Issue 21: Flame and Smoke Video Image Detection (VID)
By Daniel T. Gottuk, Ph.D., P.E.
Fire detection provides an essential
component in many fire protection system designs. Fire alarm systems
initiate public egress, start smoke control systems, and actuate fire
suppression systems. The majority of fire alarm systems used today
consists of various smoke detection technologies. These technologies
include spot-type ionization and photoelectric detectors, air sampling
smoke detectors, and linear projected beam-type smoke detectors. Fire
alarm system designers also use optical fire detectors for rapid flame
detection in applications ranging from off-shore oil drilling platforms
to aircraft hangars. All of these smoke detection technologies continue
to improve. One goal of improvement is to provide broader coverage and
faster response to real fire sources while eliminating nuisance alarms.
In addition, improvements include features that reduce maintenance,
testing and installation costs.
An alternative detection technology that provides unique advantages
over typical fire detection systems has been gaining acceptance and use.
Computer processing and image analysis technologies have improved
substantially over the course of the past decade. This rapidly advancing
technology, along with an emphasis on video surveillance since the
occurrence of events of 9/11 has propelled the development of effective
video image detection (VID) systems for fire. Fire protection system
designers initially employed these VID systems for use mostly in large
facilities, outdoor locations and tunnels. However, video-based
detection is being used for a broadening range of applications.1
For example, these systems are currently installed in electrical power
plants, paper mills, document storage facilities, historic municipal
buildings, nuclear research facilities, automotive plants,
warehouse/distribution centers, and onshore and offshore oil platforms.
In general, a fire VID system consists of video-based analytical
algorithms that integrate cameras into advanced flame and smoke
detection systems. The video image from an analog or digital camera is
processed by proprietary software to determine if smoke or flame from a
fire can be identified in the video. The detection algorithms use
different techniques to identify the flame and smoke characteristics and
can be based on spectral, spatial or temporal properties; these include
assessing changes in brightness, contrast, edge content, motion,
dynamic frequencies, and pattern and color matching. As an active area
of research, there are multiple VID systems in development; however,
there are only about five systems that are readily commercially
available. The capabilities of these systems vary from being able to
detect only flame or smoke to being able to detect both as well as
providing motion detection and other surveillance/security features.
Smoke VID systems require a minimum amount of light for effective
detection performance, and most will not work in the dark. However,
capabilities vary between systems. In general, low light cameras can
enhance performance and some systems have been developed to operate in
the dark using IR illuminators and IR sensitive cameras.2
Flame VID systems can operate effectively in dark or illuminated spaces,
and some systems will have enhanced sensitivity to flaming fires in the
There are two basic architectures utilized by VID systems. Due to
limitations of video processing technologies, the initial systems
consisted of multiple cameras (usually a maximum of eight), each with an
analog cable connection back to a central processing unit that executed
all video capture and alarm algorithms (Figure 1). The processing unit
typically has relay contact outputs and the ability to send various
alarm signals to standard fire alarm control units. The systems can also
record still shots or video clips associated with any alarm event and
can provide instantaneous video display to a monitor.
Figure 1. VID system with CCTV cameras individually processed by a central control unit that runs the alarm algorithms.
Advancement in technologies has resulted in the second type of
architecture where both the video processing and alarm algorithm
execution are performed at the camera in a single, spot-type device
(Figure 2), just like a typical optical flame detector. These fire
detectors can have onboard storage of video and can be integrated on a
closed-circuit system with an additional central processing unit, or it
can be integrated as a spot detector on a standard fire alarm system.
These devices can also be monitored remotely via network or internet
connections. Video events of alarm conditions can be archived for each
device and can be displayed automatically to monitors for instantaneous
Figure 2. Example of a spot-type flame VID device with camera and alarm processing in unit (courtesy of Micropack).
The National Fire Alarm Code, NFPA 72,3
recognizes the use of flame and smoke VID systems. Per this Code, the
installation of these systems requires a performance-based design. There
are no prescriptive siting requirements. Flame VID systems are
classified as radiant-energy sensing fire detectors and are treated
similar to optical flame detectors. Due to the variability of VID system
capabilities and the differences in alarm algorithm technologies, NFPA
72 requires that the systems be inspected, tested, and maintained in
accordance with the manufacturer's published instructions. Currently,
there are no systems that are UL listed, but three systems have been FM
approved. These include a system that detects only smoke, one that
detects only flame and one that detects both.
