Real-time Surveillance System Research Articles

Surveillance and Security in US Medicine and Equipment Supply Chains.

The COVID-19 pandemic exposed vulnerabilities of the United States' routine and emergency supply chains of medicines and critical equipment. These vulnerabilities underscore an urgent need to prevent routine and emergency shortages by making drug manufacturing more transparent and by tracking how key supplies get to end users. Near real-time surveillance systems must be developed to monitor fluctuations in supplies of medicines and equipment. Implementation of such systems will require getting key stakeholders (clinicians, administrators, community members, manufacturers, and policy makers) to collaborate.

Can real-time surveillance systems of suspected suicide accurately reflect national suicide rates? Age-specific and sex-specific findings from the first two years of the COVID-19 pandemic in England: an observational study

Introduction‘Real-time surveillance’ (RTS) systems of suspected suicide showed no overall rise in the early COVID-19 pandemic several months before official statistics reported the same. There has to date been no...

Determining the psycho-emotional state of the observed based on the analysis of video observations

This paper develops a system for determining the psycho-emotional state of the observed people based on the analysis of video surveillance with the application of artificial intelligence technology using hardware and software tools such as PoseNet, PyTorch, SQLite, FastAPI and Flask. In many areas of human endeavor, there is an urgent need for a surveillance system that can reliably function and detect suspicious activities. To solve this problem, this paper proposes a novel framework for a real-time surveillance system that automatically detects abnormal human activities. The system has been tested and validated in real environments. The results of testing artificial intelligence program models showed the best results (f1 score with values of 0.98–0.99). The weighted average value of the f1-score metric was 0.96, which is quite a high value. The use of PoseNet implemented with PyTorch allowed to accurately determine the pose of the person in the video and extract information about the position of different body parts. The peculiarity of this work lies in the development of artificial intelligence models for automatic detection of possible physical aggression in videos, in the methods of forming an optimal set of features for the development of AI models that identify the aggressor and the victim of bullying. The developed system has the potential to be a useful tool in various fields such as psychology, medicine, security and others where it is important to analyze the emotional state of people based on their physical manifestations. The obtained applied results can be used in educational institutions and in spheres where video analysis is necessary

Conv3D-Based Video Violence Detection Network Using Optical Flow and RGB Data.

Detecting violent behavior in videos to ensure public safety and security poses a significant challenge. Precisely identifying and categorizing instances of violence in real-life closed-circuit television, which vary across specifications and locations, requires comprehensive understanding and processing of the sequential information embedded in these videos. This study aims to introduce a model that adeptly grasps the spatiotemporal context of videos within diverse settings and specifications of violent scenarios. We propose a method to accurately capture spatiotemporal features linked to violent behaviors using optical flow and RGB data. The approach leverages a Conv3D-based ResNet-3D model as the foundational network, capable of handling high-dimensional video data. The efficiency and accuracy of violence detection are enhanced by integrating an attention mechanism, which assigns greater weight to the most crucial frames within the RGB and optical-flow sequences during instances of violence. Our model was evaluated on the UBI-Fight, Hockey, Crowd, and Movie-Fights datasets; the proposed method outperformed existing state-of-the-art techniques, achieving area under the curve scores of 95.4, 98.1, 94.5, and 100.0 on the respective datasets. Moreover, this research not only has the potential to be applied in real-time surveillance systems but also promises to contribute to a broader spectrum of research in video analysis and understanding.

Computer Vision-Based Intelligent Traffic Surveillance: Multi-Vehicle Tracking and Detection

The application of vehicle detection in real-time traffic surveillance systems is one of the challenging research fields with different objectives. One of the problems is the detection of many vehicles simultaneously in a video sequence sourced from CCTV cameras. In many works, the focus is only on detecting vehicle classes such as motorcycles, buses, trucks, and cars or special vehicles such as ambulances and others. In this research, we propose to apply 13 classes of vehicle types and implement YOLOv4 in the traffic surveillance task. More specifically, all classes are labeled, and then the YOLOv4 model is trained on 800 images and tested on 23 videos from three intersections in Medan City, namely Juanda Katamso Intersection, Gatot Subroto Intersection, and Uniland Intersection. Based on the test results, YOLOv4 proves successful in detecting many vehicles in frame-by-frame sequence with various types of vehicles. All vehicle detection data will be stored in the file.

