JP2009245201A - Smoke detection device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a smoke detection device for detecting that smoke has been generated even when a camera shifts. <P>SOLUTION: This smoke detection device is provided with a reference image storage part for storing an image obtained by photographing when it is confirmed that any smoke is not detected, and configured to detect the generation of smoke within a photographed image by performing the image processing of the image obtained by photographing by a camera. This smoke detection device is provided with a detection region setting means for setting a prescribed detection region within the photographed latest image and the reference image; a luminance value calculation means for calculating a luminance value by arranging pixels within the detection region in the order of the large luminance value, and extracting the luminance values of the prescribed number of pixels in the order of the large luminance value and averaging them; a transmissivity calculation means for calculating transmissivity from the ratio of the luminance value of the reference image to the luminance value of the latest image; and a decision means for deciding the presence/absence of smoke in the detection region based on the transmissivity. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a smoke detection device that detects the generation of smoke by performing image processing on an image obtained by a surveillance camera.

There is a smoke detection device that installs a camera in a tunnel or the like and processes a captured image of the camera with an image processing device to detect smoke. In an image processing apparatus, generally, a reference image (reference image) is stored, and the latest captured image is compared with the reference image to extract a region where a change has occurred in the image. (For example, refer to Patent Document 1).
In general, a difference process is performed between the reference image and the latest image to extract a changed area. Difference processing compares pixels at the same coordinate position and performs a difference in luminance value of the coordinates. This is performed over the entire detection area, and the absolute value of the difference amount is added to calculate the amount of change. To do. Therefore, it can be determined that the greater the change in the luminance value, the greater the change from the reference image. Note that there is a case where a change amount is obtained by calculating a transmittance by performing a difference process between specific pixels.

Japanese Patent No. 3909665

Since the difference processing is a difference between pixels at the same coordinate position in the reference image and the latest image, a problem occurs in the following case. For example, if the floor is coated or polished very well, light and shadows are reflected on the floor, and subtle fluctuations in the light and shadows appear in the image, and people and smoke are substantive. Although the change does not occur, if the difference process is performed, a change area is extracted.
Further, when the visual field includes a water surface or the like, in an image obtained by photographing, it is possible that places with high luminance are changed by light entering from outside.
In such a case, if difference processing is performed by matching the coordinate position between the reference image and the latest image, the brightness of the entire image has not changed, but a large amount of difference in luminance value occurs, and the change region It may be determined that there is a possibility of misreporting, and an accurate calculation of the transmittance cannot be performed.

  The object of the present invention is not affected by subtle fluctuations in the reflected light, even if a place where bright light or the like is moving in the field of view due to the reflection characteristics of the object being photographed, It is an object of the present invention to provide a smoke detection device that reduces the possibility of false alarms.

  The smoke detection device according to the present invention has a reference image storage unit in which an image obtained by photographing when it is confirmed that there is no smoke is stored as a reference image, and an image obtained by photographing with a camera In a smoke detection device for detecting the generation of smoke within a range photographed by image processing, detection area setting means for setting a predetermined detection area in the latest photographed image and the reference image, and detection Brightness value calculation means for arranging the pixels in the region in order of brightness value, extracting the brightness values of a predetermined number of pixels from the larger brightness value, and calculating the average brightness value, and a reference image Characterized by comprising a transmittance calculating means for calculating a transmittance from a ratio between a bright luminance value and a bright luminance value of a latest image, and a determining means for determining the presence or absence of smoke in a detection region based on the transmittance. It is.

When calculating the transmittance, which is one of the smoke detection determination elements, the smoke detection device according to the present invention arranges the pixels in the detection region in the order of the luminance value, and selects a predetermined number from the larger luminance value. The luminance values of the pixels extracted and averaged are calculated as bright luminance values, and the luminance values of a predetermined number of pixels extracted from the smaller luminance values are averaged and calculated as dark luminance values. The transmittance is calculated based on the value.
For this reason, it is possible to obtain the brightness and darkness of the entire detection region without affecting the position coordinates of the pixels. Therefore, even if a place where bright light, etc., moves due to the reflection characteristics of the object being photographed moves in the field of view, it is not affected by subtle fluctuations in the reflected light, and may be misreported. And smoke can be detected by calculating an accurate transmittance.

