Automated Surveillance System Research Articles

HECOL: Homography and epipolar-based consistent labeling for outdoor park surveillance

Outdoor surveillance is one of the most attractive application of video processing and analysis. Robust algorithms must be defined and tuned to cope with the non-idealities of outdoor scenes. For instance, in a public park, an automatic video surveillance system must discriminate between shadows, reflections, waving trees, people standing still or moving, and other objects. Visual knowledge coming from multiple cameras can disambiguate cluttered and occluded targets by providing a continuous consistent labeling of tracked objects among the different views. This work proposes a new approach for coping with this problem in multi-camera systems with overlapped Fields of View (FoVs). The presence of overlapped zones allows the definition of a geometry-based approach to reconstruct correspondences between FoVs, using only homography and epipolar lines (hereinafter HECOL: Homography and Epipolar-based COnsistent Labeling) computed automatically with a training phase. We also propose a complete system that provides segmentation and tracking of people in each camera module. Segmentation is performed by means of the SAKBOT (Statistical and Knowledge Based Object Tracker) approach, suitably modified to cope with multi-modal backgrounds, reflections and other artefacts, typical of outdoor scenes. The extracted objects are tracked using a statistical appearance model robust against occlusions and segmentation errors. The main novelty of this paper is the approach to consistent labeling. A specific Camera Transition Graph is adopted to efficiently select the possible correspondence hypotheses between labels. A Bayesian MAP optimization assigns consistent labels to objects detected by several points of views: the object axis is computed from the shape tracked in each camera module and homography and epipolar lines allow a correct axis warping in other image planes. Both forward and backward probability contributions from the two different warping directions make the approach robust against segmentation errors, and capable of disambiguating groups of people. The system has been tested in a real setup of a urban public park, within the Italian LAICA (Laboratory of Ambient Intelligence for a friendly city) project. The experiments show how the system can correctly track and label objects in a distributed system with real-time performance. Comparisons with simpler consistent labeling methods and extensive outdoor experiments with ground truth demonstrate the accuracy and robustness of the proposed approach.

A Review and Comparison of Measures for Automatic Video Surveillance Systems

Today's video surveillance systems are increasingly equipped with video content analysis for a great variety of applications. However, reliability and robustness of video content analysis algorithms remain an issue. They have to be measured against ground truth data in order to quantify the performance and advancements of new algorithms. Therefore, a variety of measures have been proposed in the literature, but there has neither been a systematic overview nor an evaluation of measures for specific video analysis tasks yet. This paper provides a systematic review of measures and compares their effectiveness for specific aspects, such as segmentation, tracking, and event detection. Focus is drawn on details like normalization issues, robustness, and representativeness. A software framework is introduced for continuously evaluating and documenting the performance of video surveillance systems. Based on many years of experience, a new set of representative measures is proposed as a fundamental part of an evaluation framework.

Intelligent CCTV for Mass Transport Security: Challenges and Opportunities for Video and Face Processing

CCTV surveillance systems have long been promoted as being effective in improving public safety. However due to the amount of cameras installed, many sites have abandoned expensive human monitoring and only record video for forensic purposes. One of the sought-after capabilities of an automated surveillance system is “face in the crowd” recognition, in public spaces such as mass transit centres. Apart from accuracy and robustness to nuisance factors such as pose variations, in such surveillance situations the other important factors are scalability and fast performance. We evaluate recent approaches to the recognition of faces at large pose angles from a gallery of frontal images and propose novel adaptations as well as modifications. We compare and contrast the accuracy, robustness and speed of an Active Appearance Model (AAM) based method (where realistic frontal faces are synthesized from non-frontal probe faces) against bag-of-features methods. We show a novel approach where the performance of the AAM based technique is increased by side-stepping the image synthesis step, also resulting in a considerable speedup. Additionally, we adapt a histogram-based bag-of-features technique to face classification and contrast its properties to a previously proposed direct bag-of-features method. We further show that the two bag-of-features approaches can be considerably sped up, without a loss in classification accuracy, via an approximation of the exponential function. Experiments on the FERET and PIE databases suggest that the bag-of-features techniques generally attain better performance, with significantly lower computational loads. The histogrambased bag-of-features technique is capable of achieving an average recognition accuracy of 89% for pose angles of around 25 degrees. Finally, we provide a discussion on implementation as well as legal challenges surrounding research on automated surveillance.