Development of a real-time tritium-in-water monitor

In this paper, we report the development and performance of a detector module envisaging a tritium-in-water real-time activity monitor. The monitor is based on modular detection units whose number can be chosen according to the required sensitivity. The full system is being designed to achieve a Minimum Detectable Activity (MDA) of 100 Bq/L of tritium-in-water activity which is the limit established by the E.U. Council Directive 2013/51/Euratom for water intended for human consumption. The same system can be used as a real-time pre-alert system for nuclear power plant regarding tritium-in water environmental surveillance. The first detector module was characterized, commissioned and installed immediately after the discharge channel of the Arrocampo dam (Almaraz nuclear power plant, Spain) on the Tagus river. Due to the high sensitivity of the single detection modules, the system requires radioactive background mitigation techniques through the use of active and passive shielding. We have extrapolated a MDA of 3.6 kBq/L for a single module being this value limited by the cosmic background. The obtained value for a single module is already compatible with a real-time environmental surveillance and pre-alert system. Further optimization of the single-module sensitivity will imply the reduction of the number of modules and the cost of the detector system.

Comprehensive Evaluation of Healthcare Associated Infection Surveillance System

Background & objectives: The electronic surveillance system has been applied in healthcare associated infection surveillance. We conducted an evaluation of the real-time healthcare associated infection surveillance system (RT-HAISS) to understand the early warning effect. Methods: We evaluated our RT-HAISS on a dataset of 29074 patients at the Wuxi traditional Chinese medicine (TCM) hospital in 2020, encompassing sensitivity, specificity, positive predictive value, negative predictive value, Youden index and false alarm rate.Results: 466 HAIs were confirmed in this hospital in 2020, The RT-HAISS warned the monitors a total of 1715 cases, with 2040 early warning entries and 1736 false alarm entries. The sensitivity and the specificity were 65.24% and 95.74%, respectively. The Youden index was 0.62. In addition, the positive predictive value was 14.90%, and the negative predictive value was 83.86%, with 85.10% of false alarm rate.Conclusion: The RT-HAISS is an important technical mean for healthcare associated infection surveillance, it’s necessary to further explore the improvement of real-time surveillance system.

2077. Utility of a Real-Time Spatiotemporal Mapping Surveillance System in Detection of Healthcare-Associated Acute Respiratory Viral Infection Clusters in a Tertiary Healthcare Institution

