CCTV Operators Research Articles

Channelizing Machine Learning Towards Early Threat Detection and Prevention Against Women Using Surveillance Cameras

Abstract: Closed Circuit Television Systems (CCTV) are being used in a growing number of office buildings, housing developments, and public locations. Many towns now have CCTV surveillance systems in place. Such monitoring systems create a tremendous load for the CCTV operators because the number of cameras views a single CCTV operator can process is constrained by human considerations. Surveillance automation of public places assumes an important role in proactively detection of possible threat to public and in maintaining law and order. Based on a review of the existing approaches followed in monitoring of crowd behaviours and the techniques applied, especially in public places like bus stands, railway stations and airports, the paper proposes surveillance automation i.e., automating the process of detecting, recognizing the suspects and suspicious behaviour of the people in the crowd. Here, in this project called channelizing machine learning towards early threat detection and prevention against women using surveillance cameras, we propose algorithms that are capable of alerting the human operator when: (1) Presence of a inhuman act, after setting zones of interest and danger zones within those zones of interest, the danger is detected when someone follows a women on streets, roads or any other public places, which can reduce the number of sexual harassment and other forms of sexual Algorithms violence in public spaces; (2) An abnormal behaviour of a person such as handling some objects abnormally; (3) In this project, our main aim is towards women safety and also we concentrate on reducing the amount of false alerts and to enable a real-life use of the system.

Proactive monitoring and operator discretion: A systematic social observation of CCTV control room operations

BackgroundResearch on police discretion indicates a host of legal and extra-legal factors can impact officer decision making. The emergence of video surveillance technologies has made certain police interactions with the public more remote in nature. Decisions to enforce the law, consequently, now commonly begin outside of the context of face-to-face police/citizen interactions. This study explores police decision-making in the context of proactive video surveillance operations. MethodsData were generated from a systematic social observation of CCTV operator activity during the Newark CCTV Directed Patrol Experiment. Regression models tested how covariates affect the duration of CCTV operator targeted surveillances and CCTV operator decisions to report incidents providing reasonable suspicion and/or probable cause for a police response. ResultsA range of model covariates significantly influenced CCTV operator decision making, inclusive of surveillance targets identified as prior suspects, the CCTV site being within a commercial area, visible obstructions impeding camera view, CCTV operator rank, and CCTV operator gender. ConclusionsFindings suggest that organizational culture, CCTV operator characteristics, and land usage of target areas may foster differential surveillance behavior across CCTV operators. As remote strategies for policing continue to expand internationally, the identification of factors that impact discretionary practices is critical.

Risk Management for Analytic Closed Circuit Television based on Internet of Things in Semarang City

Successful implementation of a program is the goal of program design. However, currently, many programs fail to be implemented due to the failure of efforts to prevent the occurrence of failure factors. Especially in the era of digital governance such as today, the planning of various smart city programs has a large potential risk of implementation failure. The Semarang City of IoT CCTV program, for example, the implementation of the use of analytics-based CCTV requires long and careful preparation. Not only management preparation but risk management is also needed for the program. This research aims to analyze the map of potential risks in the Semarang City CCTV of IoT program along with analyzing how the risk management of the program. This research is based on qualitative research with primary and secondary data collection techniques. Primary data collection by interviewing related informants purposively, then the data is analyzed based on qualitative analysis. The findings of this research are that the highest level of risk occurs in the shortage of human resources and CCTV operators because there are no human resources who are data analytic experts and Artificial Intelligence experts. Other high risks are database errors / down; CCTV devices using old technology; long loading dashboards; and no Fiber Optic connection in internet management

CCTV operators’ perspectives on protecting soft target terrorist locations in England

CCTV operators’ perspectives on protecting soft target terrorist locations in England

Weapon Detection using Yolov4, CNN

Abstract: A Weapon Detection software is a very efficient way to monitor the streets and alert the operator only when there is a situation where people carrying a weapon like a gun or knife can be detected and analyzed with the help of Yolov4 object detection algorithm. And the notification of the same will be given via email to the CCTV operator with the photo where people carrying a gun or knife can be seen and also the location of the CCTV. So this will enhance the safety and security of any city with the requirement of limited man power. The proposed paper focuses on different approaches and algorithms that could potentially help in Weapon detection. Keywords: Weapon Detection, Yolov4, Google Colab, CNN

