JP2019139794A - Fire detection device and fire detection method - Google Patents

Abstract

To provide a fire detection device and a fire detection method for detecting a fire by making it possible to photograph, with a camera, a trace amount of smoke rising during fumigation burning that is frequently observed at the early stage of fire, while lighting is turned off in the night time.SOLUTION: An optical spot irradiation device 12 irradiates an irradiation surface 22 partitioning a monitoring area 18 that is photographed with a monitoring camera 10 with a plurality of rays of spot light to form spot images 20, and photographs, with the monitoring camera 10, an image of an area including the irradiation surface 22 irradiated with the plurality of rays of spot light. An image processing apparatus 14 extracts the spot images of the areas irradiated with the spot light from the image photographed with the monitoring camera 10, and detects a predetermined characteristic change due to smoke to determine the occurrence of fire.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a fire detection device and a fire detection method for detecting smoke in an early fire from an image of a monitoring area captured by a camera.

  Conventionally, various devices and systems have been proposed in which a fire is detected by performing image processing on an image of a monitoring area captured by a monitoring camera.

  In such a fire detection device, early detection of a fire is important from the viewpoint of initial extinguishing and evacuation guidance for the occurrence of a fire.

  For this reason, in the conventional device (Patent Document 1), as a phenomenon caused by smoke from a fire from an image, a decrease in transmittance or contrast, a convergence of a luminance value to a specific value, a luminance distribution range is narrowed and a luminance distribution is reduced. , A change in average value of brightness due to smoke, a decrease in the total amount of edges, and an increase in intensity in the low frequency band are derived, and these can be comprehensively judged to detect smoke.

JP 2008-046916 A JP-A-7-245757 JP 2010-238028 A US Patent Application Publication No. 2009/0244280

  However, in such a conventional fire detection system that detects a fire from a smoke image associated with a fire, the entire image captured by the surveillance camera is processed to detect a characteristic change due to smoke and determine the fire. However, there is a problem that the processing load for judging the fire from the entire image increases and the processing takes time.

  Also, when the illumination of the monitoring area is turned off during the night when nobody is present, there is a problem that the image of the monitoring area captured by the monitoring camera becomes dark and the function of determining a fire from the smoke image is lost.

  To solve this problem, you can leave the lights on at night, etc. without turning them off, but leaving the lights on in places where you are not using is a practical solution. Must not.

  The present invention reduces the processing load by processing a part of the image captured by the surveillance camera and judges a fire, thereby shortening the processing time and monitoring even when the lighting is turned off at night. It aims at providing the fire detection apparatus which can detect a fire from the image imaged with the camera.

(apparatus)
The present invention provides a fire detection device,
An imaging means for capturing an image of the monitoring area;
Spot light irradiating means for irradiating a plurality of spot lights toward the monitoring region imaged by the imaging means;
Spot image extraction means for extracting a spot image of an area irradiated with a plurality of spot lights from an image captured by the imaging means;
A fire determination means for detecting a predetermined change due to smoke in a plurality of spot images extracted by the spot image extraction means, and determining a fire;
With
The spot light irradiating means irradiates a plurality of spot lights so as to form a plurality of spot images arranged in a random pattern unique to itself on an irradiation surface of a monitoring region that is an imaging direction of the imaging means. And

(Preventing false detection)
The fire determination unit determines erroneous detection when the spot image extracted by the spot image extraction unit is not based on the random pattern formed by the spot light irradiation unit.

(Infrared area spot light)
The spot light irradiating means irradiates spot light in the infrared region, and the imaging means has imaging sensitivity including the infrared light wavelength region.

(Fire judgment)
The fire determination unit determines that a fire has occurred when a time-series change in which the number of smoke spot images in which a characteristic change due to smoke has been detected by the spot image extraction unit increases in the upward direction is detected.

  Here, the fire determination means detects, as the smoke spot image, a position where the brightness of the spot image extracted from the spot image has changed in a predetermined manner.

(Location of spot irradiation means)
The image pickup means and the spot light irradiation means are arranged side by side at a predetermined position overlooking the monitoring area.

  Alternatively, the imaging unit is arranged at a predetermined position overlooking the monitoring area, and the spot light irradiation unit is arranged at a predetermined position on the imaging direction side of the imaging unit.

(Method)
The present invention provides a fire detection method,
An image of the monitoring area is captured by the imaging means,
Irradiating a plurality of spot lights toward the monitoring area imaged by the imaging means by the spot light irradiation means,
A spot image extraction unit extracts a spot image of a region irradiated with a plurality of spot lights from an image captured by the imaging unit,
A fire is determined by detecting a characteristic predetermined change due to smoke in the plurality of spot images extracted by the spot image extracting means by the fire determining means,
The spot light irradiating means irradiates a plurality of spot lights so as to form a plurality of spot images arranged in a random pattern unique to itself on an irradiation surface of a monitoring region that is an imaging direction of the imaging means. And

  The other features of the fire detection method of the present invention are basically the same as those of the above-described fire detection device, and thus description thereof is omitted.