Abstract Background Transition to endemic COVID-19 has been associated with a rise in community respiratory viral infections (ARIs) with a corresponding increase in healthcare-associated ARIs (HA-ARIs) (Figure 1). 4D-Disease Outbreak Surveillance System (4D-DOSS) is a real-time spatiotemporal mapping surveillance system being developed to detect healthcare-associated infection clusters. We aimed to assess 4D-DOSS’s utility in detection of HA-ARI clusters. Methods 4D-DOSS is a system that integrates and maps clinical, laboratory and patient movement data onto a digital twin of the hospital’s physical space. In addition to a virtual mapping replica of Singapore General Hospital, a 2000-bedded tertiary hospital in Singapore, it constitutes detailed healthcare cloud architecture and surveillance algorithms. Respiratory specimens from inpatients with ARI symptoms are tested for 16 human respiratory viral pathogens via a respiratory virus multiplex PCR (RV16) panel. HA-ARI constitutes first positive sample beyond the maximum incubation period of the corresponding virus (from admission date). Earlier positive test is categorized as community-associated ARI (CA-ARI). Two or more patients with spatial temporal overlap of three-days or less, during the infectious period (7-days) of an index patient are deemed a HA-ARI cluster. Results Incidence of HA-ARI, as per 10,000 patient-days was 7.4 pre-COVID-19 pandemic (Jan 2018 to Dec 2019), 1.5 during pandemic (Jan 2020 to Dec 2022) and 6.4 during transition to endemicity (Jan 2023 – April 2023). Between September 2018 and December 2018, one influenza cluster of ten inpatients (Nov 2018) was identified in the proof-of-concept version of 4D-DOSS. There were four HA-ARI clusters during the Jan 2020-Dec 2022 COVID-19 pandemic phase. 19 HA-ARI clusters were identified between Jan 2023-April 2023. Conclusion 4D-DOSS can detect HA-ARI clusters and has potential to trigger an alert-response process for more effective infection prevention. It enables study of infectious disease transmission kinetics in real-time. Disclosures Sean Whiteley, BSc (IT), Axomem: Board Member|Axomem: Fees received for service and software licenses Maybelle Auw, MBBS, Axomem: Work for company that received fees for service and licensing

8.B. Workshop: Fit for purpose surveillance - strategies to implement effective surveillance systems in EU/EEA countries under the new Serious Cross Border Health Threats Regulation (2022/2371)

Abstract Background As part of the European Health Union package, the EU Regulation 2022/2371 on serious cross-border threats to health has entered into force on 26 December 2022. Chapter III (Articles 13-17) of this Regulation provide new provisions for Epidemiological surveillance, the Digital platform for surveillance, and EU reference laboratories. Methods Several Implementing Acts are currently in preparation, which will define in more detail the concrete implementation of these provisions, such as the list of communicable diseases, case definitions, procedures for the operation of the epidemiological network for epidemiological surveillance, the specifications of the digital platform for surveillance and its specific rules for functioning, as well as to designate EU reference laboratories. Via the EU4Health Programme, the European Commission is furthermore providing funding for Integrated Surveillance, Surveillance capacity building Results The workshop will inform on the current state of implementation, on the results of the preparatory works carried out so far as well as on an outlook to the short and longer term plans for the implementation of the regulation on serious cross-border threats to health. Furthermore, it will inform on a funding mechanism for the improvement of EU Surveillance Systems as one of the central lessons learnt from the COVID-19 pandemic and showcase a good national practice. Conclusions The COVID-19 pandemic has clearly illustrated that the creation and further development of sound, reliable, integrated and real-time digital surveillance systems are a very crucial component for preparedness against serious cross border health threats. The current work under the new regulation is a valuable opportunity for public health practitioners to improve surveillance at EU, national and regional level. The workshop will host panel presentations followed by a round-table discussion.

A real-time surveillance system with multi-object tracking

A real-time surveillance system with multi-object tracking

Real-Time Surveillance of Dog Bite Incidence in Islamabad: A Cross-Sectional Study from December 2019 to July 2020

Rabies is a neglected viral zoonotic disease with an almost 100% fatality rate. A pilot project was initiated by the National Institute of Health (NIH), Islamabad, in collaboration with Health Security Partners (HSP) to establish a real-time dog bite surveillance in humans in public sector secondary care hospitals of Islamabad. The main objective of this study was to analyze different characteristics of dog bite cases, identify gaps in surveillance, and recommend suggestions to improve and strengthen real-time dog bite surveillance systems. An observational cross-sectional study was conducted in two major public sector hospitals in Islamabad from December 2019 to July 2020. Data on demographic information, site of the dog bite, category of the dog bite, and treatment was collected via a WVS mobile-based application. Descriptive analysis was carried out for different variables at a p-value of <0.05. A total of 338 dog bite cases were captured in the World Veterinary Service (WVS) application, and most cases (n = 226, 85.6%) were reported in December 2019 with a male-to-female ratio of 2:1. Most cases were reported from the age group of 22 to 31 years (n = 178, 53%). Out of the total number of cases, 263 (78%) presented with a single anatomical location, and 174 (51%) had Category II wounds. All dog bite patients were given only the first dose of the anti-rabies vaccine at the hospital level. Real-time dog bite surveillance via a mobile-based application proved to be effective for the timely recording and management of dog bite cases. Young people were reported to be mostly affected by dog bites, and nearly half of the cases were managed in hospitals. Refresher training was conducted for medics and paramedics for mobile-based applications, dog bite management, and proper referral of cases to tertiary care hospitals.