Surveillance in Practice: Operators’ Collective Interpretation of CCTV Images

CCTV surveillance is a cultural practice and collective effort. CCTV not only involves a technical assemblage that is used to discipline the surveilled, it is also a social assemblage in which the informal practices of operators play a major role in the multiple interpretations of images. This paper provides insights into the daily work practices and discourses of CCTV operators and their supervisors through observations of and interviews in the control room of public CCTV surveillance in Rotterdam. By providing a better understanding of the role of people in socio-technical assemblages, this paper contributes to the discussion on human mediation in computerized networks. The paper contributes to the expanding literature on surveillance as a cultural practice by combining insights on social sorting with insights on collective evaluation of unfolding situations—i.e., how group dynamics within the control room influence how people are “judged.” Building on Goffman’s frame analysis, the paper reveals the crucial role of talk and humor in re-performing what happens on the streets as well as evaluating situations and the people watched. Moreover, it discusses how these collective re-performances of what is being watched both reproduce and reshape “othering” practices within the control room. The paper shows how humorous utterances play an important part in overcoming hierarchy and collectively managing emotions, and explores how this humor influences profiling on the basis of bodily appearance.

Firearm Recognition Using Convolutional Neural Network

Closed circuit television systems (CCTV) are becoming more popular and are being deployed in many offices, housing estates and in the most public spaces. Monitoring systems have been implemented in many foreign cities. This makes for an enormous load for the CCTV operators, as the number of camera views a single operator can monitor is limited by the human factors. The projects focus on the task of automated detection and recognition of dangerous situations for CCTV systems. We propose algorithms that are able to alert the human operator when a firearm is visible in the image and also have focussed on limiting the number of false alarms, in order to allow a real life application of the system. Managed to propose a version of a firearm detection algorithm that offers a near zero rate of false alarms and have shown that it is possible to create a system that are capable of an early warning in a dangerous situation, which may lead to faster and more effective response times and a reduction in the number of potential victims.

Jonathan Murray, The New Scottish Cinema

The cover of Jonathan Murray’s most recent book, examining the development of Scottish cinema from the 1990s to the 2000s, features a striking image from Red Road (Andrea Arnold, 2006). The actor Kate Dickie, who plays the lead character of Jackie, is rendered in dark profile against a bright yellow background, her identity obscured by the high-contrast lighting. The film, an apt choice for the book’s cover, was produced as part of the Advance Party initiative, led by Glasgow-based Sigma Films, and the Danish production company Zentropa, and is one of a number of films produced in Scotland in the 2000s that Murray describes as having demonstrated ‘the extent to which traditional ideas of a deliberate, dominant and didactic focus on the question of nation are not straightforwardly applicable to a significant proportion of contemporary Scottish filmmaking’ (p. 175). The image from Red Road was the key image used for the worldwide promotion of the film, serving to give a sense of the film’s narrative focus and Jackie’s literal position in the dark in her daily work as a CCTV operator, but also acting as an indicator of the film’s deliberate positioning as a transnational film. The image is notable for its lack of obvious identifiers with its Scottish context. The identity obscured is not only Jackie’s but Scotland’s; the film’s location, as its narrative image suggests, could be anywhere.1

The importance of gaze coherence of CCTV operators in facilitating the ability to recognise harmful intentions

