine Learning classification methods have been used in a variety of applications and for different purposes, such as recognizing different types of events from accelerometer data (e.g. in smart watches, wearables, and vibration analysis). The objective of this work is to present a novel use of accelerometer-based pattern recognition techniques to detect the urban space infrastructure characteristics based on their effect on a vehicle’s body frame dynamics. For this study, we focus on detecting speed bumps, potholes and curves based on a real-time data streaming with the objective of rendering a street digitalization to automatically update an urban space Digital Twin. The present study has achieved real-time event detection for an urban trajectory with high accuracy which yields data that can be later fused with other sensors data. This approach is effectively contributing to the infrastructure layer of a multi-layered approach of digital modelling. Furthermore, the results contribute to expand the body of knowledge of real-time accelerometer event recognition techniques.

High-precision large-area muon tracking and triggering with drift-tube chambers at future colliders

Experiments like the ATLAS detector at the HL-LHC or detectors at future hadron colliders need muon detectors with excellent momentum resolution at the percent level up to the TeV scale both at the trigger and the offline reconstruction level. This requires muon tracking chambers with high spatial resolution even at the highest background fluxes. Drift-tube chambers are the most cost effective technology for the instrumentation of large-area muon systems providing the required high rate capability and three-dimensional spatial resolution. Thanks to the advances in analog and digital electronics, modern drift-tube chambers can be used in stand-alone mode up to the highest background rates providing event times and second coordinates without the necessity of additional trigger chambers. New key developments in the integrated front-end electronics are fast baseline restoration of the shaped signal and picosecond time-to-digital converters for second coordinate measurement with double-sided read-out of the tubes. Self-triggered operation has become possible using modern high-performance FPGAs allowing for real-time pattern recognition and track reconstruction.

Computing metasurfaces for all-optical image processing: a brief review

Abstract Computing metasurfaces are two-dimensional artificial nanostructures capable of performing mathematical operations on the input electromagnetic field, including its amplitude, phase, polarization, and frequency distributions. Rapid progress in the development of computing metasurfaces provide exceptional abilities for all-optical image processing, including the edge-enhanced imaging, which opens a broad range of novel and superior applications for real-time pattern recognition. In this paper, we review recent progress in the emerging field of computing metasurfaces for all-optical image processing, focusing on innovative and promising applications in optical analog operations, image processing, microscopy imaging, and quantum imaging.

Pattern recognition on FPGA for aerospace applications

This paper presents a low power near real-time pattern recognition technique based on Mathematical Morphology-MM implemented on FPGA (Field Programmable Gate Array). The key to the success of this approach concerns the advantages of machine learning paradigm applied to the translation invariant template-matching operators from MM. The paper shows that compositions of simple elementary operators from Mathematical Morphology based on ELUTs (Elementary Look-Up Tables) are very suitable to embed in FPGA hardware. The paper also shows the development techniques regarding all mathematical modeling for computer simulation and system generating models applied for hardware implementation using FPGA chip. In general, image processing on FPGAs requires low-level description of desired operations through Hardware Description Language-HDL, which uses high complexity to describe image operations at pixel level. However, this work presents a reconfiguring pattern recognition device implemented directly in FPGA from mathematical modeling simulation under Matlab/Simulink/System Generator environment. This strategy has reduced the hardware development complexity. The device will be useful mainly when applied on remote sensing tasks for aerospace missions using passive or active sensors.

Reservoir computing with biocompatible organic electrochemical networks for brain-inspired biosignal classification.

Early detection of malign patterns in patients' biological signals can save millions of lives. Despite the steady improvement of artificial intelligence-based techniques, the practical clinical application of these methods is mostly constrained to an offline evaluation of the patients' data. Previous studies have identified organic electrochemical devices as ideal candidates for biosignal monitoring. However, their use for pattern recognition in real time was never demonstrated. Here, we produce and characterize brain-inspired networks composed of organic electrochemical transistors and use them for time-series predictions and classification tasks using the reservoir computing approach. To show their potential use for biofluid monitoring and biosignal analysis, we classify four classes of arrhythmic heartbeats with an accuracy of 88%. The results of this study introduce a previously unexplored paradigm for biocompatible computational platforms and may enable development of ultralow-power consumption hardware-based artificial neural networks capable of interacting with body fluids and biological tissues.

Monitoring Indoor People Presence in Buildings Using Low-Cost Infrared Sensor Array in Doorways.

