Automatic Identification Algorithm Research Articles

The Partnership of Citizen Science and Machine Learning: Benefits, Risks, and Future Challenges for Engagement, Data Collection, and Data Quality

Advances in artificial intelligence (AI) and the extension of citizen science to various scientific areas, as well as the generation of big citizen science data, are resulting in AI and citizen science being good partners, and their combination benefits both fields. The integration of AI and citizen science has mostly been used in biodiversity projects, with the primary focus on using citizen science data to train machine learning (ML) algorithms for automatic species identification. In this article, we will look at how ML techniques can be used in citizen science and how they can influence volunteer engagement, data collection, and data validation. We reviewed several use cases from various domains and categorized them according to the ML technique used and the impact of ML on citizen science in each project. Furthermore, the benefits and risks of integrating ML in citizen science are explored, and some recommendations are provided on how to enhance the benefits while mitigating the risks of this integration. Finally, because this integration is still in its early phases, we have proposed some potential ideas and challenges that can be implemented in the future to leverage the power of the combination of citizen science and AI, with the key emphasis being on citizen science in this article.

Automatic Identification Algorithm of Equivalent Electrochemical Circuit Based on Electroscopic Impedance Data for a Lead Acid Battery

Obtaining tools to analyze and predict the performance of batteries is a non-trivial challenge because it involves non-destructive evaluation procedures. At the research level, the development of sensors to allow cell-level monitoring is an innovative path, and electrochemical impedance spectrometry (EIS) has been identified as one of the most promising tools, as is the generation of advanced multivariable models that integrate environmental and internal-battery information. In this article, we describe an algorithm that automatically identifies a battery-equivalent electrochemical model based on electroscopic impedance data. This algorithm allows in operando monitoring of variations in the equivalent circuit parameters that will be used to further estimate variations in the state of health (SoH) and state of charge (SoC) of the battery based on a correlation with experimental aging data corresponding to states of failure or degradation. In the current work, the authors propose a two-step parameter identification algorithm. The first consists of a rough differential evolution algorithm-based identification. The second is based on the Nelder–Mead Simplex search method, which gives a fine parameter estimation. These algorithm results were compared with those of the commercially available Z-view, an equivalent circuit tool estimation that requires expert human input.

Multiphase Identification Algorithm for Fall Recording Systems Using a Single Wearable Inertial Sensor.

Fall-related information can help clinical professionals make diagnoses and plan fall prevention strategies. The information includes various characteristics of different fall phases, such as falling time and landing responses. To provide the information of different phases, this pilot study proposes an automatic multiphase identification algorithm for phase-aware fall recording systems. Seven young adults are recruited to perform the fall experiment. One inertial sensor is worn on the waist to collect the data of body movement, and a total of 525 trials are collected. The proposed multiphase identification algorithm combines machine learning techniques and fragment modification algorithm to identify pre-fall, free-fall, impact, resting and recovery phases in a fall process. Five machine learning techniques, including support vector machine, k-nearest neighbor (kNN), naïve Bayesian, decision tree and adaptive boosting, are applied to identify five phases. Fragment modification algorithm uses the rules to detect the fragment whose results are different from the neighbors. The proposed multiphase identification algorithm using the kNN technique achieves the best performance in 82.17% sensitivity, 85.74% precision, 73.51% Jaccard coefficient, and 90.28% accuracy. The results show that the proposed algorithm has the potential to provide automatic fine-grained fall information for clinical measurement and assessment.

Separation of magnetotelluric signals based on refined composite multiscale dispersion entropy and orthogonal matching pursuit