(Basic effect)
According to the fire detection device and the fire detection method of the present invention, an imaging unit that captures an image of a monitoring area, a spot light irradiation unit that irradiates a plurality of spot lights toward the monitoring area captured by the imaging unit, and an imaging unit A spot image extracting means for extracting a spot image of an area irradiated with a plurality of spot lights from the image captured in step, and detecting a predetermined characteristic change caused by smoke in the plurality of spot images extracted by the spot image extracting means. When the whole image is processed, it is only necessary to extract and process the image of the spotted area scattered in the image of the monitoring area captured by the imaging unit. Compared with, the processing load is reduced and the processing time can be shortened.

  In addition, since the spotlight is radiated, it is possible to capture the image of the place where the spotlight is to be processed even at night when the illumination is turned off. In addition, it is possible to reliably determine a fire from a change in the image due to smoke. Further, when the light is turned off, the spot light is only scattered and hits the wall surface or the like, so that the influence on the light-off state is small.

(Effect of spot light in infrared region)
In addition, since the spot light irradiating means irradiates spot light in the infrared region, and the imaging means is provided with imaging sensitivity including the infrared region, even if the illumination is turned off at night, spots in the infrared region that cannot be seen by humans Since it is light irradiation, it is possible to determine a fire from an image of a spot light irradiation region that changes due to smoke without hindering light extinction.

(Effect by fire judgment)
In addition, because the fire determination means detects a time-series change in which the number of smoke spot images in which a characteristic change due to smoke is detected by the spot image detection means increases in the upward direction, it is determined as a fire. It is possible to distinguish fires from changes caused by movement of people, small animals, etc., and to make sure that fires can be judged from changes caused by fire smoke.

(Effect of fire judgment by dividing captured images)
The fire determination means divides the captured image into a plurality of rectangular areas having a predetermined height and width, detects smoke spot images having characteristic changes due to smoke in each rectangular area, and detects the height of the rectangular area. Aligned in the direction of smoke propagation set in the vertical direction or width direction, and detected smoke from the time-series change of the aligned smoke spot images in each rectangular area, so that a fire was judged. With respect to the movement of the corresponding smoke spot image, it is possible to judge the fire by reliably capturing the upward smoke propagation without being affected by the fluctuation in the horizontal direction.

  Also, even if a noisy smoke spot image with no movement is detected due to changes in brightness due to factors other than smoke due to fire, smoke propagation is not detected because there is no time series change, and noisy spot images are detected. It is possible to reliably detect smoke from a fire without being affected by changes in concentration.

(Effects of setting propagation monitoring range and fluctuation allowable range)
In addition, the fire determination means sets either one of the height and width of the rectangular area as a range where smoke propagation is monitored, and sets the other as a range where smoke fluctuation is allowed, and further upwards the smoke. When monitoring propagation, the height of the rectangular area is set longer than the width, and when monitoring the propagation of smoke in the horizontal direction, the width of the rectangular area is set longer than the height. And width are appropriately set according to the smoke propagation monitoring range (propagation direction) and the smoke fluctuation tolerance range, even if the smoke movement fluctuates with respect to the smoke propagation monitoring direction, it is affected by the fluctuation. Without being able to detect the propagation of smoke.

Explanatory drawing which showed the monitoring area | region which installed the fire detection apparatus of this invention Explanatory drawing which showed irradiation of the spot light with respect to the monitoring area | region of FIG. 1 about a side surface, a plane, and an irradiation surface Explanatory diagram showing time-series changes in spot images whose brightness changes due to smoke Block diagram showing outline of functional configuration of image processing apparatus Explanatory drawing which showed the other example of the monitoring area | region which installed the fire detection apparatus of this invention which has arrange | positioned the spot light irradiation apparatus in front of the monitoring camera Explanatory drawing which showed irradiation of the spot light with respect to the monitoring area | region of FIG. 5 about the side surface, the plane, and the irradiation surface Explanatory diagram showing a captured image of a monitoring area divided into a plurality of rectangular areas in which the vertical direction is a range in which smoke propagation is monitored and the horizontal direction is a range in which smoke fluctuation is allowed Explanatory drawing which showed the time change of the smoke spot image detected from the smoke which propagates upward about the captured image divided | segmented into the rectangular area of FIG. Explanatory drawing which showed the time change of the smoke spot image following FIG. Explanatory drawing which showed the state which arranged the smoke spot image for every rectangular area shown in FIG. 8 in the propagation direction which becomes an upper direction Explanatory drawing which showed the state which arranged the smoke spot image for every rectangular area shown in FIG. 9 in the propagation direction which becomes an upper direction Explanatory drawing which showed the number of the smoke spot image for every rectangular area shown in FIG.10 and FIG.11 with the matrix Explanatory drawing which showed as a binary matrix which binarized the presence or absence of the upward propagation of the smoke spot image for every rectangular area shown in FIG. An explanatory diagram showing a captured image of a monitoring area divided into a plurality of rectangular areas in which the width direction is a range where smoke propagation is monitored and the height direction is a range where smoke fluctuation is allowed Explanatory drawing which showed the time change of the smoke spot image detected from the smoke which propagates along a ceiling about the captured image divided | segmented into the rectangular area of FIG. Explanatory drawing which showed the time change of the smoke spot image following FIG. Explanatory drawing which showed the state which aligned the smoke spot image for every rectangular area shown to FIG.15 and FIG.16 in the propagation direction used as a horizontal direction Explanatory drawing which showed the number of the smoke spot images for every rectangular area shown in FIG. Explanatory drawing which showed as a binary matrix which binarized the presence or absence of the propagation to the horizontal direction of the smoke spot image for every rectangular area shown in FIG.