Automated Camera Monitoring System for Selective Areas of Interest

Monitoring a number of security cameras by security guards suffers from a number of difficulties especially when the events to be watched for may take months to occur or a large number of cameras exist. Furthermore, finding and hiring people for such tasks is usually expensive. Therefore, there are immediate needs for automated surveillance systems that save money, work in a preventive way and do not miss what must be caught. In this study, we present a new algorithm for detecting motions in selected areas of interest. Although it depends on algorithms to detect scene cuts and camera breaks, but it is tailored to consider motions rather than camera breaks that usually never occurs in such surveillance video feed. The algorithm may be used in real-time by complete systems to continuously monitor surveillance video and alert security officers to suspicious events while there is still time to prevent the crime.

A MULTI-SENSORY, SYMBOLIC, KNOWLEDGE-BASED MODEL FOR HUMANLIKE PERCEPTION

Automatic surveillance systems as well as autonomous robots are technical systems which would profit from humanlike perceptive systems. We present a technical model for humanlike perception by multi-sensory fusion, symbolic information processing, and knowledge integration. The model is inspired by neuroscience. Information from diverse sensor sources is processed in parallel, transformed into symbolic representations, and further processed in a modular, hierarchical fashion. Higher-level symbolic information is gained by combination of lower-level symbols. Feedbacks from higher levels to lower levels are possible and symbol generation is also influenced by stored knowledge. Copyright © 2007 IFAC

Automated Visual Surveillance in Realistic Scenarios

In this article, we present Knight, an automated surveillance system deployed in a variety of real-world scenarios ranging from railway security to law enforcement. We also discuss the challenges of developing surveillance systems, present some solutions implemented in Knight that overcome these challenges, and evaluate Knight's performance in unconstrained environments

Automated Surveillance for Adverse Drug Events at a Community Hospital and an Academic Medical Center

To compare the rates and nature of ADEs at an academic medical center and a community hospital using a single computerized ADE surveillance system. Prospective cohort study of patients admitted to two tertiary care hospitals. Outcome Measure Adverse drug events identified by automated surveillance and voluntary reporting. We implemented an automated surveillance system across an academic medical center and a community hospital. Potential events identified by the computer were reviewed in detail by medication safety pharmacists and scored for causality and severity. Findings were compared between the two hospitals, and with voluntary reports from nurses and pharmacists. Over the 8 month study period, 25,177 patients were admitted to the university hospital and 8,029 to the community hospital. There were 1,116 ADEs in 900 patients at the university hospital for an overall rate of 4.4 ADEs per 100 admissions. At the community hospital, 399 patients experienced 501 ADEs for a rate of 6.2 events per 100 admissions. Rates of antibiotic-associated colitis, drug-induced hypoglycemia, and anticoagulation-related ADEs were significantly higher at the community hospital compared with the university hospital. Computerized surveillance detected ADEs at a rate 3.6 times that of voluntary reporting at the university hospital and 12.3 times that at the community hospital. Operation of a common automated ADE surveillance system across hospitals permits meaningful comparison of ADE rates in different inpatient settings. Automated surveillance detects ADEs at rates far higher than voluntary reporting, and the difference may be greater in the community hospital setting. Community hospitals may experience higher rates of certain types of ADEs compared with academic medical centers.

Automatic Traffic Surveillance System for Vehicle Tracking and Classification

This paper presents an automatic traffic surveillance system to estimate important traffic parameters from video sequences using only one camera. Different from traditional methods that can classify vehicles to only cars and noncars, the proposed method has a good ability to categorize vehicles into more specific classes by introducing a new "linearity" feature in vehicle representation. In addition, the proposed system can well tackle the problem of vehicle occlusions caused by shadows, which often lead to the failure of further vehicle counting and classification. This problem is solved by a novel line-based shadow algorithm that uses a set of lines to eliminate all unwanted shadows. The used lines are devised from the information of lane-dividing lines. Therefore, an automatic scheme to detect lane-dividing lines is also proposed. The found lane-dividing lines can also provide important information for feature normalization, which can make the vehicle size more invariant, and thus much enhance the accuracy of vehicle classification. Once all features are extracted, an optimal classifier is then designed to robustly categorize vehicles into different classes. When recognizing a vehicle, the designed classifier can collect different evidences from its trajectories and the database to make an optimal decision for vehicle classification. Since more evidences are used, more robustness of classification can be achieved. Experimental results show that the proposed method is more robust, accurate, and powerful than other traditional methods, which utilize only the vehicle size and a single frame for vehicle classification.

