Real-time Automatic Detection Research Articles

Aprendizado de Máquina Fuzzy Incremental para Classificação de Faltas em Sistemas de Potência

O conceito de Qualidade da Energia está relacionado a um conjunto de alterações que podem ocorrer no sistema elétrico. Tais alterações (distúrbios/faltas) podem ocorrer em várias partes do sistema de energia, sejam nas instalações elétricas dos consumidores ou no sistema supridor da concessionária, causando prejuízos financeiros a ambas as partes. Por isso, é de fundamental importância a detecção e classificação em tempo real desses distúrbios de modo automático. Para reconhecimento e classificação de padrões, são considerados modelos inteligentes evolutivos, ou seja, modelos equipados com algoritmos de aprendizado incrementais online capazes de alterar seus parâmetros e estrutura, conforme novas informações surgem em um fluxo de dados. Em particular, é considerada a modelagem evolutiva baseada em conjuntos fuzzy (FBeM). Para o pré-processamento dos dados mensurados, e extração de variáveis indicadoras da presença de distúrbios, é considerado o filtro Hodrick-Prescott, a técnica de transformada rápida de Fourier e o valor eficaz da tensão. O modelo desenvolvido neste trabalho tem alcançado um desempenho comparável àqueles de modelos estado da arte na área de qualidade de energia. Detecção e classificação de distúrbios tais como elevação de tensão, sub-harmônico, transitório oscilatório, spikes e notching, ocorrendo de forma simultânea ou não, são alcançadas com acurácia de, aproximadamente, 99%.

Artificial Intelligence-Assisted Colonoscopy for Detection of Colon Polyps: a Prospective, Randomized Cohort Study

Background and aimsImproving the rate of polyp detection is an important measure to prevent colorectal cancer (CRC). Real-time automatic polyp detection systems, through deep learning methods, can learn and perform specific endoscopic tasks previously performed by endoscopists. The purpose of this study was to explore whether a high-performance, real-time automatic polyp detection system could improve the polyp detection rate (PDR) in the actual clinical environment. MethodsThe selected patients underwent same-day, back-to-back colonoscopies in a random order, with either traditional colonoscopy or artificial intelligence (AI)-assisted colonoscopy performed first by different experienced endoscopists (> 3000 colonoscopies). The primary outcome was the PDR. It was registered with clinicaltrials.gov. (NCT047126265). ResultsIn this study, we randomized 150 patients. The AI system significantly increased the PDR (34.0% vs 38.7%, p < 0.001). In addition, AI-assisted colonoscopy increased the detection of polyps smaller than 6 mm (69 vs 91, p < 0.001), but no difference was found with regard to larger lesions. ConclusionsA real-time automatic polyp detection system can increase the PDR, primarily for diminutive polyps. However, a larger sample size is still needed in the follow-up study to further verify this conclusion. Trial Registrationclinicaltrials.gov Identifier: NCT047126265

Efficient vehicle detection and tracking strategy in aerial videos by employing morphological operations and feature points motion analysis

Real-time automatic detection and tracking of moving vehicles in videos acquired by airborne cameras is a challenging problem due to vehicle occlusion, camera movement, and high computational cost. This paper presents an efficient and robust real-time approach for automatic vehicle detection and tracking in aerial videos that employ both detections and tracking features to enhance the final decision. The use of Top-hat and Bottom-hat transformation aided by the morphological operation in the detection phase has been adopted. After detection, background regions are eliminated by motion feature points’ analysis of the obtained object regions using a combined technique between KLT tracker and K-means clustering. Obtained object features are clustered into separate objects based on their motion characteristic. Finally, an efficient connecting algorithm is introduced to assign the vehicle labels with their corresponding cluster trajectories. The proposed method was tested on videos taken in different scenarios. The experimental results showed that the recall, precision, and tracking accuracy of the proposed method were about 95.1 %, 97.5%, and 95.2%, respectively. The method also achieves a fast processing speed. Thus, the proposed approach has superior overall performance compared to newly published approaches.