We propose a device for monitoring the number of people who are physically present inside indoor environments. The device performs local processing of infrared array sensor data detecting people’s direction, which allows monitoring users’ occupancy in any space of the building and also respects people privacy. The device implements a novel real-time pattern recognition algorithm for processing data sensed by a low-cost infrared (IR) array sensor. The computed information is transferred through a Z-Wave network. On-field evaluation of the algorithm has been conducted by placing the device on top of doorways in offices and laboratory rooms. To evaluate the performance of the algorithm in varying ambient temperatures, two groups of stress tests have been designed and performed. These tests established the detection limits linked to the difference between the average ambient temperature and perturbation. For an in-depth analysis of the accuracy of the algorithm, synthetic data have been generated considering temperature ranges typical of a residential environment, different human walking speeds (normal, brisk, running), and distance between the person and the sensor (1.5 m, 5 m, 7.5 m). The algorithm performed with high accuracy for routine human passage detection through a doorway, considering indoor ambient conditions of 21–30 °C.

Anisotropic photoresponse of layered rhenium disulfide synaptic transistors

Layered ReS2 with direct bandgap and strong in-plane anisotropy shows great potential to develop high-performance angle-resolved photodetectors and optoelectronic devices. However, systematic characterizations of the angle-dependent photoresponse of ReS2 are still very limited. Here, we studied the anisotropic photoresponse of layered ReS2 phototransistors in depth. Angel-resolved Raman spectrum and field-effect mobility are tested to confirm the inconsistency between its electrical and optical anisotropies, which are along 120° and 90°, respectively. We further measured the angle-resolved photoresponse of a ReS2 transistor with 6 diagonally paired electrodes. The maximum photoresponsivity exceeds 0.515 A⋅W−1 along b-axis, which is around 3.8 times larger than that along the direction perpendicular to b axis, which is consistent with the optical anisotropic directions. The incident wavelength- and power-dependent photoresponse measurement along two anisotropic axes further demonstrates that b axis has stronger light–ReS2 interaction, which explains the anisotropic photoresponse. We also observed angle-dependent photoresistive switching behavior of the ReS2 transistor, which leads to the formation of angle-resolved phototransistor memory. It has simplified structure to create dynamic optoelectronic resistive random access memory controlled spatially through polarized light. This capability has great potential for real-time pattern recognition and photoconfiguration of artificial neural networks (ANN) in a wide spectral range of sensitivity provided by polarized light.

Image segmentation using multilevel thresholding based on type II fuzzy entropy and marine predators algorithm

The digital image segmentation is an open problem that is growing day by day and is attracting the attention of researchers from last few years. Image resolution and their speed has led to the use of thresholding approaches. Image thresholding is simple, easy and effective method for image segmentation. Multi-level image thresholding is a key perspective in several real-time pattern recognition and image processing-based applications. It identifies pixels quickly and effectively in different groups indicating multiple regions in an image. Segmentation of images based on thresholding by using various intelligent optimization techniques with fuzzy entropy is widely utilized for defining thresholds in a better way to use them precisely. In this research, a novel technique for multi-level thresholding is proposed by combining Fuzzy Entropy Type II (FE-TII) with recently developed meta-heuristics named Marine Predators Algorithm (MPA). For achieving optimal thresholds of an image, the maximization of entropy is tedious and consumes a lot of time with an increasing number of thresholds. The MPA method presented is analyzed in context with image segmentation, particularly on thresholds with TII-FE. For this reason, proposed methodology is evaluated using several images along with the distribution of histograms. For analyzing the performance efficiency of the proposed methodology, the results are compared and robustness is tested with efficiency of proposed technique to multi-level image segmentation, several images are used randomly from datasets.

Real-Time Pattern-Recognition of GPR Images with YOLO v3 Implemented by Tensorflow.