[Outline of fire detection device]
FIG. 1 is an explanatory view showing a monitoring area in which a fire detection device according to the present invention is installed.

  As shown in FIG. 1, a monitoring camera 10 that functions as an imaging unit and a spot light irradiation device 12 that functions as a spot light irradiation unit are installed in the monitoring region 18, and a plurality of spot lights are monitored from the spot light irradiation device 12. The image is taken by the monitoring camera 10 in a state of irradiating 18.

  The monitoring camera 10 is installed, for example, in the center of an upper corner of a monitoring area (monitoring space) 18 partitioned vertically, horizontally, and forward and backward, and the imaging optical axis 10a is inclined so as to be directed to the vicinity of the center of the opposed irradiation surface 22. The monitor area 18 is set to face downward so that the entire monitoring area 18 can be imaged.

  The spot light irradiation device 12 is arranged side by side next to the monitoring camera 10 in the same upper corner where the monitoring camera 10 is installed, for example, a total of 49 spot light emitting units in seven rows in a vertical and horizontal direction are two-dimensionally arranged and opposed wall surfaces. A plurality of spot lights are irradiated obliquely downward so as to face the irradiation surface 22, and a total of 49 spot images 20 in seven rows in the vertical and horizontal directions are formed on the irradiation surface 22. In the following description, an image portion when the spot image 20 is captured by the monitoring camera 10 is referred to as a spot image.

  The spot light irradiation device 12 irradiates infrared light having a predetermined infrared light wavelength band, for example, infrared light having a wavelength of 820 nanometers or a wavelength of 940 nanometers. For this reason, the spot light irradiation apparatus 12 arrange | positions several infrared LED provided with the micro condensing lens in the matrix form, for example, and irradiates several infrared beam light.

  As another example of the spotlight illumination device 12, when the irradiation distance may be short, a slit plate in which slit holes are formed in seven rows in length and width is arranged in front of the surface light source that emits light from the plurality of infrared LEDs. However, a plurality of spot lights may be irradiated.

  Correspondingly, the surveillance camera 10 has imaging sensitivity in the visible light wavelength range, and has imaging sensitivity in the wavelength range including the infrared light wavelength of the wavelength 820 nanometer or the wavelength 940 nanometer by the spot light irradiation device 12. The element is used.

  In this way, with the spot light irradiation device 12 irradiating the monitoring area 18 with a plurality of spot lights, a combustible material in the fire source 24 such as a trash can placed in the monitoring area 18 causes a fire for some reason. Smoke 25 rises from the fire source 24.

  When the smoke 25 is present in the monitoring area 18, the monitoring camera 10 captures an infrared image including the irradiation surface 22 on which the spot image 20 whose luminance changes on the imaging screen due to the smoke 25 exists, and is captured by the monitoring camera 10. The transmitted image is transmitted to an image processing apparatus 14 installed in a manager's room or the like via a transmission path, and smoke is detected from a fire source 24 such as a garbage can by image processing to determine a fire, and a fire is detected. A signal is output to the fire alarm facility 16 to output a fire alarm.

[Detection principle]
The principle of detecting smoke according to the present invention will be described as follows. FIG. 2 is an explanatory diagram showing the side (FIG. 2 (A)), the plane (FIG. 2 (B)), and the irradiation surface (FIG. 2 (C)) of spot light irradiation on the monitoring region of FIG.

  As shown in FIG. 2, a plurality of spot lights are emitted obliquely downward as indicated by alternate long and short dash lines from the spot light irradiation device 12, and 49 spot images 20 are formed on the irradiation surface 22 in seven columns in the vertical and horizontal directions. The monitoring camera 10 captures the spot image 20 on the irradiation surface 22 and acquires an infrared image of the monitoring area including the spot image.

  If the smoke 25 rises from the fire source 24 at this time, the smoke 25 crosses a part of the passage positions of the plurality of spot lights irradiated from the spot light irradiation device 12 and is captured by the monitoring camera 10. Part of the light is blocked by the smoke 25, and at the same time, the spot light hits the smoke 25, causing a scattering phenomenon due to the smoke, and the brightness decreases when viewed in the infrared light wavelength region.

  For this reason, when the image captured by the monitoring camera 10 is viewed, the brightness of the spot image at the position corresponding to the smoke 25 decreases. Further, the smoke 25 caused by the fire causes a time-series movement that rises from the fire source 24 due to the hot air flow, and therefore causes a time-series change in which the number of spot images whose luminance is decreased due to scattered light increases in the upward direction. .