Automatic Dependent Surveillance-Broadcast/Cockpit Display of Traffic Information: Pilot Use of the Approach Spacing Application

An operational evaluation of the automatic dependent surveillance system and cockpit display of traffic information (CDTI) provided an opportunity to examine how pilot use of CDTI enabled them to maintain predefined spacing intervals from other aircraft. Seventy percent of the controller -issued traffic calls resulted in positive visual acquisitions (83% displayed on CDTI). Approximately 55% of the approach clearances that included instructions to "follow that traffic" involved communication problems and operational concerns. In light of the findings and the participants' comments, changes to proposed procedures and supporting phraseology will be constructed and evaluated for the approach spacing application.

People tracking in surveillance applications

This paper presents a real-time algorithm that allows robust tracking of multiple objects in complex environments. Foreground pixels are detected using luminance contrast and grouped into blobs. Blobs from two consecutive frames are matched creating the matching matrices. Tracking is performed using direct and inverse matching matrices. This method successfully solves blobs merging and splitting. Some application in automatic surveillance systems are suggested by linking trajectories and blob position information with the events to be detected.

Automatic Detection of Anomalous Events for Advanced Real-Time Video Surveillance

This paper introduces the architecture and the implementation details of an automatic real-time video surveillance system, capable of autonomously detecting anomalous events. Nowadays, video surveillance systems are not flexible in adapting to different operative scenarios (they only work well in known and structured environments) and generally need human assistance to recognize and tag specific visual events. The proposed system can automatically adapt to different scenarios without human intervention (but the placement of the TV sensors), and automatically leans the “typical” targets behavior in each specific operative environment by means of robust self-learning techniques. The learned knowledge is used by the system to give alerts in an automatic way by analyzing the trajectories of visual targets in the controlled scene. Robustness is obtained through an improved version of the Altruistic Vector Quantization algorithm (AVQ). The modified AVQ autonomously establishes the number of trajectory prototypes, and improves the representativeness of the prototypes themselves. Anomalies are detected if visual trajectories deviate from the “typical” learned prototypes. Standard PCs and TV cameras have been used for the actual implementation, that has been tested in many real indoor and outdoor environments. Currently real time performance has been obtained (15 fps for each camera). A preliminary learning period (about 20 minutes to grant a suitable time interval to learn all the “typical” visual trajectories) is necessary. After that period the system gives automatic alerts about events that do not conform to typical behaviors. The field view can be changed (by panning the TV camera, for example), then the system relearns the new scenario without any human intervention (no thresholds or other settings) and accurately detects anomalous events.

Intelligent distributed surveillance systems: a review

This survey describes the current state-of-the-art in the development of automated visual surveillance systems so as to provide researchers in the field with a summary of progress achieved to date and to identify areas where further research is needed. The ability to recognise objects and humans, to describe their actions and interactions from information acquired by sensors is essential for automated visual surveillance. The increasing need for intelligent visual surveillance in commercial, law enforcement and military applications makes automated visual surveillance systems one of the main current application domains in computer vision. The emphasis of this review is on discussion of the creation of intelligent distributed automated surveillance systems. The survey concludes with a discussion of possible future directions.

Advanced Surveillance: From Tracking To Event Detection

The authors suggest an automatic surveillance system based on a robust and simple algorithm for foreground detection and tracking of multiple objects in complex environments. It is possible to obtain enough information from the characteristics of the detected objects, their centroid's position and their trajectories as to detect some predefined events. Working without the operator supervision the application analyses the video streams from the CCTV surveillance system, discarding empty scenes for monitoring or recording, showing to the operator those fulfilling predefined requirements, like crowding, and rising an alarm to call the operator attention when it decides that an event may be taking place. The application is based on very simple algorithms in order to allow a single computer the analysis of up to eight video streams in real-time. Blob and scene characteristics are used to create a low-level description of predefined events. The application detects events by comparing the sequence's parameters with those associated with the events for the current scenario. The provided examples are tailored to the specific demands of a surveillance system in a public transport environment.