Experienced CCTV operators are better at predicting violent behaviour than novices. It has been suggested that experienced, trained operators acquire gaze strategies to process complex visual scenes more efficiently. We examined whether the coherence of the gaze path of an operator aids in perceiving harmful intent and how this compares to the situation when coherence is altered but spatial location information is retained. To achieve this, we obtained 18 real-life surveillance videos, 16-seconds each, grouped in three categories ('fight', 'confrontation', 'play'). 'Fight' and 'confrontation' videos included aggressive behaviour, but only 'fight' videos resulted in a violent outcome (after the video ended). 'Play' videos portrayed playful behaviour without the involvement of violence. For each video, we used eye-movement data from experienced CCTV operators who successfully predicated harmful intent above chance. Naïve participants then viewed a filtered version of each video by overlaying a high resolution foveal window with a blurred periphery. Videos were filtered according to an intact gaze path from one operator or shuffled gaze locations from two operators. The shuffled gaze path was computed by alternating the two operator's eye-coordinates every 0.33 seconds. Participants viewed processed videos in both intact and shuffled conditions in two blocks. For each video, they rated the likelihood of a violent incident occurring after the video ended based on a 6-point scale. Preliminary results indicate that ratings of violent outcome are significantly greater when viewing the videos processed with the intact gaze patterns than with the shuffled gaze patterns, particularly in 'fight' and 'play' videos. This suggests that the gaze path coherence aids in perception of intent and that the spatial locations of fixations are less informative in guiding performance. Therefore, allowing novices to follow the learned gaze strategies of operators may provide effective training in improving their ability to recognise harmful intentions. Meeting abstract presented at VSS 2017

CCTV as a tool for early police intervention: Preliminary lessons from nine case studies

This study explores the prospect of utilizing CCTV as an early intervention mechanism to detect and disrupt street-level activity that can lead to violence. The analysis focuses on nine case studies in Newark, NJ, incorporating data from several sources, including video footage, computer-aided dispatch (CAD) system incident data, CAD event chronologies and face-to-face interviews with CCTV operators. The findings suggest that the benefits offered by CCTV, namely the instantaneous discovery and reporting of crime, may be rendered inconsequential by the process times associated with the differential-response policy of police dispatch. Potential methods by which police can more proactively utilize CCTV to prevent crime are discussed.

A comparison of eye-movement patterns between experienced observers and novices in detecting harmful intention from surveillance video

Understanding the intentions of others by viewing their actions in a complex visual scene is a challenging task. Does experience change looking behavior in that gaze to specific action cues creates a fixation unique to experienced observers? To address this question we analyzed eye movements between experienced surveillance (CCTV) operators and novices when observing social interactions from CCTV footage. 11 experienced operators and 10 novices observed 36 unique CCTV clips, while obtaining point-of-gaze coordinates from an eye tracker. Each clip was 16 seconds in duration, and classified based on one of four contexts. 'Fight' and 'Confront' clips displayed aggressive behavior, except that fighting actually occurred for 'Fight' clips after the end of the clip (not seen by the participants). 'Play' clips showed playful interactions between people with no aggression, while 'Nothing' clips showed typical everyday behavior with no aggression. We used a sequence matching algorithm to compare eye-movements (scan paths) between individuals examining the same video. We split sequences into segments (first, middle, and last) to analyze how viewing behavior changes with the accumulation of visual input over time. We found that on average, experienced operators yielded higher fixation similarity scores suggesting in-group consistency among experienced observers, with the consistency varying by context and time segment. For both the 'Confront' and 'Fight' contexts, differences in gaze patterns between experienced observers and novices were largest for the middle segment, whereas for the 'Play' context, the largest difference was at the end. The 'Nothing' context yielded higher scores for experts but no time differences. These results suggest that expert CCTV operators predict actions similarly by attending to the relevant events within a scene, and for critical events such as aggressive behaviors, important cues (markers that result in signature looking patterns) are attended well before the onset of the disruptive action.

Recognizing harmful intent from surveillance video viewed through the eye-movements of novice and experienced observers