Object detection and activity recognition in video surveillance using neural networks

PurposeThis paper aims to implement and extend the You Only Live Once (YOLO) algorithm for detection of objects and activities. The advantage of YOLO is that it only runs a neural network once to detect the objects in an image, which is why it is powerful and fast. Cameras are found at many different crossroads and locations, but video processing of the feed through an object detection algorithm allows determining and tracking what is captured. Video Surveillance has many applications such as Car Tracking and tracking of people related to crime prevention. This paper provides exhaustive comparison between the existing methods and proposed method. Proposed method is found to have highest object detection accuracy.Design/methodology/approachThe goal of this research is to develop a deep learning framework to automate the task of analyzing video footage through object detection in images. This framework processes video feed or image frames from CCTV, webcam or a DroidCam, which allows the camera in a mobile phone to be used as a webcam for a laptop. The object detection algorithm, with its model trained on a large data set of images, is able to load in each image given as an input, process the image and determine the categories of the matching objects that it finds. As a proof of concept, this research demonstrates the algorithm on images of several different objects. This research implements and extends the YOLO algorithm for detection of objects and activities. The advantage of YOLO is that it only runs a neural network once to detect the objects in an image, which is why it is powerful and fast. Cameras are found at many different crossroads and locations, but video processing of the feed through an object detection algorithm allows determining and tracking what is captured. For video surveillance of traffic cameras, this has many applications, such as car tracking and person tracking for crime prevention. In this research, the implemented algorithm with the proposed methodology is compared against several different prior existing methods in literature. The proposed method was found to have the highest object detection accuracy for object detection and activity recognition, better than other existing methods.FindingsThe results indicate that the proposed deep learning–based model can be implemented in real-time for object detection and activity recognition. The added features of car crash detection, fall detection and social distancing detection can be used to implement a real-time video surveillance system that can help save lives and protect people. Such a real-time video surveillance system could be installed at street and traffic cameras and in CCTV systems. When this system would detect a car crash or a fatal human or pedestrian fall with injury, it can be programmed to send automatic messages to the nearest local police, emergency and fire stations. When this system would detect a social distancing violation, it can be programmed to inform the local authorities or sound an alarm with a warning message to alert the public to maintain their distance and avoid spreading their aerosol particles that may cause the spread of viruses, including the COVID-19 virus.Originality/valueThis paper proposes an improved and augmented version of the YOLOv3 model that has been extended to perform activity recognition, such as car crash detection, human fall detection and social distancing detection. The proposed model is based on a deep learning convolutional neural network model used to detect objects in images. The model is trained using the widely used and publicly available Common Objects in Context data set. The proposed model, being an extension of YOLO, can be implemented for real-time object and activity recognition. The proposed model had higher accuracies for both large-scale and all-scale object detection. This proposed model also exceeded all the other previous methods that were compared in extending and augmenting the object detection to activity recognition. The proposed model resulted in the highest accuracy for car crash detection, fall detection and social distancing detection.