Artificial intelligence (AI) is widely used in pattern recognition and positioning. In most of the geological exploration applications, it needs to locate and identify underground objects according to electromagnetic wave characteristics from the ground-penetrating radar (GPR) images. Currently, a few robust AI approach can detect targets by real-time with high precision or automation for GPR images recognition. This paper proposes an approach that can be used to identify parabolic targets with different sizes and underground soil or concrete structure voids based on you only look once (YOLO) v3. With the TensorFlow 1.13.0 developed by Google, we construct YOLO v3 neural network to realize real-time pattern recognition of GPR images. We propose the specific coding method for the GPR image samples in Yolo V3 to improve the prediction accuracy of bounding boxes. At the same time, K-means algorithm is also applied to select anchor boxes to improve the accuracy of positioning hyperbolic vertex. For some instances electromagnetic-vacillated signals may occur, which refers to multiple parabolic electromagnetic waves formed by strong conductive objects among soils or overlapping waveforms. This paper deals with the vacillating signal similarity intersection over union (IoU) (V-IoU) methods. Experimental result shows that the V-IoU combined with non-maximum suppression (NMS) can accurately frame targets in GPR image and reduce the misidentified boxes as well. Compared with the single shot multi-box detector (SSD), YOLO v2, and Faster-RCNN, the V-IoU YOLO v3 shows its superior performance even when implemented by CPU. It can meet the real-time output requirements by an average 12 fps detected speed. In summary, this paper proposes a simple and high-precision real-time pattern recognition method for GPR imagery, and promoted the application of artificial intelligence or deep learning in the field of the geophysical science.

Digital Hardware Spiking Neuronal Network with STDP for Real-time Pattern Recognition

By mimicking or being inspired by the nervous system, neuromorphic systems are designed to realize robust and power-efficient information processing by highly parallel architecture. Spike Timing Dependent Plasticity (STDP) is a common learning method for Spiking Neural Networks (SNNs). Here, we present a real-time SNN with STDP implementation on Field Programmable Gate Array (FPGA) using digital spiking silicon neuron model. Equipped with Ethernet Interface, FPGA allows online configuration as well as real-time processing data input and output. We show that this hardware implementation can achieve real-time pattern recognition tasks and allows the connection between multi-SNNs to extend the scale of networks.

Understanding Limb Position and External Load Effects on Real-Time Pattern Recognition Control in Amputees.

Limb position is a factor that negatively affects myoelectric pattern recognition classification accuracy. However, prior studies evaluating impact on real-time control for upper-limb amputees have done so without a physical prosthesis on the residual limb. It remains unclear how limb position affects real-time pattern recognition control in amputees when their residual limb is supporting various weights. We used a virtual reality target achievement control test to evaluate the effects of limb position and external load on real-time pattern recognition control in fourteen intact limb subjects and six major upper limb amputee subjects. We also investigated how these effects changed based on different control system training methods. In a static training method, subjects kept their unloaded arm by their side with the elbow bent whereas in the dynamic training method, subjects moved their arm throughout a workspace while supporting a load. When static training was used, limb position significantly affected real-time control in all subjects. However, amputee subjects were still able to adequately complete tasks in all conditions, even in untrained limb positions. Moreover, increasing external loads decreased controller performance, albeit to a lesser extent in amputee subjects. The effects of limb position did not change as load increased, and vice versa. In intact limb subjects, dynamic training significantly reduced the limb position effect but did not completely remove them. In contrast, in amputee subjects, dynamic training eliminated the limb position effect in three out of four outcome measures. However, it did not reduce the effects of load for either subject population. These findings suggest that results obtained from intact limb subjects may not generalize to amputee subjects and that advanced training methods can substantially improve controller robustness to different limb positions regardless of limb loading.

Committee classifier based on linear discriminant analysis for the detection of radioisotopes from airborne gamma-ray spectra

A committee classifier was developed for use in the application of real-time pattern recognition to gamma-ray spectra collected from airborne surveys. This technique was designed to enhance detection performance relative to that of a single linear discriminant analysis model. The approach was based on utilizing multiple classifiers to check one another through a signal averaging method. This resulted in an ability to reject random false detections while maximizing detection sensitivity. Making use of spectral preprocessing algorithms previously studied, the committee classifiers were applied to the detection of cesium-137 and cobalt-60 in spectra collected in the field during airborne surveys. Applying a z-score methodology to the classification scores allowed classifiers developed with different processing parameters to operate in the same dataspace for the purpose of classifying the target spectra. The optimized classifiers were tested over 13 diverse locations, with nine of the sites containing the respective target isotopes. Results of the committee classifiers indicated an improvement in missed and false detection performance for both radioisotopes. In addition, work was performed to confirm that several suspected false detections were actually weak target signals only visible once co-added with other similar spectra. This result suggested the committee classifier performance may have exceeded the capabilities of the visual spectral inspection on which the performance statistics were based.

