Classifications

• • G—PHYSICS
  • G08—SIGNALLING
  • G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
  • G08B13/00—Burglar, theft or intruder alarms
  • G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
  • G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
  • G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
  • G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
  • G08B13/19602—Image analysis to detect motion of the intruder, e.g. by frame subtraction
• • G—PHYSICS
  • G08—SIGNALLING
  • G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
  • G08B17/00—Fire alarms; Alarms responsive to explosion
  • G08B17/12—Actuation by presence of radiation or particles, e.g. of infrared radiation or of ions
  • G08B17/125—Actuation by presence of radiation or particles, e.g. of infrared radiation or of ions by using a video camera to detect fire or smoke

Definitions

• the present invention relates to a method and a device or system for detecting fires based on image analysis, in particular on the analysis of sequences of digital moving images.
• the camera can detect smoke and flames remotely, before they reach the sensor, such a system is therefore capable of filling the gaps of traditional systems outdoors or in large premises.
• the images taken by the camera can not only be - processed, but also used for viewing the incident by an operator. This is useful for removing doubts in case of false detection: the visualization of the image or the sequence of images by a human avoids many unnecessary movements.
• a fouling of the sensor is visible on the image, and according to the invention can even be detected automatically, unlike UV radiation sensors which lose their effectiveness without this being detectable.
• a camera breakdown or sabotage is automatically detectable.
• the camera used for fire detection can be used simultaneously for conventional video surveillance applications, which simplifies wiring.
• WOOO / 23959 describes a smoke detection system, consisting of video camera equipment, a unit for digitizing video signals and a unit for processing digital data.
• Smoke is detected by image processing algorithms based on the comparison of pixels between successive images.
• the comparison methods used aim, for example, to detect whether a significant change has occurred between an image and a reference image, which may indicate the appearance of smoke but also of another object in the filmed visual field.
• Another algorithm detects the Color convergence of several pixels to an average value, which may indicate a decrease in contrast caused by smoke. Such convergence may also indicate a change in lighting conditions.
• a third algorithm measures changes in the sharpness of the transition zones, affected by the smoke but also by the characteristics of the optics which are modified for example during zooms or changes of aperture. These methods are only suitable for detecting smoke, but no flames giving off little or no smoke. The algorithms used are complex and require significant computing power.
• WO97 / 16926 describes a method of detecting change in an image sequence in order to detect events.
• the detection method is based on taking a reference image which contains the Background information of the recorded scene.
• the appearance of new objects is detected by thresholding and pixel grouping methods.
• the algorithms used make it difficult to distinguish between the appearance of smoke or another object in the filmed visual field.
• EP0818766 describes a system for detecting forest fires by processing moving images. To detect fire, a smoke detection algorithm is used. This document describes a method for detecting temporal variations in the intensity of pixels at low frequency (between 0.3 and 0.1 Hz). The system is therefore rather slow to react since many cycles of a few tenths of seconds are necessary to detect a decorrelation which can indicate the presence of smoke.
• FR-A-2696939 describes an automatic forest fire detection system by image processing.
• the processing algorithms are based on the detection and analysis of volute and smoke cloud movements; they are however not very suitable for detecting flames or smoke developing in an unusual way, for example under the effect of wind or ventilation.
• Existing video image analysis fire detection systems are well suited to detecting particular types of fire in well-defined environments.
• a company wishing to specialize in fire monitoring at different sites must however acquire and familiarize itself with different software; there is currently no sufficiently robust and versatile solution for detecting very different fires using the same software.
• An object of the present invention is to provide a method and a device for detecting fire that is more reliable, faster and more versatile than the methods and systems of the prior art.
• Another object is to propose a fire detection method and system which can be implemented using a video surveillance system already installed on the site to be monitored.
• FIG. 1 a block diagram of an automatic fire detection system making it possible to implement the method of the invention.
• FIG. 2 a block diagram of a variant of an automatic fire detection system making it possible to implement the method of the invention, in which different elements are integrated into an intelligent video camera.
• FIG. 3 is a block diagram of a variant of an automatic fire detection system comprising several cameras connected to a computer via a processing unit.
• Figure 4 a schematic representation of an algorithm for frequency analysis of images for smoke detection.
• FIG. 5 a representation of slider buttons of a graphical interface making it possible to separately adjust the sensitivity of the flame and smoke detection.
• FIG. 1 illustrates a block diagram of an automatic fire detection system making it possible to implement the method of the invention.
• the illustrated system makes it possible to acquire images from different sources, for example from a PAL or NTSC 3 video camera, from a digital video camera, from a recording medium such as hard disk 2 or optical disk or of a video tape 1.