Preliminary assessment of an automated surveillance system for infection control.

Rapid identification and investigation of potential outbreaks is key to limiting transmission in the healthcare setting. Manual review of laboratory results remains a cumbersome, time-consuming task for infection control practitioners (ICPs). Computer-automated techniques have shown promise for improving the efficiency and accuracy of surveillance. We examined the use of automated control charts, provided by an automated surveillance system, for detection of potential outbreaks. A 656-bed academic medical center. We retrospectively reviewed 13 months (November 2001 through November 2002) of laboratory-patient data, comparing an automated surveillance application with standard infection control practices. We evaluated positive predictive value, sensitivity, and time required to investigate the alerts. An ICP created 75 control charts. A standardized case investigation form was developed to evaluate each alert for the likelihood of nosocomial transmission based on temporal and spatial overlap and culture results. The 75 control charts were created in 75 minutes and 18 alerts fired above the 3-sigma level. These were independently reviewed by an ICP and associate hospital epidemiologist. The review process required an average of 20 minutes per alert and the kappa score between the reviewers was 0.82. Eleven of the 18 alerts were determined to be potential outbreaks, yielding a positive predictive value of 0.61. Routine surveillance identified 5 of these 11 alerts during this time period. Automated surveillance with user-definable control charts for cluster identification was more sensitive than routine methods and is capable of operating with high specificity and positive predictive value in a time-efficient manner.

Automatic Dependent Surveillance--Broadcast/Cockpit Display of Traffic Information: Innovations in Pilot-Managed Departures

An operational evaluation of the automatic dependent surveillance system and cockpit display of traffic information (CDTI) was performed in October 2000. It provided an opportunity to examine how the use of a CDTI influences pilot/controller operational communications. Departure profiles were established to evaluate the ability of flight crews and air traffic controllers to manage long (6 nautical miles [nm]) and short (4.5 nm) spacing intervals between departing aircraft. Approximately 4% of the departures conducted when CDTI was not used and 9% when it was involved communication problems. Fortunately, controllers and pilots detected and corrected these problems in midstream. When pilots executed CDTI departures, there was an overall increase in time on frequency, more time spent under local control during the day, and departures were completed in less time when assigned the short spacing interval.

Telesurveillance of elderly patients by use of passive infra-red sensors in a 'smart' room

We have developed an automated surveillance system based on passive infra-red sensors. Eight sensors were installed in a hospital room. A computer automatically captured data from the sensors every night from 21:00 until 06:00 the following morning. The sensors were polled twice per second and when a sensor was activated by movement, the event and time were recorded in a data file. At the end of the surveillance period the program analysed the data and generated a report showing the activities taking place in the room and their times. Four elderly patients were observed for a total of 97 nights. A total of 1637 possible sequences of movements by the patient and the hospital staff were detected. The computer was able to identify 1450 sequences (89%) correctly, in comparison with manual analysis. Only 10 movements (0.6%) were undetected by the system; all were very short sequences (five or six activations of the sensors). The system was generally capable of detecting and classifying all major movements in the room.

Time series modeling for syndromic surveillance.

