Classifying and detecting group behaviour from visual surveillance data

We outline an approach to classifying and detecting behaviours from surveillance data. Simple pairwise movement patterns are learned and used as building blocks to describe behaviour over a temporal sequence, or compared with other pairs to detect group behaviour. As the pair primitives are easy to redefine and learn, and complex behaviour over time is specified by the user as a sequence of pair primitives, this approach provides a flexible yet robust method of detecting complex movement in a wide variety of domains.

Context aware privacy in visual surveillance

In this paper we present preliminary work implementing dynamic privacy in public surveillance. The aim is to maximise the privacy of those under surveillance, while giving an observer access to sufficient information to perform their duties. As these aspects are in conflict, a dynamic approach to privacy is required to balance the system's purpose with the system's privacy. Dynamic privacy is achieved by accounting for the situation, or context, within the environment. The context is determined by a number of visual features that are combined and then used to determine an appropriate level of privacy.

A visual mining system for theme development evolution analysis of scientific literature

Theme development evolution analysis of literature is a significant tool to help the scientific scholars find and study the frontier problems more efficiently. This paper designs and develops a visual mining system for theme development evolution analysis to deal with the large number of literature information. The analysis of related themes based on sub-themes, together with the dynamic threshold strategy are adopted for improving the accuracy of system. Experiments results prove that correlations of themes obtained from the system are accurate and achieve better practical effect in comparison with that of our early work.

Multiple camera coordination in a surveillance system

We present a distributed surveillance system that uses multiple cheap static cameras to track multiple people in indoor environments. The system has a set of Camera Processing Modules and a Central Module to coordinate the tracking tasks among the cameras. Since each object in the scene can be tracked by a number of cameras, the problem is how to choose the most appropriate camera for each object. This is important given the need to deal with limited resources (CPU, power etc.). We propose a novel algorithm to allocate objects to cameras using the object-to-camera distance while taking into account occlusion. The algorithm attempts to assign objects in the overlapping field of views to the nearest camera, which can see the object without occlusion. Experimental results show that the system can coordinate cameras to track people and can deal well with occlusion.

Cleanup worker exposures to hazardous chemicals at a former nuclear weapons plant : piloting of an exposure surveillance system

Cleanup of former U.S. Department of Energy (DOE) nuclear weapons production facilities involves potential exposures to various hazardous chemicals. We have collaboratively developed and piloted an exposure database and surveillance system for cleanup worker hazardous chemical exposure data with a cleanup contractor at the Rocky Flats Environmental Technology Site (RFETS). A unique system feature is the incorporation of a 34-category work task-coding scheme. This report presents an overview of the data captured by this system during development and piloting from March 1995 through August 1998. All air samples collected were entered into the system. Of the 859 breathing zone samples collected, 103 unique employees and 39 unique compounds were represented. Breathing zone exposure levels were usually low (86% of breathing zone samples were below analytical limits of detection). The use of respirators and other exposure controls was high (87 and 88%, respectively). Occasional high-level excursions did occur. Detailed quantitative summaries are provided for the six most monitored compounds: asbestos, beryllium, carbon tetrachloride, chromium, lead, and methylene chloride. Task and job title data were successfully collected for most samples, and showed specific cleanup activities by pipe fitters to be the most commonly represented in the database. Importantly, these results demonstrate the feasibility of the implementation of integrated exposure database and surveillance systems by practicing industrial hygienists employed in industry as well as the preventive potential and research uses of such systems. This exposure database and surveillance system--the central features of which are applicable in any industrial work setting--has enabled one of the first systematic quantitative characterizations of DOE cleanup worker exposures to hazardous chemicals.

Development and piloting of an exposure database and surveillance system for DOE cleanup operations

An industrial hygiene exposure database and surveillance system was developed in partnership between National Institute for Occupational Safety and Health (NIOSH)-funded independent investigators and practicing industrial hygienists at the Rocky Flats Environmental Technology Site (RFETS) in Golden, Colo. RFETS is a former U.S. Department of Energy nuclear weapons plant that is now in cleanup phase. This project is presented as a case study in the development of an exposure database and surveillance system in terms that are generalizable to most other industries and work contexts. Steps include gaining organizational support; defining system purpose and scope; defining database elements and coding; planning practical and efficient analysis strategies; incorporating reporting capabilities; and anticipating communication strategies that maximize the probability that surveillance findings will feed back to preventive applications. For each of these topics, the authors describe both general considerations as well as the specific choices made for this system. An important feature of the system is a two-tier task-coding scheme comprising 33 categories of task groups. Examples of grouped analyses of exposure data captured during the system pilot period demonstrate applications to exposure control, medical surveillance, and other preventive measures. Reprinted by permission of the publisher.

