Real-time Extraction Research Articles

Onshore versus offshore wind power trends and recent study practices in modeling of wind turbines’ life-cycle impact assessments

Wind energy has gradually become one of the most efficient and reliable forms of sustainable energy, and it is being globally utilized to accelerate the expansion of green power production. However, recent projections of onshore & offshore wind energy systems indicate that further improvements are continuously required in terms of the deployments, capacity factors, costs, etc. to achieve an optimal wind power generation. According to these projections, the most critical current issue is the offshore systems' high levelized cost of energy, which demands the enhancements of the technologies to ensure the cost reduction in the future. Moreover, a number of reports are underscoring that offshore power systems can also leave a larger carbon footprints through their life-cycles compared to the onshore counterpart. On the other hand, the modelling methods of the power plants' life-cycle impact assessment (LCIA) have nowadays become among the most compelling research problems in the field of wind power systems engineering. In addition to comparatively assessing the characteristics of onshore & offshore energy systems based on some reports & studies, this work is therefore aimed at rigorously exploring the existing study practices in modelling the wind power plants' LCIA. A number of these practices were based on the implementation of various conventional modelling methods that may not help to make accurate assessments of the wind power plants’ life-cycle impacts. Hence, one of the main contributions of this study is to summarize a novel LCIA modelling approach based on Internet of Things (IoT) and predictive digital twin technologies. Unlike the conventional methods, real-time data extraction from the power plants makes these technologies possible to conduct the most accurate assessments of the life-cycle impacts.

Real-Time Feature Extraction from Electrocochleography with Impedance Measurements during Cochlear Implantation using Linear State-Space Models.

Electrocochleography (ECochG) is increasingly used to monitor the inner ear function of cochlear implant (CI) patients during surgery. Current ECochG-based trauma detection shows low sensitivity and specificity and depends on visual analysis by experts. Trauma detection could be improved by including electric impedance data recorded simultaneously with the ECochG. However, combined recordings are rarely used because the impedance measurements produce artifacts in the ECochG. In this study, we propose a framework for automated real-time analysis of intraoperative ECochG signals using Autonomous Linear State-Space Models (ALSSMs). We developed ALSSM based algorithms for noise reduction, artifact removal, and feature extraction in ECochG. Feature extraction includes local amplitude and phase estimations and a confidence metric over the presence of a physiological response in a recording. We tested the algorithms in a controlled sensitivity analysis using simulations and validated them with real patient data recorded during surgeries. The results from simulation data show that the ALSSM method provides improved accuracy in the amplitude estimation together with a more robust confidence metric of ECochG signals compared to the state-of-the-art methods based on the fast Fourier transform (FFT). Tests with patient data showed promising clinical applicability and consistency with the findings from the simulations. We showed that ALSSMs are a valid tool for real-time analysis of ECochG recordings. Removal of artifacts using ALSSMs enables simultaneous recording of ECochG and impedance data. The proposed feature extraction method provides the means to automate the assessment of ECochG. Further validation of the algorithms in clinical data is needed.

Vehicle Trajectory Data Mining for Artificial Intelligence and Real-Time Traffic Information Extraction

It aims to improve the efficiency of information collection and extraction in the current intelligent transportation system, and accurately mine the vehicle trajectory data By using Artificial Intelligence (AI) and Deep Learning methods, the trajectory data generated during vehicle driving are deeply mined and analyzed, and the characteristics of driving behavior of vehicle drivers are modeled and analyzed in detail. Then, a method of mining driving behavior characteristics based on Convolutional Neural Network (CNN) and vehicle trajectory is proposed. Based on the mathematical principle of wavelet packet and Least Square Support Vector Machine (LSSVM), a combined model of trajectory mining is constructed and applied to the short-term prediction of traffic flow. The traffic flow of Binjiang Road and Renmin Road in Guangzhou, Guangdong Province from August 19 to August 21, 2021 is predicted to verify the accuracy of the trajectory mining combined model. The results show that the combination model of data mining has good fitting effect, and the average accuracy is above 0.8. Besides, the effectiveness of the Deep Learning model in driver behavior classification is verified. The accuracy of the classification model is 75.2% for trajectory, and that is 76.8% for driver behavior characteristics. It is of great significance to effectively utilize the knowledge data in Intelligent Transportation System (ITS) and extract valuable information from it, which has certain reference value for the subsequent refined prediction of vehicle behavior.