To discern intention, based on the observation of complex human behaviour, typically requires eye-movements that actively explore the visual context. We examined whether the gaze patterns of novice and experienced observers would differentially inform intention judgments of naive viewers. To achieve this we first obtained real-life surveillance (CCTV) videos and organised them into 24 videos of 16-second duration in four categories ('fight', 'confront', 'nothing', 'play'). 'Fight' and 'confront' included aggressive behaviour, but only 'fight' videos resulted in a violent incident (after the video ended). 'Play' videos showed playful behaviour and the 'nothing' videos showed a variety of everyday activity that did not include aggressive behavior. We then obtained eye-movement data from novice and experienced CCTV operators who viewed these videos with the goal of judging hostile intent. Next, a high resolution foveal window with a blurred periphery was overlaid on each video according to the gaze coordinates of a novice or experienced operator. Twenty-four of these videos were created and shown to naive participants, with half from each category being processed through an operator's eye-movements and half through a novice's eye-movements. Participants were instructed to follow the foveated area and try to ignore the blurred surrounding. For each trial they made a Yes/No judgement of whether a violent incident occurred after the video ended, and rated their confidence on a 7-point scale. Signal detection analysis was carried out on the Yes/No responses with a 'yes' on the 'fight' videos coded as a hit, and on any other video as a false alarm. Results showed no significant differences in sensitivity or bias. Results of confidence judgments revealed a trend of increased confidence when following novices' eye-movements. These preliminary results suggest that naive participants were not able to extract more information from the operator scanpaths but subjectively found the experience different.

Automated Detection of Firearms and Knives in a CCTV Image.

Closed circuit television systems (CCTV) are becoming more and more popular and are being deployed in many offices, housing estates and in most public spaces. Monitoring systems have been implemented in many European and American cities. This makes for an enormous load for the CCTV operators, as the number of camera views a single operator can monitor is limited by human factors. In this paper, we focus on the task of automated detection and recognition of dangerous situations for CCTV systems. We propose algorithms that are able to alert the human operator when a firearm or knife is visible in the image. We have focused on limiting the number of false alarms in order to allow for a real-life application of the system. The specificity and sensitivity of the knife detection are significantly better than others published recently. We have also managed to propose a version of a firearm detection algorithm that offers a near-zero rate of false alarms. We have shown that it is possible to create a system that is capable of an early warning in a dangerous situation, which may lead to faster and more effective response times and a reduction in the number of potential victims.

Neuroscience perspectives on security

Neuroscience perspectives on security

CCTV and crime displacement: A quasi-experimental evaluation

The installation of CCTV cameras in the historic centre of Malaga in March 2007 was the main crime prevention initiative implemented in the city during the past few years. This was one of the first multi-camera CCTV installations on streets in Spain, which is why a proper evaluation of its effectiveness is so important for the future use of such cameras. Using a quasi-experimental design with a pre/post test, we collected data from interviews with CCTV operators, police officers and local authority officials and from surveys of pedestrians and shopkeepers. The team also examined police crime data and CCTV incident data. In this paper we will discuss the results in terms of the following three hypotheses relating to crime reduction, displacement and public security: (a) the use of cameras reduces the levels of crime, whether property crime (robberies and burglaries), crimes against people or both; (b) some of those crimes are displaced to nearby areas within or close to the city centre where there is no camera coverage but where there are similar opportunities to commit crimes; and (c) people claim to feel safer in the city centre after dark after the cameras were introduced.

Looking for trouble: a description of oculomotor search strategies during live CCTV operation

Recent research has begun to address how CCTV operators in the modern control room attempt to search for crime (e.g., Howard et al., 2011). However, an often-neglected element of the CCTV task is that the operators have at their disposal a multiplexed wall of scenes, and a single spot-monitor on which they can select any of these feeds for inspection. Here we examined how 2 trained CCTV operators used these sources of information to search from crime during a morning, afternoon, and night-time shift. We found that they spent surprisingly little time viewing the multiplex wall, instead preferentially spending most of their time searching on the single-scene spot-monitor. Such search must require a sophisticated understanding of the surveilled environment, as the operators must make their selection of which screen to view based on their prediction of where crime is likely to occur. This seems to be reflected in the difference in the screens that they selected to view at different times of the day. For example, night-clubs received close monitoring at night, but were seldom viewed in mid-morning. Such narrowing of search based on a contextual understanding of an environment is not a new idea (e.g., Torralba et al., 2006), and appears to contribute to operator's selection strategy. This research prompts new questions regarding the nature of representation that operators have of their environment, and how they might develop expectation-based search strategies to countermand the demands of the large influx of visual information. Future research should ensure not to neglect examination of operator behavior “in the wild” (Hutchins, 1995a), as such insights are difficult to gain from laboratory based paradigms alone.