Measurement for cracks at the bottom of bridges based on tethered creeping unmanned aerial vehicle

The detection of bridge bottom cracks is required for bridge maintenance. In order to realise the requirement of automatic real-time detection of a bridge structure, an image detection method for cracks in the bottom of the bridge structure using a tethered creeping unmanned aerial vehicle (UAV) is proposed. A high-precision unlimited endurance detection plan based on the tethered creeping UAV is designed to use for the bottom cracks of the bridge structure. The detection scheme applies a high-precision image stitching measurement algorithm for cracks at the bottom of the beam body, which is able to restore a panoramic image. All mainstream filtering methods were evaluated, and it turned out that they are practicable/applicable in various crack images of different shapes. The method is applied to the crack detection in the bottom of bridge structures to ensure the accuracy and efficiency of the system measurement. According to the actual measurement by the laboratory platform, the measurement error using this method is less than 0.1 mm, which meets the requirements of measurement automation. The results of the research represent an initial step towards developing an automatic bridge health monitoring and evaluating system.

Artificial Intelligence Medical Ultrasound Equipment: Application of Breast Lesions Detection.

Breast cancer ranks first among cancers affecting women's health. Our work aims to realize the intelligence of the medical ultrasound equipment with limited computational capability, which is used for the assistant detection of breast lesions. We embed the high-computational deep learning algorithm into the medical ultrasound equipment with limited computational capability by two techniques: (1) lightweight neural network: considering the limited computational capability of ultrasound equipment, a lightweight neural network is designed, which greatly reduces the amount of calculation. And we use the technique of knowledge distillation to train the low-precision network helped with the high-precision network; (2) asynchronous calculations: consider four frames of ultrasound images as a group; the image of the first frame of each group is used as the input of the network, and the result is respectively fused with the images of the fourth to seventh frames. An amount of computation of 30 GFLO/frame is required for the proposed lightweight neural network, about 1/6 of that of the large high-precision network. After trained from scratch using the knowledge distillation technique, the detection performance of the lightweight neural network (sensitivity = 89.25%, specificity = 96.33%, the average precision [AP] = 0.85) is close to that of the high-precision network (sensitivity = 98.3%, specificity = 88.33%, AP = 0.91). By asynchronous calculation, we achieve real-time automatic detection of 24 fps (frames per second) on the ultrasound equipment. Our work proposes a method to realize the intelligence of the low-computation-power ultrasonic equipment, and successfully achieves the real-time assistant detection of breast lesions. The significance of the study is as follows: (1) The proposed method is of practical significance in assisting doctors to detect breast lesions; (2) our method provides some practical and theoretical support for the development and engineering of intelligent equipment based on artificial intelligence algorithms.

Change Detection in Multitemporal Monitoring Images Under Low Illumination

Video surveillance may involve the simultaneous monitoring of a large number of areas. Real-time automatic change detection of a monitoring area (such as involving the movement of people or vehicles) can reduce risks incurred in negligent manual observation. However, the low signal-to-noise ratio (SNR) of dark environments can significantly corrupt camera images, making it difficult for machine learning surveillance systems to detect small changes in monitored images. In addition, in the absence of changes between two multitemporal monitoring images, sensor noise can lead to false alarms. The objective of this paper is to reduce the effect of sensor noise on change detection of monitored images and the run time of change detection algorithms. For these purposes, we proposed a novel multitemporal monitoring image change detection algorithm based on morphological structure filtering and normalized fusion difference image. First, the random noise in two surveillance images was removed using a multidirectional weighted multiscale series of a morphological filter. Next, two difference images were obtained by using the compression log-ratio operator and the mean ratio operator, and a fusion difference image was obtained by equal-weight fusion of the two difference images. Then, the sigmoid function was used to compress the fusion difference map to obtain a normalized fusion difference image, and a median filter was used to obtain a final difference image. Finally, the k-means clustering algorithm was utilized to obtain the change detection results. The experimental results demonstrate that the proposed method can accurately detect changes in a night monitoring scene in real time. Subjective and objective evaluation of the experimental results demonstrate that the proposed method is superior to reference algorithms in terms of change detection accuracy, time and robustness.