Nanosensor Chemical Cytometry: Advances and Opportunities in Cellular Therapy and Precision Medicine.

With the definition of therapeutics now encompassing transplanted or engineered cells and their molecular products, there is a growing scientific necessity for analytics to understand this new category of drugs. This Perspective highlights the recent development of new measurement science on label-free single cell analysis, nanosensor chemical cytometry (NCC), and their potential for cellular therapeutics and precision medicine. NCC is based on microfluidics integrated with fluorescent nanosensor arrays utilizing the optical lensing effect of a single cell to real-time extract molecular properties and correlate them with physical attributes of single cells. This new class of cytometry can quantify the heterogeneity of the multivariate physicochemical attributes of the cell populations in a completely label-free and nondestructive way and, thus, suggest the vein-to-vein conditions for the safe therapeutic applications. After the introduction of the NCC technology, we suggest the technological development roadmap for the maturation of the new field: from the sensor/chip design perspective to the system/software development level based on hardware automation and deep learning data analytics. The advancement of this new single cell sensing technology is anticipated to aid rich and multivariate single cell data setting and support safe and reliable cellular therapeutics. This new measurement science can lead to data-driven personalized precision medicine.

Non-invasive human thermal comfort assessment based on multiple angle/distance facial key-region temperatures recognition

Accurate monitoring of human thermal comfort is crucial in optimizing HVAC system control scheme and enhancing building energy efficiency. The method based on facial temperature infrared recognition is a non-invasive and real-time thermal comfort predictive strategy, which has been proven great application potential. In response to the limitations of previous fixed-position facial infrared detection in practical applications, this study aims to explore a less restrictive non-intrusive method for assessing human thermal comfort. The temperature features of multiple regions on human face are identified by an infrared camera, and six temperature-sensitive regions are selected through skin thermal sensitivity analysis. Thirty subjects participate in a series of field experiments. To reduce the measurement limitations caused by human postures, the correlation between these six facial regions, indoor air temperature, thermal sensation votes (TSV), and thermal comfort votes (TCV) are analyzed in details to determine two key facial regions for the non-intrusive thermal comfort assessment. By applying the YOLOv5 algorithm, the real-time extraction of facial region temperatures from multiple angles and distances is achieved. On this basis, a data-driven thermal comfort predictive strategy based on facial temperature and optimized SVM model is designed. Results indicate that the method achieves a general accuracy of 85.68% in predicting thermal comfort based on its successful recognition rate of 88.7% for facial key regions.

An 8-Channel Ambulatory EEG Recording IC With In-Channel Fully-Analog Real-Time Motion Artifact Extraction and Removal.

Wereport the design, implementation, and experimental characterization of an 8-channel EEG recording IC (0.13 μm CMOS, 12 mm 2 total area) with a channel architecture that conducts both the extraction and removal of motion artifacts on-chip and in-channel. The proposed dual-path feed-forward method for artifact extraction and removal is implemented in the analog domain, hence is needless of a DSP unit for artifact estimation, and its associated high-DR ADCs and DACs employed by the state of the art for artifact replica generation. Additionally, the presented architecture improves system's scalability as it enables channels' stand-alone operation, and yields the lowest reported channel power consumption among works featuring motion artifact detection/removal. Following an experimental study on electrode-skin interface electrical characteristics for dry electrodes in the absence and presence of motions, the article presents the channel architecture, its detailed signal transfer function analysis, circuit-level implementation, and experimental characterization results. Our measurement results show an amplification voltage gain of 48.3 dB, a bandwidth of 300 Hz, rail-to-rail input DC offset tolerance, and 41.5 dB artifact suppression, while consuming 55 μW per channel. The system's efficacy in EEG motion artifact suppression is validated experimentally, and system- and circuit-level features and performance metrics of the presented design are compared with the state of the art.