S1-4: Experience Dependent Differences in Brain Mechanisms of Action Observation: From Watching Dance to CCTV Surveillance

Our recent work has focused on using fMRI to investigate how experience in observing action is reflected in brain activity. In this talk I will concentrate on two studies that span a range of experience. The first study investigates brain activity of inexperienced observers when they view a complex 6-minute dance to which they have no familiarity. Comparison of brain activity to both a novel motion index and a measure of human event segmentation revealed a right hemisphere system including the inferior frontal gyrus and occipitotemporal cortex. This neural system can be interpreted as providing an ability to use motion and information about action hierarchies to parse unfamiliar streams of human activity. The second study contrasts the brain activity of naive observers to experienced CCTV operators when they view selected video clips of real street activity while predicting whether physical violence ensues after completion of the video clip. Comparison of the brain activity between groups revealed a decre...

Semi-automated CCTV surveillance: The effects of system confidence, system accuracy and task complexity on operator vigilance, reliance and workload

Recent advances in computer vision technology have lead to the development of various automatic surveillance systems, however their effectiveness is adversely affected by many factors and they are not completely reliable. This study investigated the potential of a semi-automated surveillance system to reduce CCTV operator workload in both detection and tracking activities. A further focus of interest was the degree of user reliance on the automated system. A simulated prototype was developed which mimicked an automated system that provided different levels of system confidence information. Dependent variable measures were taken for secondary task performance, reliance and subjective workload. When the automatic component of a semi-automatic CCTV surveillance system provided reliable system confidence information to operators, workload significantly decreased and spare mental capacity significantly increased. Providing feedback about system confidence and accuracy appears to be one important way of making the status of the automated component of the surveillance system more ‘visible’ to users and hence more effective to use.

Unblinking eyes: the ethics of automating surveillance

In this paper I critique the ethical implications of automating CCTV surveillance. I consider three modes of CCTV with respect to automation: manual (or non-automated), fully automated, and partially automated. In each of these I examine concerns posed by processing capacity, prejudice towards and profiling of surveilled subjects, and false positives and false negatives. While it might seem as if fully automated surveillance is an improvement over the manual alternative in these areas, I demonstrate that this is not necessarily the case. In preference to the extremes I argue in favour of partial automation in which the system integrates a human CCTV operator with some level of automation. To assess the degree to which such a system should be automated I draw on the further issues of privacy and distance. Here I argue that the privacy of the surveilled subject can benefit from automation, while the distance between the surveilled subject and the CCTV operator introduced by automation can have both positive and negative effects. I conclude that in at least the majority of cases more automation is preferable to less within a partially automated system where this does not impinge on efficacy.

The importance of perceiving social contexts when predicting crime and antisocial behaviour in CCTV images

Purpose. The aim of this study was to establish which visual cues are used within CCTV images to accurately predict a criminal act and to investigate observer descriptions of the meaning of such cues. This study also tested whether professional CCTV surveillance experience influenced the ability to predict antisocial acts using real CCTV footage. Method. CCTV operators (N = 12) and inexperienced observers (N = 12) were shown 24 short scenes in which half led to an antisocial act. An eye‐tracker was used to identify the visual cues attended whilst participants predicted whether or not each scene led to criminal behaviour. Participants also verbalized their impressions of selected scenes and reported any features arousing suspicion. Results. More accurate predictions of impending criminal action were associated with looking at the face/head of lone individuals and the bodies of individuals engaged in reciprocal social interaction or communication. Qualitative data revealed that accurate observers attended to the spatial configuration of individuals within the scene and were able to appropriately identify distinct social entities (i.e., subgroups of people that appear to belong together within the scene as a whole). Conclusion. The perception of social entities sensitizes successful observers to violations of normative conduct. Thus, better prediction accuracy in CCTV images appears to be associated with a more sophisticated awareness of the social context present in the scene.