Magnetotelluric (MT) data processing can increase the reliability of measured data. Traditional MT data denoising methods are usually applied to entire MT time-series, which results in the loss of useful MT signals and a decrease of imaging accuracy of electromagnetic inversion. However, targeted MT noise separation can retain part of the signal unaffected by strong noise and enhance the quality of MT responses. Thus, we propose a novel method for MT noise separation that uses the refined composite multiscale dispersion entropy (RCMDE) and the orthogonal matching pursuit (OMP) algorithm. First, the RCMDE is extracted from each segment of the MT data. Then, the RCMDEs for each segment are input to the fuzzy c-mean (FCM) clustering algorithm for automatic identification of the MT signal and noise. Next, the OMP method is utilized to remove the identified noise segments independently. Finally, the reconstructed signal consists of the denoised signal segments and the identified useful signal segments. We conducted simulation experiments and algorithm evaluations on electromagnetic transfer function (EMTF) data, simulated data and measured sites. The results indicate that the RCMDE can improve the stability of multiscale dispersion entropy (MDE) and multiscale entropy (ME) by analyzing the characteristics of the signal samples library, effectively distinguishing MT signals and noise. Compared with the existing technique of denoising entire time series, the proposed method uses the RCMDE as characteristic parameter and uses the OMP algorithm for noise separation, simplifies the multi-feature fusion, and improves the accuracy of signal-noise identification. Moreover, the denoising efficiency is accelerated, and the MT response in the low-frequency band is greatly improved.

Spatiotemporal Variability of Mesoscale Eddies in the Indonesian Seas

Mesoscale eddies are ubiquitous in the world ocean and well researched both globally and regionally, while their properties and distributions across the whole Indonesian Seas are not yet fully understood. This study investigates for the first time the spatiotemporal variations and generation mechanisms of mesoscale eddies across the whole Indonesian Seas. Eddies are detected from altimetry sea level anomalies by an automatic identification algorithm. The Sulu Sea, Sulawesi Sea, Maluku Sea and Banda Sea are the main eddy generation regions. More than 80% of eddies are short-lived with a lifetime below 30 days. The properties of eddies exhibit high spatial inhomogeneity, with the typical amplitudes and radiuses of 2–6 cm and 50–160 km, respectively. The most energetic eddies are observed in the Sulawesi Sea and Seram Sea. Eddies feature different seasonal cycles between anticyclonic and cyclonic eddies in each basin, especially given that the average latitude of the eddy centroid has inverse seasonal variations. About 48% of eddies in the Sulawesi Sea are highly nonlinear, which is the case for less than 30% in the Sulu Sea and Banda Sea. Instability analysis is performed using high-resolution model outputs from Bluelink Reanalysis to assess mechanisms of eddy generation. Barotropic instability of the mean flow dominates eddy generation in the Sulu Sea and Sulawesi Sea, while baroclinic instability is slightly more in the Maluku Sea and Banda Sea.

Towards validation and standardization of automatic gait event identification algorithms for use in paediatric pathological populations

BackgroundTo analyse and interpret gait patterns in pathological paediatric populations, accurate determination of the timing of specific gait events (e.g. initial contract – IC, or toe-off – TO) is essential. As currently used clinical identification methods are generally subjective, time-consuming, or limited to steps with force platform data, several techniques have been proposed based on processing of marker kinematics. However, until now, validation and standardization of these methods for use in diverse gait patterns remains lacking. Research questions1) What is the accuracy of available kinematics-based identification algorithms in determining the timing of IC and TO for diverse gait signatures? 2) Does automatic identification affect interpretation of spatio-temporal parameters?. Methods3D kinematic and kinetic data of 90 children were retrospectively analysed from a clinical gait database. Participants were classified into 3 gait categories: group A (toe-walkers), B (flat IC) and C (heel IC). Five kinematic algorithms (one modified) were implemented for two different foot marker configurations for both IC and TO and compared with clinical (visual and force-plate) identification using Bland-Altman analysis. The best-performing algorithm-marker configuration was used to compute spatio-temporal parameters (STP) of all gait trials. To establish whether the error associated with this configuration would affect clinical interpretation, the bias and limits of agreement were determined and compared against inter-trial variability established using visual identification. ResultsSagittal velocity of the heel (Group C) or toe marker configurations (Group A and B) was the most reliable indicator of IC, while the sagittal velocity of the hallux marker configuration performed best for TO. Biases for walking speed, stride time and stride length were within the respective inter-trial variability values. SignificanceAutomatic identification of gait events was dependent on algorithm-marker configuration, and best results were obtained when optimized towards specific gait patterns. Our data suggest that correct selection of automatic gait event detection approach will ensure that misinterpretation of STPs is avoided.