  FIG. 3 shows a time-series change of a captured image including a spot image in which the brightness falls when the spot light hits the smoke.

  Since the captured image 42a in FIG. 3A is the initial stage of the fire and smoke starts to rise, the smoke spot image 40 in which the brightness of the spot image has decreased as a characteristic change due to smoke exists below the irradiation surface. doing.

  In the subsequent captured image 42b of FIG. 3 (B), the number of smoke spot images 40 increases in a vertical direction due to the rise of smoke accompanying the expansion of the fire, and when further time passes, the captured image of FIG. 3 (C). As shown in the image 42c, the number of smoke spot areas 40 increases to extend further vertically.

  For this reason, in the present invention, as the characteristic predetermined change due to smoke, the number of smoke spot images whose brightness has decreased due to the generation of scattered light due to the spot light hitting the smoke increases in the upward direction. A fire is judged when a time series change is detected. This makes it possible to reliably detect fire smoke and distinguish it from changes in the brightness of spot images that accompany movement of people, small animals, etc., and reliably determine fire from changes due to fire smoke.

[Fire detection device]
(Functional configuration of fire detection device)
FIG. 4 is a block diagram showing an outline of the functional configuration of the fire detection device according to the present invention. As shown in FIG. 4, the fire detection device includes a monitoring camera 10, a spot light irradiation device 12, and an image processing device 14. The image processing device 14 includes a CPU, a memory, various input / output ports and the like as hardware. As a function realized by execution of a program by the CPU, a control unit 26, a spot image extraction unit 32 that functions as a spot image extraction unit, and an image change detection unit that functions as an image change detection unit 34. A fire judgment unit 36 that functions as a fire judgment means is provided.

  In addition, an alarm display unit 37 and an acoustic alarm unit 38 are provided to notify an alarm when a fire is determined.

  The transmission unit 30 uses an appropriate transmission interface for receiving image data captured by the monitoring camera 10, and the irradiation drive unit 28 drives the spot light irradiation device 12 to turn on.

  The control unit 26 performs overall control of each function provided in the monitoring camera 10, the spot light irradiation device 12, and the image processing device 14.

  The control of the spot light irradiation device 12 by the control unit 26 is based on continuous lighting driving during operation, but for example, it is possible to drive the lighting by determining a time zone during which it is unattended by setting a timer.

  Further, as another control of the spot light irradiation device 12 by the control unit 26, the brightness of the room is measured by measuring the illuminance of the monitoring space with an illuminometer or measuring the total luminance of the image captured by the monitoring camera 10. The spot light irradiation device 12 may be driven when the illuminance decreases or when the total luminance of the image becomes a predetermined threshold value or less.

  Further, spot light irradiation by the spot light irradiation device 12 may be driven (irradiated) in a pulsed manner, and an image may be taken by the monitoring camera 10 in synchronization therewith.

  The monitoring camera 10 functioning as an imaging unit operates in response to an instruction from the control unit 26, transmits image data of a monitoring area, for example, 30 frames per second as moving image data by transmission control of the transmission unit 30, and images The data is stored in a memory (not shown) provided in the processing device 14.

  As shown in FIG. 3, the spot image extraction unit 32 extracts a plurality of spot images that are irradiated with the spot light from the frame-unit images stored in the memory. Prior to the start of operation of the apparatus, the spot image extraction unit 32 sets the spot image extraction range by displaying a monitoring screen on the monitor screen and setting a circular extraction region for each spot image in the monitor screen. This is prepared as a template for spot image extraction, and automatic extraction (image cutting) using the template is performed.

  For each spot image extracted by the spot image extraction unit 32, the image change detection unit 34 obtains, for example, the sum of the luminances of the pixels constituting the spot image, stores it as a reference value by, for example, a moving average, and creates a new spot. Every time the total luminance of the image is calculated, the difference (absolute value) from the reference value is calculated, and if this difference exceeds the predetermined threshold, it is estimated that the luminance change due to smoke (luminance change due to scattered light) has occurred. , Detected as a smoke spot image.

  The fire determination unit 36 stores the smoke spot image detected by the image change detection unit 34, detects its time-series change, determines a fire when a characteristic time-series change due to smoke is detected, and determines a fire detection signal. Is output to the fire alarm facility shown in FIG. In the present embodiment, as shown in FIG. 3, the fire determination unit 36 determines a fire when a time-series change in which the number of smoke spot images increases in the upward direction is detected.

(Fire detection operation)
The fire detection operation by the image processing apparatus of FIG. 4 is as follows. During operation, the spot light irradiation device 12 is always turned on to irradiate the monitoring area with a plurality of spot lights, and the monitoring camera 10 captures images of the monitoring area including a large number of spot images.

  The image processing device 14 acquires an image of a monitoring area captured at, for example, 30 frames / second as a moving image from the monitoring camera 10 and stores it in a memory. After the spot image extraction unit 32 cuts out the spot image, the image change detection unit The total luminance of each spot image is also obtained at 34, and when the difference from the reference value is greater than or equal to a predetermined value, it is detected as a smoke spot image whose luminance has changed due to smoke.