Automatic detection and damage quantification of multiple cracks on concrete surface from video

Real-time automatic detection of multiple cracks from a video stream of a concrete surface is addressed in this paper. Robust principal component analysis is used to detect multiple cracks forming at different instances of time in an unsupervised manner using the Gini index as a metric to quantify the presence of an observable crack. The relative positions of the relevant pixels around the crack are monitored using the Kanade Lucas Tomasi feature tracking algorithm. Further, Hu's invariant moments of those pixel positions are computed which acts as a robust damage indicator even for breathing cracks under time-varying service loads. The proposed method is experimentally validated using two small scale under-reinforced beams undergoing three-point bending tests. The method successfully detects the onset of multiple cracks, at varied locations, at different time instants and further tracks their propagations.

Automatic detection and damage quantification of multiple cracks on concrete surface from video

Real-time automatic detection of multiple cracks from a video stream of a concrete surface is addressed in this paper. Robust principal component analysis is used to detect multiple cracks forming at different instances of time in an unsupervised manner using the Gini index as a metric to quantify the presence of an observable crack. The relative positions of the relevant pixels around the crack are monitored using the Kanade Lucas Tomasi feature tracking algorithm. Further, Hu's invariant moments of those pixel positions are computed which acts as a robust damage indicator even for breathing cracks under time-varying service loads. The proposed method is experimentally validated using two small scale under-reinforced beams undergoing three-point bending tests. The method successfully detects the onset of multiple cracks, at varied locations, at different time instants and further tracks their propagations.

Real-Time Object Detection for AUVs Using Self-Cascaded Convolutional Neural Networks

This article presents an automatic real-time object detection method using sidescan sonar (SSS) and an onboard graphics processing unit (GPU). The detection method is based on a modified convolutional neural network (CNN), which is referred to as self-cascaded CNN (SC-CNN). The SC-CNN model segments SSS images into object-highlight, object-shadow, and seafloor areas, and it is robust to speckle noise and intensity inhomogeneity. Compared with typical CNN, SC-CNN utilizes crop layers which enable the network to use local and global features simultaneously without adding convolution parameters. Moreover, to take the local dependencies of class labels into consideration, the results of SC-CNN are postprocessed using Markov random field. Furthermore, the sea trial for real-time object detection via the presented method was implemented on our autonomous underwater vehicle (AUV) named SAILFISH via its GPU module at Jiaozhou Bay Bridge, Qingdao, China. The results show that the presented method for SSS image segmentation has obvious advantages when compared with the typical CNN and unsupervised segmentation methods, and is applicable in real-time object detection task.

Methodology for Full Waveform Near Real-Time Automatic Detection and Localization of Microseismic Events Using High (8 kHz) Sampling Rate Records in Mines: Application to the Garpenberg Mine (Sweden)

AbstractRecent studies have demonstrated the success of automatic full-waveform detection and location methods in analyzing and monitoring natural and induced seismicity. These approaches have been shown to provide a significant improvement in events detectability, increasing the significance of statistical analysis that permits to identify small changes of seismicity rates in space and time. Although currently nontrivial and by far nonstandard, application of such methods to seismic monitoring of active mines could significantly improve forecasting of potential destructive rockburst events. The main challenges of such applications are related to the presence of a wide range of seismic noise sources that have to do with mining activity and a high sampling rate of recorded data (several kHz), posing problems for real-time data transfer and processing.In this study, we propose an adapted full-waveform-based automatic method for the detection and location of microseismic events that makes use of continuous seismic records from an in-mine seismic network and can be adjusted to a near-real-time monitoring scheme. The method consists of two steps: (1) event extraction and amplitude ratio-based preliminary location and (2) event relocation using a coherency-based backprojection approach. The event extraction, based on multiband signal characterization implemented in the first step, allows us to overcome the challenge of high sampling rate data (8 kHz), reducing the overall volume of transferred data and providing an energy-based signal classification scheme. This allows us to remove a significant number of machinery noise sources. The technique is developed and tested on the case study of the Garpenberg mine (Sweden) monitored by a local seismic network that is maintained by Ineris. We demonstrate the improvement in event detection capacity by a factor of 50, compared with the standard triggered-based monitoring schemes. This increased number of detected microseismic events permits us to investigate the migration pattern of induced microseismicity that is generated in response to production blast.