• the image sequences are digitized if necessary by a digitizer 4 and transmitted to a digital processing system 6, for example an industrial PC, which executes the algorithms for detecting flames and smoke described below .
• the digitizer 4 is constituted for example by a card for digitizing the video sequences coming from the camera or from the video recorder inserted in the digital processing system 6.
• Certain algorithms can use one or more images or sequences of reference images, for example a view of the background of the image without fire, in a memory 5.
• the results of the detection algorithms can be displayed locally on the screen of the digital processing system 6 or processed by a results interpretation and decision-making system 7 capable of generating fire or smoke alarms or pre-alarms when certain conditions predefined are met.
• This alarm can be transmitted to an alarm center 8, to an apparatus 9 generating an acoustic alarm and / or to an operator via a graphical interface 10 on one of the systems 7 or 8.
• the alarm center alarm manages all alarms coming from the results interpretation and decision-making system.
• the system 7 can be implemented by an industrial computer close to the monitored area or by a program or set of programs executed by the digital processing system 6.
• the alarm center can be located remotely and manage the alarms coming from different sites under surveillance.
• FIG. 2 illustrates a variant of the system making it possible to implement the invention, in which most of the elements of FIG. 1 are integrated into a single smart camera 3, that is to say a camera integrating processing means digital images.
• the camera incorporates optics 30, an image sensor not shown, for example a random access sensor, and an image acquisition and digital processing system 6 for acquiring the image sequences of the camera in a form digital and to execute the different flame and smoke detection algorithms described below on these image sequences.
• the intelligent camera 3 also integrates a memory 5 for storing these algorithms as well as one or more images or sequences of reference images used by these algorithms.
• a system for interpreting the results and making a decision 7 can be implemented, for example, in the form of a computer module loaded into the memory 5 and executed by the digital processing system 6.
• the intelligent camera 3 can also integrate a event management system 70 for managing the events detected by the system 7 and triggering, for example, the sending of an alarm or a pre-alarm.
• the intelligent camera 2 can be connected through a communication interface to a screen 15 to display either the sequences of images acquired live, or recorded images corresponding to detected events.
• the camera 3 is also capable of communicating its results to a computer 12.
• a control unit 11 makes it possible to choose areas of interest in the image, to vary the sensitivity of the detection, to program camera movements, etc. Camera 3 therefore constitutes a complete intelligent camera system capable of detecting flames and smoke and of generating warning signals accordingly.
• FIG. 3 illustrates another variant of the system making it possible to implement the invention, in which one or more video cameras 3 for detecting smoke 13 or flames 14 supply sequences of images directly processed by the digital processing system d 'images 6, for example an industrial PC on the monitored site.
• the system 6 executes the fire detection algorithms by image processing and the interpretation of the results.
• Processed images and events detected are transmitted to a remote operator provided with a computer 12 integrating a graphic interface making it possible to visualize the video images coming from the cameras 3 and to inform the operator in the event of alarm detection.
• the digital image processing system 6 and the system interpretation of results and decision-making 7 use several separate image processing algorithms and combine them.
• the algorithms used can be based on the following methods:
• the presence of smoke reduces the sharpness of the contours of objects present in the scene, which corresponds to a low-pass spatial smoothing filter.
• the high frequencies of the image 31 are therefore attenuated by the presence of smoke with respect to the reference image 32 stored in the memory 5 and corresponding for example to an image of the background without smoke or flames.
• the method therefore consists in calculating the frequency transform of each image 31 or portion of image acquired using a module 33 of fast Fourier transformation FFT or FHT for example and in comparing it using a system comparison 35 with the frequency transform of the reference image 32 calculated by a module 34.
• a decision module 36 can indicate a smoke alarm or a probability of a smoke alarm.
• This algorithm can be used on the whole image. In order to more clearly and more quickly detect the appearance of smoke, this algorithm is preferably applied to one or more sub-portions or zoes of the filmed image; an alarm being triggered as soon as one or minimum number of zones indicate an attenuation of the high spatial frequencies compared to the reference image. It is also possible to apply this algorithm only on the portions of the image on which smoke is likely to appear or in which another algorithm has indicated a probability of fire event. Finally, this algorithm can either be applied to a grayscale image or another component, or separately on the different components of a color image. Depending on the smoke colors likely to appear, it is possible to weight the different chromatic components differently.
• the appearance of an object whose contours, chrominance or brightness oscillate at a frequency greater than 0.5 Hz is a sign of the possible presence of flames. This can be detected using a frequency analysis method using successive images of a sequence of images. To do this analysis, the computer must have a whole sequence of images in its memory and detect objects in the spatial domain using a shape recognition algorithm.