Comparison of filtering methods for real-time extraction of the volitional EMG component in electrically stimulated muscles

ObjectiveRecorded electromyograms (EMG) of electrically stimulated muscles can contain both an exogenous-evoked potential (M-wave) and an endogenous, or volitional, component. This study evaluated the effectiveness of three filtering methods (i.e., high-pass, adaptive, and comb), commonly used in neurorehabilitation, in extracting the volitional component of simulated and experimental EMG during upper-limb tasks. MethodsVolitional EMG and M-wave were simulated through a physiological model of muscle recruitment, comprising of a motor neuron pool and associated muscle fibres, superimposed to a stimulation artefact. Experimental EMG data during different levels of volitional muscle contraction in isometric and dynamic tasks were recorded from five unimpaired individuals. Electrical stimulation artefact was removed with different techniques (i.e., none, removing samples, blanking, and interpolation) to assess filter performance across time and frequency domains, and information content (i.e., Kolmogorov-Smirnov D-value). ResultsThe experimental results agreed with the simulations, wherein the adaptive filter outperformed the other filters when using no artefact removal or removing artefact samples from the signal, while for the blanking and interpolation artefact removal methods, the adaptive and comb filters outperformed the high-pass filter. ConclusionThe adaptive and comb filters best estimated volitional muscle activity in electrically stimulated muscles. SignificanceResults from this study will enable the enhanced design of real-time neuroprosthesis control.

LRE-Net: A Vision-Based real-time rail extraction method under Metro tunnel low-light scene

Vision-based methods for detecting the dynamic clearance intrusion of trains in underground tunnels encounter challenges in extracting rails from low-contrast images caused by low-light scenes. This study introduces a solution for Rail Extraction in the Low-light scenes of underground tunnels (LRE-Net). Exposure features were integrated into the backbone to enhance the capability of the model in capturing global features in low-light scenes. Additionally, this study proxies the track extraction as a row-grid-based classification formulation. Two auxiliary tasks were introduced as regularization terms for generic feature learning during the training phase to address the potential overfitting resulting from the sparse nature of the track-extraction task. Experiments were conducted on a dataset consisting of 7,643 images, demonstrating that LRE-Net achieved the highest extraction accuracy of 93.94% at 88 frames per second (FPS). Furthermore, the impact of noise in low-light images on track extraction was analyzed, and strategies were proposed to enhance accuracy.

Drivable Road Region Segmentation in Real Time with High Precision using Deep Learning

Abstract: This paper presents a novel approach that addresses the challenging task of real-time drivable road region extraction in computer vision. Semantic segmentation, which involves accurately identifying and segmenting objects in real-time data, is a complex problem. However, deep learning has proven to be a powerful technique for achieving semantic segmentation by automatically identifying patterns without the need for explicit programming. To tackle this task, the paper proposes a fusion of the YOLO algorithm and UNET architecture, leveraging their respective strengths. The YOLO algorithm enables high-speed object detection, while the UNET architecture provides advantages in global location utilization, contextual understanding, and performance, even with limited training samples. Importantly, the proposed method is lightweight, making it suitable for deployment on embedded systems with limited computational power. To optimize memory usage and capture context at different scales, the system employs dilated convolutions for efficient feature extraction. The algorithm exhibits exceptional performance in accurately segmenting irregular objects and handles diverse input data types, including images and videos, in real-time. Overall, this paper contributes significantly to the advancement of computer vision technologies and offers a valuable solution for real-time drivable road region extraction. Its potential applications include addressing driving challenges and enhancing safety in autonomous vehicles and intelligent transportation systems.

Migratory Perception in Edge-Assisted Internet of Vehicles

Autonomous driving technology heavily relies on the accurate perception of traffic environments, mainly through roadside cameras and LiDARs. Although several popular and robust 2D and 3D object detection methods exist, including R-CNN, YOLO, SSD, PointPillar, and VoxelNet, the perception range and accuracy of an individual vehicle can be limited by blocking from other vehicles or buildings. A solution is to harness roadside perception infrastructures for vehicle–infrastructure cooperative perception, using edge computing for real-time intermediate features extraction and V2X networks for transmitting these features to vehicles. This emerging migratory perception paradigm requires deploying exclusive cooperative perception services on edge servers and involves the migration of perception services to reduce response time. In such a setup, competition among multiple cooperative perception services exists due to limited edge resources. This study proposes a multi-agent reinforcement learning (MADRL)-based service scheduling method for migratory perception in vehicle–infrastructure cooperative perception, utilizing a discrete time-varying graph to model the relationship between service nodes and edge server nodes. This MADRL-based approach can efficiently address the challenges of service placement and migration in resource-limited environments, minimize latency, and maximize resource utilization for migratory perception services on edge servers.