As noted, commercially available systems range in capabilities. They
also vary considerably in their setup and manner of use. Some systems
have little or no user definable settings and are almost plug-and-play,
and some systems require a trained manufacturer's representative to
customize the system to the application. Once set up, there is little
maintenance required of a VID system. Similar to any field-of-view
detector, the primary issue is keeping the optical windows clean and the
camera position fixed and unobstructed. Most systems monitor the video
image, such as low or exceedingly high light levels, video loss,
significant image changes, or obscured camera images, and provide a
warning if the image quality is degrading or not sufficient for proper
detection performance. Currently, VID systems can be tested/checked in
one of three ways: 1) using a target smoke or fire source, 2) feeding
the system a pre-recorded image of a flame or smoke event, or 3) using a
product-specific electronic device that directs a pre-set "light"
signal to the VID detector.
VID systems provide unique advantages in a wide range of
applications. One advantage these systems offer is the ability to
protect a larger area, while still achieving fast detection. This is
particularly unique for smoke VID systems compared to spot or beam smoke
detectors. In many large facilities with excessive ceiling heights,
designers find it impractical to use conventional smoke detection
devices. VID systems are able to detect smoke or flame anywhere within
the field of view of the camera; whereas conventional smoke detectors
require smoke to migrate to the detector. Figure 3 shows an example of a
large facility application. VID systems can also be used for outdoor
applications, such as train stations and off-shore oil platforms.
Figure 3. Smoke VID system detecting smoke in a power plant (courtesy of axonX)
The ability to use the basic hardware of the VD system (i.e., the
cameras and wiring) for multiple purposes is clearly one of the primary
advantages of this technology. Integrating video-based fire detection
with video surveillance inherently minimizes certain installation,
maintenance and service costs and can increase system reliability due to
more frequent use of and attention to the video equipment. Providing
fire protection for historic buildings poses many challenges to not
disturb the historic features of the structure. Running wire and
mounting devices of typical fire alarm systems is just not possible in
many of these applications for both aesthetic and practical installation
reasons. Many museums and historic buildings already have surveillance
cameras installed, which makes the use of VID systems attractive.
Another advantage of VID systems is the ability to have live video
immediately available upon detecting a pre-alarm or an alarm condition.
Immediate situational awareness allows monitoring personnel to easily
view the protected area to determine the extent of the fire and to more
accurately identify the location. In the case of nuisance alarms, the
live video allows better assessment and a more appropriate response to
the event. Video archiving of events provides a means to diagnose fires
and potential problems and a basis to make system adjustments in the
case of fires or nuisance sources.
The potential for nuisance sources is highly dependent on the
specific VID technology. Some systems have the ability to ignore areas
of the field of view that may have potential nuisance sources, to adjust
sensitivity and to adjust the persistence time of the event before an
alarm signal is issued. Specific alarm algorithms have also been
developed by manufacturers to avoid common nuisance events.
In an age of ever-increasing use of video for surveillance and
security, the potential to utilize the video images for multiple
purposes offers a number of advantages. The development of flame and
smoke VID technology capitalizes on the need for video, but more
importantly, it provides a significant step forward in fire protection.
VID technology provides advantages in many applications that cannot be
effectively covered by typical fire alarm systems. Besides the
commercially available VID systems, multiple systems are being developed
and investigated by a range of manufacturers. As the technology becomes
more widely accepted, used and recognized in the codes and standards,
fire protection engineers should consider the potential benefits of
using flame and smoke VID systems in their fire protection system
Daniel Gottuk is with Hughes Associates
Gottuk, D.T., Lynch, J.A., Rose-Pehrsson, S.L, Owrutsky, J.C. and
Williams, F.W., "Video Image Fire Detection for Shipboard Use," AUBE '04 – Proceedings of the13th International Conference on Automatic Fire Detection, Duisburg, Germany, September 14–16, 2004.
Zakrzewski, R.R., Sadok, M., and Zeliff, B., "Video-based Cargo Fire
Verification System for Commercial Aircraft," AUBE '04 – Proceedings of
the 13th International Conference on Automatic Fire Detection, Duisburg,
Germany, September 14–16, 2004.
NFPA 72, National Fire Alarm Code, National Fire Protection Association, Quincy, MA, 2007.
For questions concerning delivery of this e-Newsletter, please contact our Customer Service Department at (216) 931-9934 or magazine.sfpe.org.