Automatic classification and identification algorithms for single-environment risk and their application in social changes

This article focuses on the risk assessment about social environmental safety-. Based on multi-rule, multi-standard, multi-level, multi-factor, and multi-dimension classification principles, this paper proposes a layer-by-layer classification and automatic identification algorithm for single-environment risks. The layer-by-layer classification, k-classification, single-environment risk, and risk information are defined. This paper first proposes an algorithm for the layer-by-layer classification of risk, and -an automatic identification algorithm of layer-by-layer classificationfor single-environment risk. By designing a rule for thelayer-by-layer classification of risk, this paper innovatively studies the automatic identification, evaluation, measurement, and classification of risk in a single environment. The algorithms are applied to a real case where a woman in Dalian, China was suddenly beaten by a stranger in the middle of the night on 22 June 2019. Through four iterations of gradual classification and automatic identification, the environment is finally labeled as "top risk," and a corresponding numerical risk value is calculated. This research is practical and can be used to provide guidance to populations on how to identify and avoid certain risks.

Automatic Roadblock Identification Algorithm for Unmanned Vehicles Based on Binocular Vision

In order to improve the accuracy of automatic obstacle recognition algorithm for driverless vehicles, an automatic obstacle recognition algorithm for driverless vehicles based on binocular vision is constructed. Firstly, the relevant parameters of the camera are calibrated around the new car coordinate system to determine the corresponding obstacle position of the vehicle. At the same time, the three‐dimensional coordinates of obstacle points are obtained by binocular matching method. Then, the left and right cameras are used to capture the feature points of obstacles in the image to realize the recognition of obstacles. Finally, the experimental results show that for obstacle 1, the recognition error of the algorithm is 0.03 m; for obstacle 2, the recognition error is 0.02 m; for obstacle 3, the recognition error is 0.01 m. The algorithm has small recognition error. The vehicle coordinate system is added in the camera calibration process, which can accurately measure the relative position information between the vehicle and the obstacle.

Automatic Identification Algorithm of the Rice Tiller Period Based on PCA and SVM

The tillering period of rice is the crucial phenological period for the cultivation of high-quality and high-yield rice. Currently, human inspection is mainly used for identification, but it is time-consuming, laborious, and prone to mistakes. To efficiently and accurately identify the start date of rice tillering in the monitoring area, this paper proposed a new algorithm called rice tiller period recognition combining principal component analysis (PCA) and a support vector machine (SVM) (RTR-CPS). This algorithm characterizes the problem of identifying the rice tiller stage as a binary classification problem of rice either entering or not entering the tiller stage. To improve the image segmentation quality of traditional visual segmentation methods, the algorithm was designed to extract five image feature values to describe the rice tiller stage in a multi-featured way, reducing the impact of single feature value bias on the identification of the rice tiller stage. To improve the performance of the rice tiller stage recognition model, the algorithm selects ideal principal features in limited sample data by the PCA algorithm and optimizes SVM classification model hyper-parameters by combining 5-fold cross-validation. The experimental results showed that the accuracy of the algorithm for identifying the tillering date of potted rice was as high as 97.76%, which is significantly higher than other competitive methods, and the maximum error between the detection results and human inspection of potted rice tillering period was no more than 2 days. The rice tillering stage recognition model was applied to the field, and the images of field rice planted by two different methods were tested, which verified that the algorithm proposed in this paper is generalizable. These results fully demonstrate the feasibility and superiority of the algorithm in this paper.

Railway Panorama: A Fast Inspection Method for High-Speed Railway Infrastructure Monitoring

At present, the comprehensive inspection trains of all countries in the world are equipped with an environmental video inspection system. The running environment is monitored by manual analysis of the video data, which is labor intensive and inefficient. In this paper, we propose a fast and intelligent inspection method for high-speed railway infrastructure monitoring. A railway environment video is acquired by an imaging device installed on a high-speed integrated detection train, and the forward-looking video is used to establish a panoramic imaging model by constructing the rectangular sampling area having a dynamic width from each frame of the video sequence, which covers the four side scenes of the railway and generate four panoramic images through image stitching technology. The experimental analysis indicates that proposed method makes the examination and analysis of video faster, simpler and more intelligent, which not only simplifies video content and reduces the time cost for users to browse video; but is also more feasible to run automatic identification algorithm on the compressed panoramic image for realizing the automatic intelligent detection of railway environment anomalies.