Advanced Predictive Tile Selection Using Dynamic Tiling for Prioritized 360° Video VR Streaming

The widespread availability of smart computing and display devices such as mobile phones, gaming consoles, laptops, and tethered/untethered head-mounted displays has fueled an increase in demand for omnidirectional (360°) videos. 360° video applications enable users to change their viewing angles while interacting with the video during playback. This allows users to have a more personalized and interactive viewing experience. Unfortunately, these applications require substantial network and computational resources that the conventional infrastructure and end devices cannot support. Recently proposed viewport adaptive fixed tiling solutions stream only relevant video tiles based on user interaction with the virtual reality (VR) space to use existing transmission resources more efficiently. However, achieving real-time accurate viewport extraction and transmission in response to both head movements and bandwidth dynamics can be challenging, which can impact the user’s Quality of Experience (QoE). This article proposes innovative dynamic tiling-based adaptive 360° video streaming solutions in order to achieve high viewer QoE. First, novel and easy-to-scale tiling layout selection methods are introduced, and the best tiling layouts are employed in each adaptation interval based on the prediction-assisted visual quality metric and the observed viewport divergence. Second, a novel proactive tile selection approach is presented, which adaptively extracts tiles for each selected tiling layout based on two low-complex viewport prediction mechanisms. Finally, a practical dynamic tile priority-oriented bitrate adaptation scheme is introduced, which uniformly distributes the bitrate budget among different tiles during 360° video streaming. Extensive trace-driven experiments are conducted to evaluate the proposed solutions using head motion traces from 48 VR users for five 360° videos with tiling layouts of 4 × 3, 6 × 4, and 8 × 6 and segment durations of 1s, 1.5s, and 2s. The experimental evaluations show that the dynamic video tiling solutions achieve up to 11.2% more viewport matches and an average improvement in QoE of 9.7% to 18% compared to state-of-the-art 360° streaming approaches.

FEBench: A Benchmark for Real-Time Relational Data Feature Extraction

As the use of online AI inference services rapidly expands in various applications (e.g., fraud detection in banking, product recommendation in e-commerce), real-time feature extraction (RTFE) systems have been developed to compute the requested features from incoming data tuples in ultra-low latency. Similar to relational databases, these RTFE procedures can be expressed using SQL-like languages. However, there is a lack of research on the workload characteristics and specialized benchmarks for RTFE, especially in comparison with existing database workloads and benchmarks (e.g., concurrent transactions in TPC-C). In this paper, we study the RTFE workload characteristics using over one hundred real datasets from open repositories (e.g. Kaggle, Tianchi, UCI ML, KiltHub) and those from 4Paradigm. The study highlights the significant differences between RTFE workloads and existing database benchmarks in terms of application scenarios, operator distributions, and query structures. Based on these findings, we propose to develop a realtime feature extraction benchmark named FEBench based on the four important criteria for a domain-specific benchmark proposed by Jim Gray. FEBench consists of selected representative datasets, query templates, and an online request simulator. We use FEBench to evaluate the effectiveness of feature extraction systems including OpenMLDB and Flink and find that each system exhibits distinct advantages and limitations in terms of overall latency, tail latency, and concurrency performance.

Real-Time field road freespace extraction for agricultural machinery autonomous driving based on LiDAR

Efficient road freespace (RFS) extraction is an important element in autonomous driving. In this article, a method based on line segments is proposed for the extraction RFS of low curvature (maximum curvature of the boundary of 27). The process is made up of three parts. First, the multisource data were fused to correct the distorted point cloud. Second, we extracted the road surface point cloud within the region of interest (ROI) and generated road boundary lines. Third, the obstacles on the road were segmented in accordance with the spatial association features of the line segments, followed by the elimination of pseudo obstacles. Finally, the road boundary, the road surface point cloud, the pseudo obstacles point cloud, and the obstruct obstacles point cloud were merged to obtain and visualize the RFS. The experimental results showed that the TPR, FPR and accuracy of the proposed method with Velodyne VLP-16 LiDAR point cloud data were 98.54%, 9.27% and 97.57% for semistructured roads and 97.78%, 4.57% and 97.55% for unstructured roads, respectively. The mean processing speeds for a single frame were 57 ms/frame for semistructured roads and 61 ms/frame for unstructured roads. The RFS acquired in this article can provide the fundamental outcomes of perception for dynamic planning and decision-making of agricultural machinery autonomous driving.