Nosocomial infections in adult patients receiving extracorporeal membrane oxygenation in China: A retrospective cohort study

Extracorporeal membrane oxygenation (ECMO) has been increasingly used in China, but nosocomial infections (NI) in patients receiving ECMO remain poorly characterized. Thus, this study aimed to investigate the incidence rate, causative was pathogens, and risk factors of NIs in ECMO patients. A retrospective cohort study of patients receiving ECMO between January 2015 and October 2021 was conducted in a tertiary hospital. General demographics and clinical data of the included patients were collected from the electronic medical record system and the real-time NI surveillance system. A total of 86 infected patients with 110 episodes of NIs were identified in 196 patients receiving ECMO. The incidence of NI was 59.2/1000 ECMO days. The median time for the first NI in ECMO patients was 5days (interquartile range: 2-8days). Hospital-acquired pneumonia and bloodstream infections were common types of NIs in ECMO patients, and the main pathogens were gram-negative bacteria. Pre-ECMO invasive mechanical ventilation (OR=2.40, 95%CI:1.12-5.15) and prolonged duration of ECMO (OR=1.26, 95%CI:1.15-1.39) were risk factors for NIs during ECMO support. This study identified the main infection sites and pathogens of NIs in ECMO patients. Although NIs may not affect successful ECMO weaning, additional measures should be implemented to reduce the incidence of NI during ECMO support.

Vaccine intention and hesitancy among Australian women who are currently pregnant or have recently given birth: the Birth in the Time of COVID-19 (BITTOC) national online survey

ObjectiveTo examine the prevalence of COVID-19 vaccination, and factors associated with vaccination intention and hesitancy in pregnant and postnatal women in Australia.Design and settingA national online survey was conducted over...

Language and vision based person re-identification for surveillance systems using deep learning with LIP layers

Real-time surveillance systems have become a necessity of today's life owing to their relevance in the contemporary era for security reasons to ensure a secure and safe environment. Presently, Person re-identification (Re-ID)-based surveillance systems are becoming increasingly more prevalent and sophisticated since they do not require human intervention and are more reliable to deploy in public spaces leveraging multi-camera networks. However, one of the major problems in Person ReID is the visual appearance i-e the appearance of a person in an image is greatly affected by different camera views. As a result, the discriminative set of features must be learned in a deep learning model in order to re-identify persons from opposing camera viewpoints. To address this challenge, we propose an image/text-retrieval-based Person ReId method in which both visual and text-based features are exploited to carry out person re-identification. More precisely, the textual descriptions of the images are taken into account as text features with Glove Word Embedding followed by 1D-MAPCNN and fused with image-level features extracted using the GoogLeNet model. In addition, the feature discriminability is enhanced using local importance-based pooling (LIP) layers in which adaptive significance weights are learned during downsampling. Moreover, from two different modalities, feature refinement is done during training with the help of attention mechanisms using the Convolutional Block Attention module (CBAM) and the proposed shared attention neural network. It is observed that LIP layers along with both vision and textual features are playing a key role in acquiring discriminative features even if the visual appearance of the same person is greatly affected due to camera pose conditions. The proposed method is validated on the CUHK-PADES dataset and has 15.34% and 24.39% rank-1 improvement in text and image-based retrievals.

Real-Time Tunnel Abnormal Sound Detection Algorithm Using Convolutional Neural Networks

In the traffic industry, the automatic accident detection system is a major concern. Although image-based and radar-based traffic accident detection systems are commonly employed, they have several drawbacks, including the need to secure the camera’s field of view, a high rate of false alarms, and a lengthy detection time. Using a real-time acoustic surveillance system and the classification algorithm via Convolutional Neural Network (CNN), this article proposes several methods for identifying abnormal situations, such as a car crash or tire skid sound, to overcome the limitations of existing methods. We create an audio database by collecting sounds from two tunnels in South Korea using self-made microphones for eight months and classifying them into three categories: car crash, tire skid, and normal environmental sounds. We establish a three-step classification procedure using an algorithm. We compare the detection rate and false alarm rate of our proposed method to those of deep learning techniques including MLP (Multi-Layer Perceptron), Long-Short Term Memory, ShuffleNetv2, and MobileNetv2. In addition, we present a method for filtering out irrelevant sound data to improve the computational efficiency of our approach.