  Subsequently, when a time series change in which the number of the plurality of smoke spot images increases in the upward direction is detected as a time series change of the smoke spot images detected for each screen by the fire determination unit 36, it is determined as a fire. The fire detection signal is output to the fire alarm facility 16 in FIG. 1 to output a fire alarm.

  The alarm display unit 37 is operated to display a fire alarm, and the acoustic alarm unit 38 is operated to output a fire alarm sound.

[Other arrangement of surveillance camera and spot light irradiation device]
FIG. 5 is an explanatory view showing another arrangement of the monitoring camera and the spot light irradiation device in the monitoring area, and FIG. 6 is a side view (FIG. 6A) and a plan view (FIG. 6) of spot light irradiation to the monitoring area in FIG. (B)), it is explanatory drawing shown about the irradiation surface (FIG.6 (C)).

  As shown in FIG. 5, the monitoring camera 10 is installed at the center of a predetermined upper corner of the monitoring area 18, and the imaging optical axis 10 a is set obliquely downward so as to face the vicinity of the center of the opposed irradiation surface 22. The monitoring area 18 can be imaged.

  On the other hand, the spot light irradiation device 12 is installed in the upper part near the irradiation surface 22 in front of the monitoring camera 10. A plurality of spot light beams are irradiated obliquely downward so as to face, and a total of 49 spot images 20 in seven rows in the vertical and horizontal directions are formed on the irradiation surface 22.

  When the spot light irradiation device 12 is arranged in front of the monitoring camera 10 in this way, as shown in FIG. 6, the space of the monitoring area through which the spot light passes is limited to the narrow area on the right side in the drawing. For example, even if smoke 25 rises from the fire source 24 in the space between the monitoring camera 10 and the spot light irradiation device 12, the light from the spot light irradiation device 12 is not blocked by the smoke 25, and is imaged by the monitoring camera 10. The light from the spot image 20 on the irradiated surface 22 is blocked by the smoke 25.

  Therefore, when smoke 25 is generated, the light from the spot image 20 corresponding to the smoke 25 toward the monitoring camera 10 is blocked by the smoke, and the position corresponding to the smoke 25 in the image of the irradiation surface 22 captured by the monitoring camera 10 This causes a change in the brightness of the spot image. Therefore, the image change detection unit 34 in FIG. 4 detects a spot image whose luminance has decreased due to smoke as a smoke spot image. Other configurations and functions are the same as those in the embodiment of FIG.

  The arrangement positions of the monitoring camera 10 and the spot light irradiation device 12 can take various forms as necessary. In this case, depending on the smoke generation position in the monitoring area, a change in the spot image in the image captured by the monitoring camera 10 due to smoke increases in luminance due to the generation of scattered light by the spot light hitting the smoke, and the smoke As a result, the light from the spot image on the irradiated surface may attenuate and the brightness of the spot image may decrease. On the other hand, the image change detection unit 34 in FIG. 4 can detect the brightness change spot image in any case, and the fire determination unit 36 detects the time series change in which the number increases in the increasing direction of the brightness change spot image. The fire can be judged reliably.

[Fire judgment by segmenting captured images]
(Outline of fire judgment based on detection of smoke propagating in the height direction)
FIG. 7 is an explanatory view showing a captured image of a monitoring area divided into a plurality of rectangular areas in which the vertical direction is a range in which smoke propagation is monitored and the horizontal direction is a range in which smoke fluctuation is allowed.

  As another embodiment of the fire determination based on the smoke spot image by the image processing device 14 shown in FIG. 4, the fire determination unit 36 provided in the image processing device 14 takes an image with the monitoring camera 10 as shown in FIG. The captured image 42 of the monitoring area is divided into, for example, 10 rectangular areas A11 to A25 having a predetermined height H1 and width W1, and smoke spot images having characteristic changes due to smoke for each of the rectangular areas A11 to A25. Then, the smoke spot image is aligned with the smoke propagation direction set in the upward direction of the rectangular area, and from the time-series change of the aligned smoke spot image of each rectangular area A11 to A25, the smoke propagation direction is changed to the smoke propagation direction. A fire is judged by detecting the propagation of smoke.

  Here, the captured image 42 in FIG. 7 has a height H and a width W, and is divided into, for example, 10 sections of rectangular areas A11 to A25 having a predetermined height H1 and a width W1, in this case. In order to detect smoke propagating upward, the height H1 of the rectangular areas A11 to A25 is set longer than the width W1. Note that the height and width of the captured image 42 are determined by the number of pixels.

(Detection of smoke spot image)
8 and 9 are explanatory views showing temporal changes in smoke spot images detected from smoke propagating upward in the captured image divided into the rectangular regions in FIG.

  A smoke spot image 40 indicated by a black circle is detected by detecting a decrease in luminance due to smoke in a spot image indicated by a circle in the captured image 42a shown in FIG. 8A captured in a state in which smoke due to fire has started to rise in the monitoring area. Is detected. At the time of FIG. 8 (A), eight smoke spot images 40 are present in the rectangular area A23 due to smoke rising while fluctuating in the horizontal direction.