A DEEP LEARNING APPROACH FOR URBAN UNDERGROUND OBJECTS DETECTION FROM VEHICLE-BORNE GROUND PENETRATING RADAR DATA IN REAL-TIME

Abstract. GPRs (Ground Penetrating Radar) are widely adopted in underground space survey and mapping, because of their advantages of fast data acquisition, convenience, high imaging resolution and NDT (Non Destructive Testing) inspection. However, at present, the automation of the GPR data post-processing is low and the identification of underground objects needs expert interpretation. The heavy manual interpretation labor limits the GPR applications in large-scale urban scenarios. According to the latest research, it is still an unsolved problem to detect targets or defects in GPR data automatically and needs further exploration. In this paper, we propose a deep learning method for real-time detection of underground targets from GPR data. Seven typical targets in urban underground space are identified and labelled to construct the training dataset. The constructed dataset is consist of 489 labelled samples including rainwater wells, cables, metal/nonmetal pipes, sparse/dense steel reinforcement, voids. The training dataset is further augmented to produce more samples. DarkNet53 convolutional neural network (CNN) is trained using the constructed training dataset including realistic data and augmented data to extract features of the buried objects. And then the end-to-end YOLO detection framework is used to classify and locate the seven specific categories buried targets in the GPR data in real time. Experiments show that the automatic real-time detection method proposed in this paper can effectively detect the buried objects in the ground penetrating radar image in real time at Shenzhen test site (typical urban road scene).

Real-time automatic detection system increases colonoscopic polyp and adenoma detection rates: a prospective randomised controlled study

ObjectiveThe effect of colonoscopy on colorectal cancer mortality is limited by several factors, among them a certain miss rate, leading to limited adenoma detection rates (ADRs). We investigated the effect...

Spell checker for consumer language (CSpell).

Automated understanding of consumer health inquiries might be hindered by misspellings. To detect and correct various types of spelling errors in consumer health questions, we developed a distributable spell-checking tool, CSpell, that handles nonword errors, real-word errors, word boundary infractions, punctuation errors, and combinations of the above. We developed a novel approach of using dual embedding within Word2vec for context-dependent corrections. This technique was used in combination with dictionary-based corrections in a 2-stage ranking system. We also developed various splitters and handlers to correct word boundary infractions. All correction approaches are integrated to handle errors in consumer health questions. Our approach achieves an F1 score of 80.93% and 69.17% for spelling error detection and correction, respectively. The dual-embedding model shows a significant improvement (9.13%) in F1 score compared with the general practice of using cosine similarity with word vectors in Word2vec for context ranking. Our 2-stage ranking system shows a 4.94% improvement in F1 score compared with the best 1-stage ranking system. CSpell improves over the state of the art and provides near real-time automatic misspelling detection and correction in consumer health questions. The software and the CSpell test set are available at https://umlslex.nlm.nih.gov/cSpell.

The impact of position errors on crowd simulation

In large crowd events, there is always a potential possibility that a stampede accident will occur. The accident may cause injuries or even death. Approaches for controlling crowd flows and predicting dangerous congestion spots would be a boon to on-site authorities to manage the crowd and to prevent fatal accidents. One of the most popular approaches is real-time crowd simulation based on position data from personal Global Positioning System (GPS) devices. However, the accuracy of spatial data varies for different GPS devices, and it is also affected by an environment in which an event takes place. In this paper, we would like to assess the effect of position errors on stampede prediction. We propose an Automatic Real-time dEtection of Stampedes (ARES) method to predict stampedes for large events. We implement three different stampede assessment methods in Menge framework and incorporate position errors. Our analysis suggests that the probability of simulated stampede changes significantly with the increase of the magnitude of position errors, which cannot be eliminated entirely with the help of classic techniques, such as the Kalman filter. Thus, it is our position that novel stampede assessment methods should be developed, focusing on the detection of position noise and the elimination of its effect.

An approach to automatic process deviation detection in a time-critical clinical process.