Capacitive micromachined ultrasonic transducers leak detection by dye penetrant test

Dye penetrant test has been used to detect non hermetic capacitive micromachined ultrasonic transducers (CMUTs). The method has proven its ability to highlight faulty CMUTs membranes among more than hundreds of CMUTs membranes. The fluorescence microscopy image highlights the defective CMUTs with a strong contrast and respecting the shape of the CMUTs, which makes it possible to design robust faulty CMUT automatic identification algorithms. Scanning electron microscopy has been performed in order to check the reliability the detection of the non-hermetic CMUTs. Observations confirmed the reliability of the method.

Early damage assessment in large-scale structures by innovative statistical pattern recognition methods based on time series modeling and novelty detection

Time series analysis and novelty detection are effective and promising methods for data-driven structural health monitoring (SHM) based on the statistical pattern recognition paradigm. However, processing substantially large volumes of vibration measurements may represent a serious limitation, especially for long-term SHM programs of large-scale civil structures. Moreover, shortcomings like the choice of an appropriate time series model in an automatic manner, the determination of optimal orders of the identified model and the classification of random high-dimensional features for damage detection, can strongly affect the performance of these approaches. This study is intended to propose statistical pattern recognition methods regarding time series modeling for feature extraction and novelty detection in feature classification in the presence of big data. These methods include an automatic model identification algorithm, an improved order determination approach and a hybrid distance-based novelty detection through a combination of Partition-based Kullback-Leibler divergence and Mahalanobis-squared distance. Experimental datasets relevant to a cable-stayed bridge are considered to validate the effectiveness of the proposed methods. Results demonstrate that: the AutoRegressive-AutoRegressive with eXogenous input (AR-ARX) model turns out to be the most suitable representation for feature extraction; the orders of this model are efficiently and automatically determined; the proposed novelty detection approach is highly successful in detecting damage, even in case of large volumes of random high-dimensional features.

Mathematical model of fire load type identification by electromagnetic radiation of the optical range using an unmanned aerial vehicle at oil and gas industry facilities

Providing comprehensive security of the pipeline is the main task of monitoring linear facilities in the oil and gas industry. Proposed mathematical model is a part of an automated identification algorithm to determine the fire load type. The study identified four groups of combustible substances characteristic of oil and gas facilities: oil, refined products, liquefied petroleum gas and vegetation. There was a study of electromagnetic radiation of the optical range, recorded by an UAV’S digital camera for each type of substance. An algorithm for comparing the area color with a standard was proposed and implemented. The obtained mathematical correspondence allows identifying the fire color and the fire load type in conditions of different luminosity. The information on the flaming substance allows authorized rescue units to make operational decisions about the fire extinguishing method.

A method for checking high-redshift identification of radio AGNs

ABSTRACT In large-scale optical spectroscopic surveys, there are many objects found to have multiple redshift measurements due to the weakness of their emission lines and the different automatic identification algorithms used. These include some suspicious high-redshift $(z \gtrsim 5)$ active galactic nuclei (AGNs). Here, we present a method for inspecting the high-redshift identification of such sources provided that they are radio-loud and have very long baseline interferometry (VLBI) imaging observations of their milli-arcsec (mas) scale jet structure available at multiple epochs. The method is based on the determination of jet component proper motions, and the fact that the combination of jet physics (the observed maximal values of the bulk Lorentz factor) and cosmology (the time dilation of observed phenomena in the early Universe) constrain the possible values of apparent proper motions. As an example, we present the case of the quasar J2346 + 0705 that was reported with two different redshifts, z1 = 5.063 and z2 = 0.171, in the literature. We measured the apparent proper motions (μ) of three components identified in its radio jet by utilizing VLBI data taken from 2014 to 2018. We obtained μJ1 = 0.334 ± 0.099 mas yr−1, μJ2 = 0.116 ± 0.029 mas yr−1, and μJ3 = 0.060 ± 0.005 mas yr−1. The maximal proper motion is converted to an apparent transverse speed of $\beta _{\rm app} = 41.2\pm 12.2\, c$, if the source is at redshift 5.063. This value exceeds the blazar jet speeds known to date. This and other arguments suggest that J2346 + 0705 is hosted by a low-redshift galaxy. Our method may be applicable for other high-redshift AGN candidates lacking unambiguous spectroscopic redshift determination or having photometric redshift estimates only, but showing prominent radio jets allowing for VLBI measurements of fast jet proper motions.