  In addition, a small number of smoke spot images 40 are also present in the other rectangular areas, but this is not a decrease in luminance due to smoke, but noise-like smoke detected by a decrease in luminance due to factors other than smoke. The spot image 40 is the majority.

  For the captured image 42b at the next time point shown in FIG. 8B, the smoke spot image 40 existing in the rectangular area A23 propagates to the rectangular area A13 above it, and the smoke spot image continues to the rectangular area A23. 40 appears. 9C, the smoke spot image 40 propagated to the rectangular area A23 hits the ceiling surface and spreads laterally.

(Judgment of smoke propagation direction)
In the plurality of smoke spot images 40 in the rectangular areas A23 and A13 shown in FIG. 8 and FIG. 9, there is random movement in the horizontal direction simultaneously with the upward movement, and the movement of all the smoke spot images 40 is sometimes In order to determine the movement of smoke by tracking in a sequential manner, the processing load on the image processing apparatus 14 increases, and there is a possibility that processing takes time.

  Therefore, in the present embodiment, as shown in FIGS. 10 and 11, the smoke spot images 40 for each of the rectangular regions A11 to A25 shown in FIGS. 8 and 9 are arranged in a line in the upward propagation direction. The aligned images 44a to 44c are generated.

  For example, taking the alignment image 44a as an example, the alignment images 44a to 44c are generated by sequentially assigning numbers 1 to 8 to the eight smoke spot images 40 in the rectangular area A23 shown in FIG. Then, by performing a sorting process that arranges them in numerical order, it is possible to generate an aligned image 44a arranged in one vertical column as shown in the rectangular area A23 of FIG.

  By generating the aligned images 44a to 44c in which the smoke spot images 40 of the rectangular areas A11 to A25 are arranged in the upward propagation direction, the rectangular areas A11 to A of the captured images 42a to 42c of FIGS. The fluctuation of the A25 smoke spot image 40 propagating in the horizontal direction is removed, and the movement only in the height direction, which is the direction of smoke propagation, can be extracted.

  FIG. 12 shows matrices 46a to 46c indicating the number of smoke spot images 40 for each of the rectangular areas A11 to A25 shown in FIGS. 10 and 11, and the matrix corresponding to the rectangular areas A23 and A13 positioned in the vertical relationship. As the time passes, the number of smoke spots changes as (0, 8) (6, 10) (11, 9) as indicated by the thick frame, and the other rectangular areas indicate the presence of smoke. Therefore, it can be determined that the number of smoke spot images is not sufficient.

  Accordingly, for example, a predetermined threshold TH = 5 is set for the smoke spot number matrices 46a to 46c in FIG. 12, and binarization is performed such that 0 is less than the threshold TH and 1 is greater than or equal to the threshold TH. The time-varying binary matrices 48a to 48c shown are generated. The values of the rectangular areas A23 and A13 in the binary matrices 48a to 48c change in a time series as (1,0) (1,1) (1,1) as shown by the thick frames, and the rectangular areas A23, It can be seen that smoke exists in A13 and propagates upward, and based on this, smoke can be detected to determine a fire.

(Fire judgment based on detection of smoke propagating in the horizontal direction)
FIG. 14 is an explanatory diagram showing a captured image of a monitoring area divided into a plurality of rectangular areas in which the horizontal direction is a range in which smoke propagation is monitored and the vertical direction is a range in which smoke fluctuation is allowed.

  As shown in FIG. 14, the captured image 42 having a height H and a width W sets the width W2 of the rectangular areas A11 to A52 longer than the height H2 in order to detect smoke propagating in the lateral direction along the ceiling surface. doing.

(Detection of smoke spot image)
15 and 16 are explanatory diagrams showing temporal changes in smoke spot images detected from smoke propagating in the horizontal direction for the captured image divided into rectangular regions in FIG.

  By detecting a decrease in brightness due to smoke in a spot image indicated by a circle in the imaged image 42a shown in FIG. 15 (A) taken in a state where smoke from a fire hits the ceiling surface and begins to spread laterally in the monitoring area, A smoke spot image 40 indicated by a black circle is detected. At the time of FIG. 15 (A), there are five smoke spot images 40 due to smoke propagating in the rectangular direction A11 near the ceiling surface while fluctuating in the height direction.

  In addition, a small number of smoke spot images 40 are also present in the other rectangular areas, but this is not a decrease in luminance due to smoke, but noise-like smoke detected by a decrease in luminance due to factors other than smoke. The spot image 40 is the majority.

  The captured image 42b at the next time point shown in FIG. 15 (B) propagates when the smoke spot image 40 continues to appear in the rectangular area A11 and increases in number in the right direction, and also in the rectangular area A12 next to it. Is propagating. Further, at the time of FIG. 16C, the smoke spot image 40 propagated to the rectangular area A12 further moves to the right and increases in number.