Weak Spatial Target Extraction Based on Small-Field Optical System.

Compared to wide-field telescopes, small-field detection systems have higher spatial resolution, resulting in stronger detection capabilities and higher positioning accuracy. When detecting by small fields in synchronous orbit, both space debris and fixed stars are imaged as point targets, making it difficult to distinguish them. In addition, with the improvement in detection capabilities, the number of stars in the background rapidly increases, which puts higher requirements on recognition algorithms. Therefore, star detection is indispensable for identifying and locating space debris in complex backgrounds. To address these difficulties, this paper proposes a real-time star extraction method based on adaptive filtering and multi-frame projection. We use bad point repair and background suppression algorithms to preprocess star images. Afterwards, we analyze and enhance the target signal-to-noise ratio (SNR). Then, we use multi-frame projection to fuse information. Subsequently, adaptive filtering, adaptive morphology, and adaptive median filtering algorithms are proposed to detect trajectories. Finally, the projection is released to locate the target. Our recognition algorithm has been verified by real star images, and the images were captured using small-field telescopes. The experimental results demonstrate the effectiveness of the algorithm proposed in this paper. We successfully extracted hip-27066 star, which has a magnitude of about 12 and an SNR of about 1.5. Compared with existing methods, our algorithm has advantages in both recognition rate and false-alarm rate, and can be used as a real-time target recognition algorithm for space-based synchronous orbit detection payloads.

Current status and operation of the H.E.S.S. array of imaging atmospheric Cherenkov telescopes

The High Energy Stereoscopic System (H.E.S.S.) is an array of five imaging atmospheric Cherenkov telescopes (IACTs) to study gamma-ray emission from astrophysical objects in the Southern hemisphere. It is the only hybrid array of IACTs, composed of telescopes with different collection area and footprint, individually optimised for a specific energy range. Collectively, the array is most sensitive to gamma rays in the range of 100GeV to 100TeV. The array has been in operation since 2002 and has been upgraded with new telescopes and cameras multiple times. Recent hardware upgrades and changes in the operational procedures increased the amount of observing time, which is of key importance for time-domain science. H.E.S.S. operations saw record data taking in 2020 and 2021 and we describe the current operations with specific emphasis on system performance, operational processes and workflows, quality control, and (near) real-time extraction of science results. In light of this, we will briefly discuss the early detection of gamma-ray emission from the recurrent nova RS Oph and alert distribution to the astrophysics community.

A satellite component contour extraction method for lightweight space mobile platforms

PurposeOwing to the complex space environment and limited computing resources, traditional and deep learning-based methods cannot complete the task of satellite component contour extraction effectively. To this end, this paper aims to propose a high-quality real-time contour extraction method based on lightweight space mobile platforms.Design/methodology/approachA contour extraction method that combines two edge clues is proposed. First, Canny algorithm is improved to extract preliminary contours without inner edges from the depth images. Subsequently, a new type of edge pixel feature is designed based on surface normal. Finally, surface normal edges are extracted to supplement the integrity of the preliminary contours for contour extraction.FindingsExtensive experiments show that this method can achieve a performance comparable to that of deep learning-based methods and can achieve a 36.5 FPS running rate on mobile processors. In addition, it exhibits better robustness under complex scenes.Practical implicationsThe proposed method is expected to promote the deployment process of satellite component contour extraction tasks on lightweight space mobile platforms.Originality/valueA pixel feature for edge detection is designed and combined with the improved Canny algorithm to achieve satellite component contour extraction. This study provides a new research idea for contour extraction and instance segmentation research.

Nanopore Filter: A Method for Counting and Extracting Single DNA Molecules Using a Biological Nanopore.