Addressing Challenges in Rig-Based Drilling Advisory System Deployment

This article, written by JPT Technology Editor Judy Feder, contains highlights of paper SPE 194184, “Change-Management Challenges Deploying Drilling Advisory System,” by Michael Behounek, SPE, Blake Millican, and Brian Nelson, Apache, et al., prepared for the 2019 SPE/IADC Drilling International Conference and Exhibition, The Hague, 5–7 March. The paper has not been peer reviewed. Sophisticated drilling-analysis software can help drillers set and modify weight on bit (WOB), rev/min, and other drilling parameters, but achieving acceptance of these software-based recommendations by a driller is complicated. Additionally, acceptance of changes to drilling techniques and modified work flows by a driller on one test rig is insufficient. The challenge is to scale buy-in across a mixed rig contractor fleet. Many projects fail when the change-management process is not properly executed. The complete paper presents a process used to successfully implement a rig-based drilling advisory system (RDAS) across a mixed group of rig contractors. The paper documents the direction and effort taken to implement change from the rig site to the office, along with management support. Change-Management Process Many drilling performance projects fail, not because of technology, but because of factors such as the lack of consideration given to how the project will be perceived by those it affects (stakeholders), and the lack of thought given to the project’s scalability. To address these issues, change-management principles must be applied consciously. The list of stakeholders may include drillers, wellsite supervisors, drilling contractors, drilling engineers, managers, and others. Once this list has been compiled, mapping of the stakeholders on a grid that gauges their influence and interest is beneficial. Stakeholder management involves determining how best to engage individual stakeholders to move them to the high-interest side of the grid. High levels of interest and engagement from stakeholders lead to good feedback, which, if incorporated properly into the system, leads to even higher interest and subsequent permanent adoption of the technology. The best way to ensure permanent change is to ensure that each stakeholder has benefitted from the project in some way, a goal accomplished through continuous monitoring and tracking of stake-holder interest level. RDAS Overview To address shortcomings in existing industry approaches to using both low- and high-frequency data to improve drilling performance, and to achieve a scalable, economic drilling advisory system, Apache developed an RDAS that is agnostic to any high-frequency data stream—surface or downhole—including those from rig-control systems. The RDAS consists of data-acquisition hardware and scalable off-the-shelf software, with an open API layer and a plug-and-play backend software that runs both physics- and data-based analytical models at the rig site on reasonable computing power. The RDAS displays new advisory information in the driller’s cabin, running real-time pattern recognition algorithms to detect drilling dysfunctions. When a drilling dysfunction is encountered, a change in drilling parameters is suggested. Additionally, drilling parameters from offset wells are made available automatically for the driller’s use on the drilling screen. Through this process, the operator entrusts field personnel with a slightly higher level of technical responsibility. The team has improved the system iteratively using feedback from drillers who used the RDAS.

Real-time pattern recognition implementation on FPGA in multi-SNNs

By mimicking or being inspired by the nervous system, Neuromorphic systems are designed to realize robust and power-efficient information processing by highly parallel architecture. Spike timing dependent plasticity (STDP) is a common method for training Spiking Neural Networks (SNNs) for pattern recognition. Here, we present a real-time STDP implementation on FPGA in SNN using digital spiking silicon neuron (DSSN) model. Equipped with Ethernet Interface, FPGA allows online configuration as well as data input and output all in real-time. We show that this STDP implementation can achieve pattern recognition task and the connection between multi-SNNs enlarge the scale of networks and application.

Modification of the WaldBoost algorithm to improve the efficiency of solving pattern recognition problems in real-time

The implementation of the WaldBoost algorithm is considered, and its modification is proposed, which allows to significantly reduce the number of weak classifiers to achieve a given classification accuracy. The efficiency of the proposed algorithm is shown by specific examples. The paper studies modifications of compositions (ensembles) of algorithms for solving real-time pattern recognition problems. The aim of the study is to improve the known machine learning algorithms for pattern recognition using a minimum amount of time (the minimum number of used classifiers) and with a given accuracy of the results. We consider the implementation of the WaldBoost algorithm, which combines two algorithms: adaptive boosting of weak classifiers – AdaBoost (adaptive boosting), which has a high generalizing ability, and the sequential probability ratio test – SPRT (Wald test), which is the optimal rule of decision-making when distinguishing two hypotheses. It is noted that when using the WaldBoost, the values of the actual probability of classification errors, as a rule, are less than given because of the approximate boundaries of the SPRT, so that the classification process uses an excessive series of weak classifiers. In this regard, we propose a modification of the WaldBoost based on iterative refinement of the decision boundaries, which can significantly reduce the number of used weak classifiers required for pattern recognition with a given accuracy. The efficiency of the proposed algorithm is shown by specific examples. The results are confirmed by statistical modeling on several data sets. It is noted that the results can be applied in the refinement of other cascade classification algorithms.