(Judgment of lateral propagation of smoke)
In the plurality of smoke spot images 40 in the rectangular areas A11 and A12 shown in FIGS. 15 and 16, the movement in the height direction is random at the same time as the movement in the horizontal direction, and the movements of all the smoke spots are time-sequentially. Tracking and judging the movement of smoke increases the processing load on the image processing apparatus 14 and may take time.

  Therefore, in this embodiment, as shown in FIG. 17, the smoke spot images 40 for each of the rectangular areas A11 to A52 shown in FIGS. 15 and 16 are arranged in a line in the horizontal propagation direction. Images 44a-44c are generated.

  For example, taking the alignment image 44a as an example, the alignment images 44a to 44c are generated by sequentially assigning numbers 1 to 5 to the five smoke spot images 40 in the rectangular area A11 shown in FIG. Then, by performing a sorting process in which the images are arranged in numerical order, an aligned image 44a arranged in one horizontal row as shown in the rectangular area A11 in FIG. 17 can be generated.

  By generating the aligned images 44a to 44c in which the smoke spot images 40 of the rectangular regions A11 to A52 are arranged in the propagation direction which is the horizontal direction, the rectangular regions A11 to A of the captured images 42a to 42c of FIGS. The fluctuation in the height direction of the smoke spot image 40 of A52 is removed, and only the movement in the lateral direction, which is the smoke propagation direction, can be extracted.

  FIG. 18 shows matrices 46a to 46c indicating the number of smoke spot images 40 for each of the rectangular areas A11 to A52 shown in FIG. 17, and the smoke in the matrix portion corresponding to the rectangular areas A11 and A12 located in the left-right relationship. The number of spots changes with time as (5, 1) (10, 2) (8, 9) as indicated by a thick frame, and the other rectangular areas are sufficient to indicate the presence of smoke. The number of smoke spot images does not exist, and it can be determined that the number of smoke spots is noisy.

  Accordingly, for example, a predetermined threshold value TH = 5 is set for the smoke spot number matrices 46a to 46c in FIG. 18, and binarization is performed so that 0 is less than the threshold value TH and 1 is greater than or equal to the threshold value TH. The time-varying binary matrices 48a to 48c shown are generated.

  The values of the rectangular areas A11 and A12 in the binary matrices 48a to 48c change in a time series as (1,0) (1, 0) (1, 1) as shown by the thick frame, and the rectangular areas A11 It can be seen that smoke exists and spreads in the lateral area A12 and propagates in the lateral direction, and based on this, smoke can be detected to determine a fire.

(Parallel processing of smoke judgments propagating in the height and lateral directions)
As another embodiment of the present invention, the functions of both the judgment process for smoke propagating in the height direction shown in FIGS. 7 to 13 and the judgment process for smoke propagating in the lateral direction shown in FIGS. 4 is provided in the fire determination unit 36 of the image processing apparatus 14 shown in FIG. 4, and the fire is detected by detecting smoke propagating in the height direction or smoke propagating in the horizontal direction by parallel processing of one image processing apparatus 14. You may make it do.

[Modification of the present invention]
(Multiple units installed)
In the above embodiment, one spot light irradiation device and one irradiation device are installed for one monitoring camera, but a plurality of spot light irradiation devices may be installed. As a result, a large area that cannot be covered by one irradiation device can be covered by installing a plurality of spot light irradiation devices. In this case, the surveillance camera can be a wide-angle type, or a wide area can be monitored by performing an automatic turning operation.

  A plurality of monitoring cameras may be installed for one spot light irradiation device.

(Irradiation pattern)
The irradiation pattern by the spot light irradiation device may be a random irradiation pattern unique to each spot light irradiation device. Accordingly, it is possible to determine whether or not the spot light is emitted from the spot light irradiation device and to prevent erroneous detection.

(Spot light irradiation device)
In the above embodiment, the spot light irradiation device emits spot light having an infrared light wavelength. However, visible light spot light may be emitted.

  Further, the spot light irradiation device is not limited to a device that irradiates a plurality of spot lights in a specific direction. For example, a spot hole is opened on a spherical surface, and light from an internal light source is irradiated in a 360 ° direction. May be. In this case, it is possible to monitor by installing a plurality of monitoring cameras or by installing an omnidirectional monitoring camera having a dome-shaped field of view.

(Fire detection)
In the above embodiment, as a characteristic time series change due to smoke, a fire is determined when a time series change is detected in which the number of a plurality of brightness change spot images increases in the upward direction. A fire may be judged by detecting characteristic time-series changes due to other appropriate smoke.

(Division area of captured image)
In the above embodiment, the captured image is divided into a plurality of rectangular areas having a predetermined width in the horizontal direction (x-axis direction) and a predetermined height in the vertical direction (y-axis direction). As an embodiment, the image may be divided into a plurality of rectangular regions inclined so as to have a predetermined angle with respect to the vertical direction or horizontal direction of the captured image.