An empirical analysis of stateful operator migration for online scheduling in distributed stream processing systems

Streaming data is the data from sensors as well as other real-time surveillance systems. Distributed stream processing systems are the software that manages such data. Such frameworks have to deliver outcomes on the go instantly. They are susceptible to delay and malfunction or system failures. The system must be tolerant of faults and always accessible. Many variables, such as improved network arrival rates, node failures, and so on, disrupt the system's reliability. Some operators need to be relocated online from one physical resource to another to manage or reimburse a slow or failing node. In this study, we propose a co-location based systematic migration heuristic for live operator migration between physical resources using a migration map revised with costs for each migration. The suggested method evaluates continuous operator performance patterns and makes online scheduling decisions based on the same. The decisions include migrating operators during a node failure or straggling.

Syndromic surveillance system during mass gathering of Panchkroshi Yatra festival, Ujjain, Madhya Pradesh, India

BackgroundThe health implications surrounding a mass gathering pose significant challenges to public health officials. The use of syndromic surveillance provides an ideal method for achieving the public health goals and objectives at such events. In the absence of published reports of systematic documentation of public health preparedness in mass gatherings in the local context, we describe the public health preparedness and demonstrate the operational feasibility of a tablet-based participatory syndromic surveillance among pilgrims during the annual ritual circumambulation- Panchkroshi Yatra. MethodsA real-time surveillance system was established from 2017-2019 to capture all the health consultations done at the designated points (medical camps) in the Panchkroshi yatra area of the city Ujjain in Madhya Pradesh. We also surveyed a subset of pilgrims in 2017 to gauge satisfaction with the public health measures such as sanitation, water, safety, food, and cleanliness. ResultsIn 2019, injuries were reported in the highest proportion (16.7%; 794/4744); most numbers of fever cases (10.6%; 598/5600) were reported in 2018, while 2017 saw the highest number of patient presentations of abdominal pain (7.73%; 498/6435). ConclusionPublic health and safety measures were satisfactory except for the need for setting up urinals along the fixed route of the circumambulation. A systematic data collection of selected symptoms among yatris and their surveillance through tablet could be established during the panchkroshi yatra, which can complement the existing surveillance for detecting early warning signals. We recommend the implementation of such tablet-based surveillance during such mass gathering events.

Monitoring Indonesian online news for COVID-19 event detection using deep learning

Even though coronavirus disease 2019 (COVID-19) vaccination has been done, preparedness for the possibility of the next outbreak wave is still needed with new mutations and virus variants. A near real-time surveillance system is required to provide the stakeholders, especially the public, to act in a timely response. Due to the hierarchical structure, epidemic reporting is usually slow particularly when passing jurisdictional borders. This condition could lead to time gaps for public awareness of new and emerging events of infectious diseases. Online news is a potential source for COVID-19 monitoring because it reports almost every infectious disease incident globally. However, the news does not report only about COVID-19 events, but also various information related to COVID-19 topics such as the economic impact, health tips, and others. We developed a framework for online news monitoring and applied sentence classification for news titles using deep learning to distinguish between COVID-19 events and non-event news. The classification results showed that the fine-tuned bidirectional encoder representations from transformers (BERT) trained with Bahasa Indonesia achieved the highest performance (accuracy: 95.16%, precision: 94.71%, recall: 94.32%, F1-score: 94.51%). Interestingly, our framework was able to identify news that reports the new COVID strain from the United Kingdom (UK) as an event news, 13 days before the Indonesian officials closed the border for foreigners.

Police-led real-time surveillance system for suspected suicides in Great Britain

It has become increasingly apparent that rapidly available information about the occurrence of suicides is needed, particularly to support suicide prevention efforts. Concerns about the potential impact of the COVID-19...