This paper describes a method for the real-time counting and extraction of DNA molecules at the single-molecule level by nanopore technology. As a powerful tool for electrochemical single-molecule detection, nanopore technology eliminates the need for labeling or partitioning sample solutions at the femtoliter level. Here, we attempt to develop a DNA filtering system utilizing an α-hemolysin (αHL) nanopore. This system comprises two droplets, one filling with and one emptying DNA molecules, separated by a planar lipid bilayer containing αHL nanopores. The translocation of DNA through the nanopores is observed by measuring the channel current, and the number of translocated molecules can also be verified by quantitative polymerase chain reaction (qPCR). However, we found that the issue of contamination seems to be an almost insolvable problem in single-molecule counting. To tackle this problem, we tried to optimize the experimental environment, reduce the volume of solution containing the target molecule, and use the PCR clamp method. Although further efforts are still needed to achieve a single-molecule filter with electrical counting, our proposed method shows a linear relationship between the electrical counting and qPCR estimation of the number of DNA molecules.

Data-Driven-Based Terahertz Image Restoration

Terahertz (THz) has been widely used in radar, remote sensing, atmospheric and environmental monitoring, real-time biological information extraction, and medical diagnosis. However, the long-wavelength property of THz leads to low imaging resolution. Due to the expensive nature of experimental equipment and the difficulty of obtaining terahertz sources, most scientific institutions have difficulty in obtaining large amounts of terahertz dataset. In addition, it is difficult for convolution neural networks to recover low-resolution complex images because of the phase information contained in the images. In this paper, we build a terahertz imaging system and propose a Progressive Multi-scale Augmented Neural Network (PMANN) to address this problem. PMANN integrates self-attention and channel-attention for image contextual feature extraction by the proposed progressive network. Compared with the Complex convolution Neural Network (CCNN) in private cartoon datasets, PMANN improves the quantitative metrics on PSNR by about 18.6%. It also obtains better visual quality than CCNN. Moreover, we publicly and synthesize the first grayscale image MinistTHz dataset. Compared to SwinTransformer, PMANN improves the PSNR by 28.2%. The experimental results on the THz dataset demonstrate that the proposed model achieves better performances than the state-of-the-art models. Our synthetic THz dataset is open at https://github.com/szpxmu/PMANN.

Hardware architecture design for real-time SIFT extraction with reduced memory usage

Hardware architecture design for real-time SIFT extraction with reduced memory usage

Hyper-parameter tuning and feature extraction for asynchronous action detection from sub-thalamic nucleus local field potentials.

Decoding brain states from subcortical local field potentials (LFPs) indicative of activities such as voluntary movement, tremor, or sleep stages, holds significant potential in treating neurodegenerative disorders and offers new paradigms in brain-computer interface (BCI). Identified states can serve as control signals in coupled human-machine systems, e.g., to regulate deep brain stimulation (DBS) therapy or control prosthetic limbs. However, the behavior, performance, and efficiency of LFP decoders depend on an array of design and calibration settings encapsulated into a single set of hyper-parameters. Although methods exist to tune hyper-parameters automatically, decoders are typically found through exhaustive trial-and-error, manual search, and intuitive experience. This study introduces a Bayesian optimization (BO) approach to hyper-parameter tuning, applicable through feature extraction, channel selection, classification, and stage transition stages of the entire decoding pipeline. The optimization method is compared with five real-time feature extraction methods paired with four classifiers to decode voluntary movement asynchronously based on LFPs recorded with DBS electrodes implanted in the subthalamic nucleus of Parkinson's disease patients. Detection performance, measured as the geometric mean between classifier specificity and sensitivity, is automatically optimized. BO demonstrates improved decoding performance from initial parameter setting across all methods. The best decoders achieve a maximum performance of 0.74 ± 0.06 (mean ± SD across all participants) sensitivity-specificity geometric mean. In addition, parameter relevance is determined using the BO surrogate models. Hyper-parameters tend to be sub-optimally fixed across different users rather than individually adjusted or even specifically set for a decoding task. The relevance of each parameter to the optimization problem and comparisons between algorithms can also be difficult to track with the evolution of the decoding problem. We believe that the proposed decoding pipeline and BO approach is a promising solution to such challenges surrounding hyper-parameter tuning and that the study's findings can inform future design iterations of neural decoders for adaptive DBS and BCI.