Development of an exposure database and surveillance system for use by practicing OSH professionals

This report summarizes the development of an occupational exposure database and surveillance system for use by health and safety professionals at Rocky Flats Environmental Technology Site (RFETS), a former nuclear weapons production facility. The site itself is currently in the cleanup stage with work expected to continue into 2006. The system was developed with the intent of helping health and safety personnel not only to manage and analyze exposure monitoring data, but also to identify exposure determinants during the highly variable cleanup work. Utilizing a series of focused meetings with health and safety personnel from two of the major contractors at RFETS, core data elements were established. These data elements were selected based on their utility for analysis and identification of exposure determinants. A task-based coding scheme was employed to better define the highly variable work. The coding scheme consisted of a two-tiered hierarchical list with a total of 34 possible combinations of work type and task. The data elements were incorporated into a Microsoft Access database with built-in data entry features to both promote consistency and limit entry choices to enable stratified analyses. In designing the system, emphasis was placed on the ability of end users to perform complex analyses and multiparameter queries to identify trends in their exposure data. A very flexible and user-friendly report generator was built into the system. This report generator allowed users to perform multiparameter queries using an intuitive system with very little training. In addition, a number of automated graphical analyses were built into the system, including ex posure levels by any combination of building, date, employee, job classification, type of contaminant, work type or task, exposure levels over time, exposure levels relative to the permissible exposure limit (PELS), and distributions of exposure levels. Both of these interfaces, allow the user to ''drill down'' or gradually narrow query criteria to identify specific exposure determinants. A number of other industrial hygiene processes were automated by the use of this database. Exposure calculations were coded into the system to allow automatic calculation of time-weighted averages and sample volumes. In addition, a table containing all the PELs and other relevant occupational exposure limits was built into the system to allow automatic comparisons with the current standards. Finally, the process of generating reports for employee notification was automated. The implementation of this system demonstrates that an integrated database system can save time for a practicing hygienist as well as provide useful and more importantly, timely information to guide primary prevention efforts.

Vision-based target tracking and surveillance with robust set-valued state estimation

Tracking a target from a video stream (or a sequence of image frames) involves nonlinear measurements in Cartesian coordinates. However, the target dynamics, modeled in Cartesian coordinates, result in a linear system. We present a robust linear filter based on an analytical nonlinear to linear measurement conversion algorithm. Using ideas from robust control theory, a rigorous theoretical analysis is given which guarantees that the state estimation error for the filter is bounded, i.e., a measure against filter divergence is obtained. In fact, an ellipsoidal set-valued estimate is obtained which is guaranteed to contain the true target location with an arbitrarily high probability. The algorithm is particularly suited to visual surveillance and tracking applications involving targets moving on a plane.

Visual tracking of vehicles using multiresolution analysis and neural network

This paper describes the procedure for detection and tracking of a vehicle from an on-road image sequence taken by a monocular video capturing device in real time. The main objective of such a visual tracking system is to closely follow objects in each frame of a video stream, such that the object position as well as other geometric information are always known. In the tracking system described, the video capturing device is also moving. It is a challenge to detect and track a moving vehicle under a constantly changing environment coupled to real time video processing. The system suggested is robust to implement under different illuminating conditions by using the monocular video capturing device. The vehicle tracking algorithm is one of the most important modules in an autonomous vehicle system, not only it should be very accurate but also must have the safety of other vehicles, pedestrians, and the moving vehicle itself. In order to achieve this an algorithm of multi resolution technique based on Haar basis functions were used for the wavelet transform, where a combination of classification was carried out with the multilayer feed forward neural network. The classification is done in a reduced dimensional space, where principle component analysis (PCA) dimensional reduction technique has been applied to make the classification process much more efficient. The results show the effectiveness of the proposed methodology.