A: Regarding the Principle of Smoke Detection First, before explaining the smoke detection device of the present embodiment, what principle is used to detect smoke from an image will be described with reference to FIGS. In FIG. 1 in which a room is imaged, FIGS. 1 (a) and 1 (b) show an original image taken by a surveillance camera, FIG. 1 (a) shows a state without smoke, and FIG. 1 (b) shows a detection. Shows smoke in the area. 2 (a) and 2 (b) show the luminance distribution in the detection region where the horizontal axis indicates the luminance and the vertical axis indicates the number of pixels. FIG. 1 (a) and FIG. The drawing corresponds to b). Also, in FIG. 3, FIGS. 3A and 3B show the results of the differential processing of the detection region W1, respectively corresponding to FIGS. 1A and 1B. ing.

  Here, the detection region refers to a region W1 in which the periphery of the window glass is surrounded by a rectangle in the room which is the monitoring region shown in FIGS. 1A and 1B, and smoke is generated in the room. This is an area (area) to be monitored. In FIG. 2A, it can be seen that in a state where there is no smoke in the room, the luminance is distributed over a wide range from a high value to a low value. When the luminance value of each pixel is added and the average of the luminance values (average luminance value), which is the value divided by the total number of pixels, is obtained, naturally, the variance that is a deviation from the average value shows a large value. Become.

  On the other hand, when smoke enters the detection area of FIG. 1B, the area becomes blurred. When this is seen in the luminance distribution diagram of FIG. 2B, the range of luminance that can be taken is narrower than in the state without smoke. Similarly, when the average value of the luminance values is obtained and the variance is calculated, it can be seen that the deviation (dispersion) from the average value becomes small. With the inflow of smoke, the fact that the brightness distribution range narrows can be said whether the smoke is black smoke or white smoke, but in the case of black smoke, the brightness value shifts in the direction of decreasing. However, in the case of white smoke, the luminance value shifts in the increasing direction. In addition, if the detection area is completely filled with smoke, the luminance distribution range is further narrowed, and is considered to converge to a specific luminance value.

In addition, the detection area W1 including the window glass has a large luminance difference between the outside bright part and the indoor dark part, and when the differential process (edge process) is performed, it corresponds to the outline of the window glass part. An edge is formed (see FIG. 3A). However, even if the differentiation process is performed in a smoke-containing state, the difference in luminance is not large and the edge is not so much compared to the normal state without smoke (see FIG. 3B). That is, when smoke is generated, the edge amount in the detection region is considered to change.
Furthermore, the smoke fluctuation was small, and it was confirmed that the low frequency band was larger than the high frequency band when frequency analysis was performed.

Based on such a viewpoint, the inventor of the present application, when smoke is generated,
(1) The field of view is blurred and the transmittance or contrast is lowered.
(2) the luminance value converges to a certain value;
(3) The luminance distribution range is narrowed and the luminance dispersion is reduced,
(4) The average value of luminance changes from a normal smoke-free state,
(5) In the detection area, the total amount of edges changes,
(6) The intensity of the low frequency band is greater than the intensity of the high frequency band,
I started. Judging from these comprehensively, smoke can be detected.

B: Basic Configuration of the Present Invention FIG. 4 is a configuration diagram of the smoke detection device according to the embodiment of the present invention. As shown in FIG. 4, the smoke detection apparatus according to the embodiment of the present invention captures a fire (smoke) generation monitoring range, for example, at a rate of 30 frames per second and outputs image data for each frame. And a detection device 3 for processing the image data to detect the generation of smoke and issue an alarm based on the detection. The camera 2 is composed of, for example, a CCD camera or a CMOS camera.

The detection device 3 includes a central processing unit (CPU) 4, a ROM 5, a RAM 6, a built-in timer 7, and an input / output interface (I / O) 8, and includes a computer in which a frame grabber 9 and an external storage device 10 are built. Has been.
The frame grabber 9 acquires image data from the NTSC video signal output from the camera 2. The image data is composed of, for example, one line of 640 pixels and one frame of 480 lines, and the pixels are represented by luminance of 256 gradations. The ROM 5 stores a procedure for processing operations performed by the CPU 4 as a program. The CPU 4 reads the program and advances the procedure for processing operations based on the program.

FIG. 5 shows various storage units secured in the ROM 5 and RAM 6 of the detection apparatus 3 according to the embodiment of the present invention.
As shown in FIG. 5, the detection device 3 includes an image storage unit 11 that stores luminance data related to an image acquired by the frame grabber 9 (hereinafter referred to as latest image data), and luminance data related to a reference image (hereinafter referred to as a reference image). A reference image storage unit 12 for storing data), a reference data storage unit 13 for storing reference data in each detection region of the reference image, a detection region storage unit 14 for storing information relating to the detection region, and a detection region in time series A luminance average dispersion storage unit 15 for storing the average and variance of the luminances, and a detection number storage unit 16 for storing the number of detection regions determined to have smoke for each smoke detection procedure are set as storage regions.