BackgroundEmergency department (ED) based syndromic surveillance systems identify abnormally high visit rates that may be an early signal of a bioterrorist attack. For example, an anthrax outbreak might first be detectable as an unusual increase in the number of patients reporting to the ED with respiratory symptoms. Reliably identifying these abnormal visit patterns requires a good understanding of the normal patterns of healthcare usage. Unfortunately, systematic methods for determining the expected number of (ED) visits on a particular day have not yet been well established. We present here a generalized methodology for developing models of expected ED visit rates.MethodsUsing time-series methods, we developed robust models of ED utilization for the purpose of defining expected visit rates. The models were based on nearly a decade of historical data at a major metropolitan academic, tertiary care pediatric emergency department. The historical data were fit using trimmed-mean seasonal models, and additional models were fit with autoregressive integrated moving average (ARIMA) residuals to account for recent trends in the data. The detection capabilities of the model were tested with simulated outbreaks.ResultsModels were built both for overall visits and for respiratory-related visits, classified according to the chief complaint recorded at the beginning of each visit. The mean absolute percentage error of the ARIMA models was 9.37% for overall visits and 27.54% for respiratory visits. A simple detection system based on the ARIMA model of overall visits was able to detect 7-day-long simulated outbreaks of 30 visits per day with 100% sensitivity and 97% specificity. Sensitivity decreased with outbreak size, dropping to 94% for outbreaks of 20 visits per day, and 57% for 10 visits per day, all while maintaining a 97% benchmark specificity.ConclusionsTime series methods applied to historical ED utilization data are an important tool for syndromic surveillance. Accurate forecasting of emergency department total utilization as well as the rates of particular syndromes is possible. The multiple models in the system account for both long-term and recent trends, and an integrated alarms strategy combining these two perspectives may provide a more complete picture to public health authorities. The systematic methodology described here can be generalized to other healthcare settings to develop automated surveillance systems capable of detecting anomalies in disease patterns and healthcare utilization.

An automated surface surveillance system

This paper describes an automated surface surveillance system, developed on behalf of the Government of Canada to detect and track illegal vessels. The scenario involves a moving target having speed significantly less than the searcher speed, slowly approaching Canada's coastline. The crux of the surveillance problem is to determine the sequence of sub-regions to search in order to maximize the probability of target detection. The complexity of our surveillance problem lies in the absence of knowledge on the target location, speeds and course. Additionally, the searcher is frequently confronted with insufficient time to area search the sub-regions. The presence of false targets and the occurrence of irregular search area further compound the problem. Our decision support system is a combination of established theories on probability maps, barrier patrol and a novel construction of heuristics for area searching irregular regions. Our approach also involves extensive use of visualization tools to aid code debugging and validation. More importantly, our automated surveillance system provides a user-friendly environment for decision planners.

オクルージョンを考慮した交通流計測

A traffic problem, such as a traffic accident, traffic congestion and environmental pollution, are serious social issues with the increase in vehicle ownership ratio. As a measure, practical application of system which provides a driver with useful information is desired. In connection with such a background, the trial in which the automated surveillance system of a traffic state and an obstacle using image processing technology is made. This research proposes the traffic flow measurement technique using image sequences taken by a CCD camera mounted on the footbridge. We perform vehicles detection using adaptive background difference processing, and vehicles tracking using the concept of a deformable template matching and a tracking cycle to provide solution to vehicles occlusion. Eventually, the traffic parameter called a traffic volume, a speed, and a space occupancy is computed. In the experimental result, the accuracy and the efficiency of our approach is shown.

Using automated medical records for rapid identification of illness syndromes (syndromic surveillance): the example of lower respiratory infection.

BackgroundGaps in disease surveillance capacity, particularly for emerging infections and bioterrorist attack, highlight a need for efficient, real time identification of diseases.MethodsWe studied automated records from 1996 through 1999 of approximately 250,000 health plan members in greater Boston.ResultsWe identified 152,435 lower respiratory infection illness visits, comprising 106,670 episodes during 1,143,208 person-years. Three diagnoses, cough (ICD9CM 786.2), pneumonia not otherwise specified (ICD9CM 486) and acute bronchitis (ICD9CM 466.0) accounted for 91% of these visits, with expected age and sex distributions. Variation of weekly occurrences corresponded closely to national pneumonia and influenza mortality data. There was substantial variation in geographic location of the cases.ConclusionThis information complements existing surveillance programs by assessing the large majority of episodes of illness for which no etiologic agents are identified. Additional advantages include: a) sensitivity, uniformity and efficiency, since detection of events does not depend on clinicians' to actively report diagnoses, b) timeliness, the data are available within a day of the clinical event; and c) ease of integration into automated surveillance systems.These features facilitate early detection of conditions of public health importance, including regularly occurring events like seasonal respiratory illness, as well as unusual occurrences, such as a bioterrorist attack that first manifests as respiratory symptoms. These methods should also be applicable to other infectious and non-infectious conditions. Knowledge of disease patterns in real time may also help clinicians to manage patients, and assist health plan administrators in allocating resources efficiently.