• This algorithm can also be implemented to detect and track objects on several successive images whose shape, size and / or color vary irregularly and according to a random frequency. Object identification and object tracking methods can be used.
• stereoscopic vision algorithms can be used to assess the position, three-dimensional shape, volume and distance of filmed objects, such as new objects appearing in relation to a reference image. It is thus possible to distinguish for example between a column of smoke appearing in front of a wall and a shadow or a reflection on this wall. In the open air, this algorithm makes it possible to distinguish between a new cloud and a much closer column of smoke. This algorithm can be used for example to identify very reliably the areas of interest of an image or of a sequence of images on which the other algorithms must focus.
• Multiple image sequences can be generated, for example, using several cameras, using a single motorized camera allowing the position or the angle of view to be changed, using a or multiple cameras and a set of mirrors, etc.
• the digital processing system 6 can also be connected to one or more external sensors which may be present and which make it possible to detect particular events, for example sensors of temperature, infrared or ultraviolet radiation, movement, etc.
• the indications provided by these sensors are transmitted to acquisition cards in the digital processing system 6 and can be used to confirm the indications provided by the image processing algorithms or to improve the performance of these algorithms.
• a motion detector can be used to trigger an optical or digital zoom movement or movement from a camera to the area where the movement occurred, or to focus image processing algorithms on portions image corresponding to the area where motion was detected.
• the results of the different algorithms are combined with one another by a process of interpretation and decision-making of the results executed for example by the system 7 in order to detect the flames and / or the smoke in a reliable manner. This process of interpretation of the results can take into account the evolution of the different detection criteria as a function of time. For example, a rapidly growing detection level is more dangerous than a stable detection level.
• the sensitivity can be modified to adapt the system to its environment.
• this adjustment can be made using a single parameter influencing all the algorithms of the system.
• This parameter can be modified via a slider button on the graphical interface 10, a potentiometer, or by any other adjustment element.
• FIG. 5 illustrates two sliders for separately setting flame detection and smoke detection.
• Those skilled in the art will understand that it is easily possible, within the framework of the invention, to imagine an advanced configuration mode making it possible to separately adjust the sensitivity of each algorithm, the sensitivity applied to each zone or to each component of colors, etc. It is thus possible to use the same device and the same fire detection program and to configure it to detect flames or smoke in very different environments, for example in a road or rail tunnel, outside, in hangars, etc.
• the various events that can occur in the systems are presented by the graphical interface 10 to the operator in order of emergency.
• the graphic interface thus displays for example at the top of the list the flame and smoke alarms starting with the most recent alarm, then the flame and smoke pre-alarms starting here also with the most recent pre-alarm, other events or alarms if necessary detected being displayed at the bottom of the list.
• These other events may include, for example, camera failures, dirty cameras, indications of insufficient brightness of the scene to be monitored, or external events detected by sensors not shown, such as stalling of fire extinguishers, door openings, etc.
• a visual message, preferably a "pop-up" window indicating the type of alarm detected and opening in a graphical interface 10, and an audible beep are preferably generated when an alarm is detected
• log file a file in the processing system 6, in the system 7 or in the computer used by the remote operator and listing all the events that have occurred.
• This file is preferably made up of an XML document also containing images or sequences of images linked to each listed event, as well as the date of the event. An operator can thus consult the XML file corresponding to the monitoring period and load the recorded images, for example remotely, to check the detected alarms and ensure, for example, that the alarms detected actually correspond to fires.
• the present invention relates to a fire detection method. It also relates to a device specially adapted for implementing this method, for example a computer or an intelligent camera programmed to implement this method, as well as a data medium comprising a computer program directly loadable in the memory of such a device and comprising portions of computer code constituting means for carrying out this process.

Landscapes

• Engineering & Computer Science (AREA)
• Physics & Mathematics (AREA)
• General Physics & Mathematics (AREA)
• Computer Vision & Pattern Recognition (AREA)
• Multimedia (AREA)
• Business, Economics & Management (AREA)
• Emergency Management (AREA)
• Fire-Detection Mechanisms (AREA)
• Alarm Systems (AREA)
• Control Of Combustion (AREA)
• Closed-Circuit Television Systems (AREA)

Applications Claiming Priority (3)

Publications (3)

Family

ID=4505312

Family Applications (1)

Country Status (5)

Families Citing this family (54)

Family Cites Families (18)

• 2002
  • 2002-02-26 ES ES02711747T patent/ES2243699T3/es not_active Expired - Lifetime
  • 2002-02-26 AT AT02711747T patent/ATE298912T1/de active
  • 2002-02-26 WO PCT/CH2002/000118 patent/WO2002069292A1/fr not_active Application Discontinuation
  • 2002-02-26 EP EP02711747A patent/EP1364351B8/de not_active Expired - Lifetime
• 2003
  • 2003-08-25 US US10/647,109 patent/US6937743B2/en not_active Expired - Fee Related

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events