  As shown in FIG. 6, the rough configuration of the detection device 3 is an initial setting means 21 that performs initial settings when the detection device 3 is newly installed, and image data output from the camera 2 is stored in the image storage unit 11. Image storage means 22, detection means 23 for detecting smoke, smoke determination means 24 for determining fire when smoke is detected, reference image update means 25 for updating reference image data, and reference image replacement means for replacing reference image data 26.

Here, first, before preparing the details of the image processing method of the present embodiment, the detection preparation to be executed at the stage before image processing will be described. The initial setting unit 21 includes a detection area setting unit 27, a light / dark luminance value calculation unit 28, and a reference data calculation unit 29.
In the detection device 3, the means for extracting the smoke generation region that becomes the abnormality generation region from the detection region is based on the difference processing between the reference image and the latest image. Therefore, a reference image serving as a reference is prepared, and the difference between the image and the current image is processed.
The initial reference image is an image taken in a state where a smoke detection device is newly installed and no abnormality has occurred as shown in FIG. 1A, for example.

The detection area setting means 27 will be described. The detection area setting means 27 is executed before the detection process is started, and determines which area of the image is set as the detection area in the reference image or the latest image captured (captured) by the monitoring camera 2. Is. If it is attempted to monitor the entire image, the amount of calculation for image processing becomes enormous. Therefore, in the present embodiment, the detection area setting unit 27 divides the image into at least two detection areas having different sizes, and calculates the feature amount of the smoke detection determination element such as average luminance for each detection area. I do.
That is, as shown in FIG. 7A, a plurality of small areas, for example, a vertical 64 × horizontal 64 pixel area are combined into one small detection area, and as shown in FIG. The 128 × 128 pixel areas are combined into one large detection area, and the small detection area and the large detection area are set (divided) in a matrix in the vertical and horizontal directions, and the arbitrary area is set as the detection area. . In this way, by treating a part of the region as a detection region instead of the entire image, the amount of calculation can be reduced and the influence of noise can be reduced.

Although the size of the small / large detection area is not limited, frequency analysis or the like is performed if the size of the large detection area is set to N times the size of the small detection area (N is an integer equal to or greater than 1). When convenient. In addition, since the feature amount of the large detection region can be calculated using the feature amount calculated in the small detection region, the total calculation amount can be reduced. That is, in the state of FIG. 7, one large detection area in FIG. 7B is a collection of a total of four small detection areas, two in length and width. The corresponding large detection area operation can be processed by simply adding or dividing the two values and dividing by four. Further, two or more detection areas may be connected and handled as one detection area.
In this embodiment, two types of detection areas having different sizes such as a large detection area and a small detection area are set. However, even if three or more types of detection areas are set by changing the size of the detection area. good.

  Here, returning to FIG. 1, in FIG. 1A, the detection regions are indicated by W1 and W2. The detection area W1 is a rectangular area surrounding the window glass, and W2 is an area surrounding the corner on the floor side of the room with a rectangle. Here, the detection area W can be set in any part of the image by changing its size and shape. Preferably, the detection area W is a place where smoke to be detected is generated, and smoke is detected. It should be set in a place that is easy to do.

  The detection area is preferably set in consideration of detection sensitivity, and an area having a large change in luminance is set as the detection area. As described with reference to the smoke detection principle of FIGS. 1 and 2, as the smoke flows in, the area has a narrow luminance distribution range. That is, in a region where the luminance distribution is narrow from the normal state where there is nothing, even if smoke enters, the change is scarce and it is difficult to detect smoke. On the other hand, the window glass region W1 contains external light and has a wider luminance distribution than the W2 region at the corner of the room. Therefore, when smoke enters, changes in the luminance distribution are easily detected. Specifically, the detection area setting means 27 performs spatial differentiation on the basis of the initial reference image data photographed after confirming that there is no possibility of affecting the detection accuracy such as smoke in the preparation stage. A region having a large spatial differential value, that is, a region having a large luminance distribution range is set as a detection region.