Prior research has shown that minor errors and deviations from recommended guidelines in complex medical processes can accumulate to increase the likelihood that a major error will go uncorrected and lead to an adverse outcome. Real-time automatic and accurate detection of process deviations may help medical teams better prevent or mitigate the effect of errors and improve patient outcomes. Our goal was to develop an approach for automatic detection of errors and process deviations in trauma resuscitation. Using video review, we coded activity traces of 95 pediatric trauma resuscitations collected in a Level 1 trauma center over two years (2014-2016). Twenty-four randomly selected activity traces were compared with a knowledge-driven model of trauma resuscitation workflow using a phase-based conformance checking algorithm for detecting true and false deviations (alarms). An analysis of false alarms identified three types of causes: (1) model gaps or discrepancies between the model ("work as imagined") and actual practice ("work as done"), (2) errors in activity traces coding, and (3) algorithm limitations. We repaired the system to remove model gaps, reduce coding errors, and address algorithm limitations. The repaired system was first evaluated with another 20 traces and then applied to the entire dataset of 95 traces. During the training, we detected 573 process deviations in 24 activity traces that include 1099 activities. Among these deviations, only 27% represented true deviations and the remaining 73% were false alarms. This initial deviation detection accuracy was only 66.6%, with a F1-score of 0.42. Detection accuracy of the repaired system increased to 95.2% (0.85 F1-score) during system validation and to 98.5% (0.96 F1-score) during testing. After deploying the repaired deviation detection system to all 95 activity traces, we detected 1060 process deviations in 5659 activities (11.2 deviations per resuscitation). Among the 5659 activities in these traces, 4893 fit the repaired knowledge-driven workflow model, 294 were errors of omission, 538 were errors of commission, and 228 were scheduling errors. Our approach to automatic deviation detection provides a method for identifying repeated, omitted and out-of-sequence activities that can be included in the design of decision support systems for complex medical processes. Our findings show the importance of assessing detected deviations for repairing a knowledge-driven model that best represents "work as done."

Automated detection and manipulation of sleep in C. elegans reveals depolarization of a sleep-active neuron during mechanical stimulation-induced sleep deprivation

Across species, sleep is characterized by a complex architecture. Sleep deprivation is a classic method to study the consequences of sleep loss, which include alterations in the activity of sleep circuits and detrimental consequences on well being. Automating the observation and manipulation of sleep is advantageous to study its regulation and functions. Caenorhabditis elegans shows sleep behavior similar to other animals that have a nervous system. However, a method for real-time automatic sleep detection that allows sleep-specific manipulations has not been established for this model animal. Also, our understanding of how sleep deprivation affects sleep neurons in this system is incomplete. Here we describe a system for real-time automatic sleep detection of C. elegans grown in microfluidic devices based on a frame-subtraction algorithm using a dynamic threshold. As proof of principle for this setup, we used automated mechanical stimulation to perturb sleep behavior and followed the activity of the sleep-active RIS neuron. We show that our system can automatically detect sleep bouts and deprive worms of sleep. We found that mechanical stimulation generally leads to the activation of the sleep-active RIS neuron, and this stimulation-induced RIS depolarization is most prominent during sleep deprivation.

Using Temporal Covariance of Motion and Geometric Features via Boosting for Human Fall Detection.

Fall induced damages are serious incidences for aged as well as young persons. A real-time automatic and accurate fall detection system can play a vital role in timely medication care which will ultimately help to decrease the damages and complications. In this paper, we propose a fast and more accurate real-time system which can detect people falling in videos captured by surveillance cameras. Novel temporal and spatial variance-based features are proposed which comprise the discriminatory motion, geometric orientation and location of the person. These features are used along with ensemble learning strategy of boosting with J48 and Adaboost classifiers. Experiments have been conducted on publicly available standard datasets including Multiple Cameras Fall (with 2 classes and 3 classes) and UR Fall Detection achieving percentage accuracies of 99.2, 99.25 and 99.0, respectively. Comparisons with nine state-of-the-art methods demonstrate the effectiveness of the proposed approach on both datasets.