Geological Survey and Unstable Rock Block Movement Monitoring of a Post-Earthquake High Rock Slope Using Terrestrial Laser Scanning

Local topography and structural features of discontinuities have a great effect on the movement characteristics and failure mechanisms of the unstable rock blocks of the rock slope. By taking full advantage of the geometric information contained in the multi-temporal terrestrial laser scanning (TLS) point clouds, a technique based on a roto-translation method has been developed to track and monitor the movement behavior of blocks in unstable rock slopes. The technique can be completed in three stages: (a) a preliminary change detection is used to distinguish the unstable area by means of a shortest distance (SD) algorithm; (b) an automatic discontinuity identification algorithm is implemented to visually identify the discontinuity sets and to exactly delineate rock blocks in unstable areas; and (c) finally a roto-translation movement monitoring method is applied to track and monitor the unstable rock blocks. This technique permits the user to visually identify the discontinuity sets and to determine their orientation, and therefore provides detailed information of geometric compounding relationships between the slope and joint systems as well as the forming rock blocks which are the key impacts to stability of the rock slope. It permits the user to exactly and visually delineate the rock blocks in the unstable area, and thus the completely tracking and monitoring of lock block movement in unstable rock slopes can be effectively conducted. It overcomes the limited precision of the single TLS points and provides an actual 3D movement measuring method for the blocks rather than a shortest distance. The technique was first used in the movement behavior monitoring of the unstable rock blocks in a post-earthquake high rock slope. The results show that the rock block movement monitoring technique is well suited for providing high-quality data in the assessment of rock failure hazards.

An algorithm competition for automatic species identification from herbarium specimens.

PremisePlant biodiversity is threatened, yet many species remain undescribed. It is estimated that >50% of undescribed species have already been collected and are awaiting discovery in herbaria. Robust automatic species identification algorithms using machine learning could accelerate species discovery.MethodsTo encourage the development of an automatic species identification algorithm, we submitted our Herbarium 2019 data set to the Fine‐Grained Visual Categorization sub‐competition (FGVC6) hosted on the Kaggle platform. We chose to focus on the flowering plant family Melastomataceae because we have a large collection of imaged herbarium specimens (46,469 specimens representing 683 species) and taxonomic expertise in the family. As is common for herbarium collections, some species in this data set are represented by few specimens and others by many.ResultsIn less than three months, the FGVC6 Herbarium 2019 Challenge drew 22 teams who entered 254 models for Melastomataceae species identification. The four best algorithms identified species with >88% accuracy.DiscussionThe FGVC competitions provide a unique opportunity for computer vision and machine learning experts to address difficult species‐recognition problems. The Herbarium 2019 Challenge brought together a novel combination of collections resources, taxonomic expertise, and collaboration between botanists and computer scientists.