  Further, as another embodiment of the divided region, for example, a plurality of parallelograms having a predetermined width in the horizontal direction of the captured image and a predetermined height in a direction inclined at a predetermined angle with respect to the vertical direction of the captured image It may be divided into shapes. Conversely, it may be divided into a plurality of parallelograms having a predetermined height in the vertical direction of the captured image and having a predetermined width in a direction inclined at a predetermined angle with respect to the horizontal direction of the captured image.

  By dividing the captured image in the diagonal direction in this way, the movement of the smoke spot image corresponding to the smoke of fire rising diagonally while fluctuating is not affected by the fluctuation in the lateral direction with respect to the moving direction. It is possible to judge the fire by surely capturing the smoke propagation in the direction.

(Other)
The present invention is not limited to the above-described embodiment, includes appropriate modifications without impairing the object and advantages thereof, and is not limited by the numerical values shown in the above-described embodiment.

10a: imaging optical axis 12: spot light irradiation device 14: image processing device 16: fire alarm equipment 18: monitoring area 20: spot image 22: irradiation surface 24: fire source 25: smoke 26: control unit 28: irradiation drive unit 30 : Transmission unit 32: Spot image extraction unit 34: Image change detection unit 36: Fire judgment unit 37: Alarm display unit 38: Acoustic alarm unit 40: Smoke spot images 42, 42 a to 42 c: Captured images 44 a to 44 c: Aligned image 46 a ~ 46c: Smoke spot number matrix 48a-48c: Binary matrix

Claims (14)

An imaging means for capturing an image of the monitoring area;
Spot light irradiating means for irradiating a plurality of spot lights toward a monitoring region imaged by the imaging means;
Spot image extraction means for extracting a spot image of an area irradiated with the plurality of spot lights from an image captured by the imaging means;
A fire determination means for detecting a predetermined fire characteristic smoke due to smoke in the plurality of spot images extracted by the spot image extraction means;
With
The spot light irradiating means irradiates the plurality of spot lights so as to form a plurality of spot images arranged in a random pattern unique to itself on an irradiation surface of the monitoring area that is an imaging direction of the imaging means. A fire detection device characterized by that.
The fire detection device according to claim 1, wherein the fire determination unit performs false detection when the spot image extracted by the spot image extraction unit is not based on a random pattern formed by the spot light irradiation unit. A fire detection device characterized by judging.
In the fire detection device according to claim 1,
The spot light irradiation means irradiates spot light in an infrared region,
The fire detection device, wherein the imaging means has imaging sensitivity including the infrared region.
2. The fire detection device according to claim 1, wherein the fire determination unit performs a time-series change in which the number of smoke spot images in which a characteristic change due to smoke is detected by the spot image extraction unit increases in an increasing direction. A fire detection device that, when detected, determines a fire.
5. The fire detection device according to claim 4, wherein the fire determination means detects, as the smoke spot image, a position where the brightness of the spot image extracted from the spot image has changed a predetermined amount. apparatus.
2. The fire detection device according to claim 1, wherein the imaging unit and the spot light irradiation unit are arranged side by side at a predetermined position overlooking the monitoring area.
In the fire detection device according to claim 1,
The imaging means is arranged at a predetermined position overlooking the monitoring area,
The fire detection device, wherein the spot light irradiation means is arranged at a predetermined position on the imaging direction side of the imaging means.
An image of the monitoring area is captured by the imaging means,
Irradiating a plurality of spot lights to the monitoring area imaged by the imaging means by the spot light irradiating means,
Extracting an image of an area irradiated with the plurality of spot lights from an image captured by the imaging unit by a spot image extracting unit;
Detecting a predetermined fire characteristic change due to smoke in a plurality of spot images extracted by the spot image extracting means by the fire judging means, and judging a fire;
The spot light irradiating means irradiates the plurality of spot lights so as to form a plurality of spot images arranged in a random pattern unique to itself on an irradiation surface of the monitoring area that is an imaging direction of the imaging means. A fire detection method characterized by:
9. The fire detection method according to claim 8, wherein the fire determination unit performs false detection when the spot image extracted by the spot image extraction unit is not based on a random pattern formed by the spot light irradiation unit. A fire detection method characterized by judging.
In the fire detection method according to claim 8,
The spot light irradiation means irradiates spot light in an infrared region,
The fire detection method, wherein the imaging means captures an image of a monitoring region including the spot light with an imaging sensitivity including the infrared region.
9. The fire detection method according to claim 8, wherein the number of spot images in which the fire detection means has detected characteristic changes due to smoke by the spot image extraction means is increased in the upward direction by the fire detection means. A fire detection method characterized by determining a fire when a time-series change is detected.
12. The fire detection method according to claim 11, wherein the fire determination means detects, as the smoke spot image, a position where a brightness of a spot image extracted from the spot image has changed a predetermined value. Method.
9. The fire detection method according to claim 8, wherein the imaging unit and the spot light irradiation unit are arranged side by side at a predetermined position overlooking the monitoring area.
In the fire detection method according to claim 8,
The imaging means is arranged at a predetermined position overlooking the monitoring area,
The fire detection method, wherein the spot light irradiation means is arranged at a predetermined position on the imaging direction side of the imaging means.

Images