F177. Automated R-peak detection algorithm for patients with epilepsy using portable ECG

Earlier studies have shown that short term heart rate variability (HRV) analysis of ECG is a promising biomarker for detection of epileptic seizures. A precise and accurate automatic R-peak detection algorithm is a necessity in a real-time, continuous measurement of HRV, in a portable ECG device. We used the portable CE marked ePatch® heart monitor to record the ECG of 14 patients, who were enrolled in the video-EEG long term monitoring unit for clinical workup of epilepsy. Recordings of the first 7 patients (356 recording hours) were used as training set of data to optimize the R-peak detection algorithm and the recordings of the last 7 patients (467.6 recording hours) were used to test the performance of the algorithm. We aimed to modify and optimize an existing QRS-detection algorithm to a more precise R-peak detection algorithm to avoid the possible jitter Q- and S-peaks can create in the tachogram, which causes error in short-term HRV-analysis. The proposed R-peak detection algorithm showed a high sensitivity (Se = 99.979%) and positive predictive value (P+ = 99.976%), which was comparable with a previously published QRS-detection algorithm for the ePatch®ECG device, when testing the same dataset. The novel R-peak detection algorithm designed to avoid jitter has very high sensitivity and specificity and thus is a suitable tool for a robust, fast, real-time HRV-analysis in patients with epilepsy. The R-peak detection algorithm is the first important step in creating a portable fully automatic real-time seizure detection for these patients.

Adaptive bulk motion exclusion for improved robustness of abdominal magnetic resonance imaging.

Non‐Cartesian magnetic resonance imaging (MRI) sequences have shown great promise for abdominal examination during free breathing, but break down in the presence of bulk patient motion (i.e. voluntary or involuntary patient movement resulting in translation, rotation or elastic deformations of the body). This work describes a data‐consistency‐driven image stabilization technique that detects and excludes bulk movements during data acquisition. Bulk motion is identified from changes in the signal intensity distribution across different elements of a multi‐channel receive coil array. A short free induction decay signal is acquired after excitation and used as a measure to determine alterations in the load distribution. The technique has been implemented on a clinical MR scanner and evaluated in the abdomen. Six volunteers were scanned and two radiologists scored the reconstructions. To show the applicability to other body areas, additional neck and knee images were acquired. Data corrupted by bulk motion were successfully detected and excluded from image reconstruction. An overall increase in image sharpness and reduction of streaking and shine‐through artifacts were seen in the volunteer study, as well as in the neck and knee scans. The proposed technique enables automatic real‐time detection and exclusion of bulk motion during MR examinations without user interaction. It may help to improve the reliability of pediatric MRI examinations without the use of sedation.

P322 Fully automated R-peak detection algorithm for patients with epilepsy: First step towards portable seizure detector

Objectives Earlier studies have shown that short term heart rate variability (HRV) analysis of ECG is a promising biomarker for detection of epileptic seizures. A precise and accurate automatic R-peak detection algorithm is a necessity in a real-time, continuous measurement of HRV, in a portable ECG device. Methods We used the portable CE marked ePatch® heart monitor to record the ECG of 14 patients, who were enrolled in the video-EEG long term monitoring unit for clinical workup of epilepsy. Recordings of the first 7 patients (356 recording hours) were used as training set of data to optimize the R-peak detection algorithm and the recordings of the last 7 patients (467.6 recording hours) were used to test the performance of the algorithm. We aimed to modify and optimize an existing QRS-detection algorithm to a more precise R-peak detection algorithm to avoid the possible jitter Q- and S-peaks can create in the tachogram, which causes error in short-term HRV-analysis. Results The proposed R-peak detection algorithm showed a high sensitivity (Se = 99.979%) and positive predictive value ( P + = 99.976%), which was comparable with a previously published QRS-detection algorithm for the ePatch® ECG device, when testing the same dataset. Conclusions/significance The novel R-peak detection algorithm designed to avoid jitter has very high sensitivity and specificity and thus is a suitable tool for a robust, fast, real-time HRV-analysis in patients with epilepsy. The R-peak detection algorithm is the first important step in creating a portable fully automatic real-time seizure detection for these patients.