Enhanced winter snowmelt in the Antarctic Peninsula: Automatic snowmelt identification from radar scatterometer

Surface snowmelt affects the energy balance through melt-albedo feedback and may endanger the ice shelves in the Antarctic Peninsula (AP) through hydrofracture. Here, we introduce an automatic snowmelt identification algorithm based on Quick Scatterometer and Advanced Scatterometer. The proposed method can provide self-adaptive thresholds for snowmelt detection based on Rosin thresholding, and is able to detect weak melt signals with a wavelet denoising procedure. Results suggest the AP surface snowmelt has slightly declined during 1999–2018 in the context of recent cooling. However, winter melt index (i.e., melt days times the area of melting) has significantly increased (above the 99% confidence level) with a rate of 83% decade−1. Unprecedented winter snowmelt was found in 2015/2016 when about one-third of the AP experienced snowmelt in late May. This may be attributed to the intrusions of marine air and the enhanced föhn-driven warming resulted from the abnormal northwesterly flow driven by a significant deepening of the Amundsen Sea Low.

Automatic impurity spectral line identification algorithm with noise reduction for fusion plasmas

Accurate diagnostics of impurity behavior inside the tokamak plasma is essential for long and stable plasma operation and also for machine protection in fusion devices including ITER. In this study, a numerical code for identifying impurity line spectra was developed and assessed by utilizing the ITER-relevant vacuum ultraviolet (VUV) spectroscopic diagnostic system in KSTAR. For real-time analysis of a spectrum, it is necessary to reduce the noise level of the data caused by high neutron and gamma photon fluxes reaching the detector such as a charge-coupled device detector. The code uses a high-order derivative method that distinguishes relatively sharp noises from spectral lines. Tests with synthetic spectra showed successful noise reduction of approximately 90%. In addition, an in-situ wavelength calibration algorithm was developed by using representative carbon emission lines (C III and C IV) as markers that appear during the current ramp-up phase in carbon-walled devices. This algorithm is followed by a matching algorithm that enables the annotation of an ionic state of an impurity on the line peak by referencing the NIST atomic database. From the integrated processing algorithm, simulated VUV spectra were tested and the results showed successful automatic annotation of the spectral lines after noise reduction.

Automatic Identification of Individual Motor Unit Firing Accuracy From High-Density Surface Electromyograms.

We introduce an algorithm for automatic identification of true positive (TP) and false positive (FP) spikes in the motor unit spike train, identified by blind source separation (BSS) of high-density surface electromyograms (HDsEMG). The algorithm selects predefined number of spikes, so called witnesses, from identified spike train. The other spikes in the spike train are called test spikes and are classified into TP or FP spikes by our algorithm. For this purpose, the algorithm constructs as many motor unit filters as there are test spikes, using the information from all the witnesses and each individual test spike. Afterwards, it applies each motor unit filter to HDsEMG to get new estimate of MU spike train for each selected test spike and calculates previously introduced Pulse-to-Noise Ratio (PNR) on preselected witnesses in this new spike train. When accumulated over all the test spikes, these PNR values exhibit bimodal distribution with the peak at lower PNR values representing FPs and the peak at higher PNR values representing TPs. Therefore, FPs and TPs can be discriminated by applying computationally efficient segmentation algorithm to corresponding PNR values. We also propose and mutually compare different witness selection strategies and show that selection of about 40 spikes with maximal amplitude in the identified spike train minimizes the selection of FPs as witnesses and maximizes the TP vs. FP discrimination power. In our tests on 20 s long experimental HDsEMG signals from biceps brachii muscle the number of FPs decreased from 23.9 ± 4.7 to 4.1 ± 4.4 when the proposed algorithm was used.

Real-time reef fishes identification using deep learning

Reef fishes is an important part in maintaining the balance of various components in the coral reef ecosystem. The existence of reef fish on coral reef ecosystems is a marker of the ecosystem in good condition. Furthermore, it is important to observe the condition of reef fish in a coral reef ecosystem to determine the population and diversity of reef fish in the ecosystem. Observation of reef fish generally by performing a manual visual census by scuba diver. In entering the industrial revolution 4.0 era there is a need to develop technology that is used to monitor the condition of reef fish in a coral reef ecosystem. The development of technology will certainly help researchers, and later on ecosystem manager, in observing the condition of reef fish with automatic identification. The technological development that can be done to observe reef fish is by applying deep learning. In this research we used YOLO deep learning algorithm for automatic identification. YOLO has the advantage of faster object detection. Application of deep learning to identify fish automatically is illustrated using underwater video recording of reef fish.