Further, the detection area setting means 27 may exclude a predetermined luminance or higher in the reference image data in order to prevent the influence of luminance saturation. For example, a spatial differential process is performed by excluding pixels with luminance of 250 or more to obtain a spatial differential value, and an area having the largest sum of the spatial differential values is set as one detection area.
The initial reference image data is scanned over the entire small detection area, for example, a total of 4096 pixels in the length 64 × width 64 direction, and the sum of the spatial differential values of each area is obtained. Then, the region having the largest sum of the spatial differential values is set as the first small detection region, and then the region having the largest sum of the spatial differential values among the regions not overlapping the first small detection region is set as the second small detection region. Set as small detection area. By repeating this operation, all small detection areas are stored in descending order of the sum of the spatial differential values, and a predetermined number of small detection areas are set as smoke detection areas except for areas where the sum of the spatial differential values is too small. The information of the set small detection area is stored in the detection area storage unit 14.

  Further, the initial reference image data is scanned over a large detection region, for example, a region of a total of 16384 pixels in the vertical 128 × horizontal 128, and the total of the spatial differential values of each region is obtained. Then, the region having the largest sum of the spatial differential values is set as the first large detection region, and then the region having the largest sum of the spatial differential values among the regions not overlapping the first large detection region is set as the second large detection region. Set as large detection area. By repeating this operation, all large detection areas are stored in the descending order of the sum of spatial differential values, and a predetermined number of large detection areas are set as smoke detection areas except for areas where the sum of spatial differential values is too small. The information of the set large detection area is stored in the detection area storage unit 14. In setting the detection area, the spatial differential value (also referred to as edge amount) is used to determine whether or not the area is suitable for the detection area. When at least one of the value and the transmittance is low, it is desirable not to use the detection area with a mask.

By using two types of detection areas having different sizes as described above, it is possible to detect the generation of smoke with the same determination criteria even if the generation of smoke is different. That is, as shown in FIG. 8 (a), in an image obtained by photographing a range with a depth, even if smoke having the same size as smoke generated on the back side and smoke generated on the near side is generated, It is quite different and it is difficult to judge the presence or absence of smoke with the same criteria. In other words, in the deep image, the smoke S1 generated far from the camera appears very small in the image, and conversely, the smoke S2 generated near the camera has the same size. Even if it appears, it appears very large. Therefore, as shown in FIG. 8 (b), the smoke S2 generated on the near side, that is, near the camera, uses a large large detection area, and on the far side, that is, far from the camera, as shown in FIG. 8 (c). If a small detection area with a small size is used for the generated smoke S1, it is possible to ignore the distance from the camera and to detect the generation of smoke with the same criteria.
In this way, an image that captures a part with such a depth by sharing the smoke that appears in a completely different size with two detection areas with different sizes on the image that shows the part to be monitored even with smoke of the same size Can stably detect smoke.

  In the present embodiment, two types of detection areas having different sizes are used to share the front side with a detection area with a large size and the back side with a detection area with a small size. It is not limited to a region, and an appropriate location may be shared using three or more types of detection regions.

The initial setting unit 21 includes a brightness / darkness luminance value calculation unit 28. The bright / dark luminance value calculating means 28 is composed of a bright luminance value calculating means and a dark luminance value calculating means. The bright luminance value calculating means arranges the pixels in the detection area in the order of the luminance values and calculates the luminance value. The luminance value of a predetermined number of pixels extracted from the larger one and averaged is calculated as the bright luminance value. Then, the dark luminance value calculation means arranges the pixels in the detection area in the order of the luminance value, and calculates the dark luminance value by extracting and averaging the luminance values of a predetermined number of pixels from the smaller luminance value. Is.
Note that the brightness / darkness brightness value calculation means 28 uses the brightness from the captured image when the detection area is newly set and when the reference image data is updated or replaced by the reference image update means 25 or the reference image replacement means 26. A group and a dark luminance group are extracted, and a bright luminance value and a dark luminance value are calculated from the extracted bright luminance group and dark luminance group, and are stored in the reference data storage unit 13.

The bright / dark luminance value calculation means 28 calculates a bright luminance representative value (bright luminance value) and a dark luminance representative value (dark luminance value) in each detection region. Used when performing calculations.
More specifically, as shown in FIG. 9, the light / dark luminance value calculation means 28 arranges the pixels in the region for each large detection region of the reference image data in the luminance buffer in the order of the luminance value, and is the brightest. The bright luminance group and the dark luminance group are extracted without using the pixels on the side and the pixels on the darkest side. That is, when arranged in order of luminance, a predetermined number of luminances are extracted from the luminance in the predetermined order and counted in the direction in which the luminance decreases, and stored in the upper luminance register as a bright luminance group. . Then, the average of the luminance stored in the upper luminance register is obtained and stored in the reference data storage unit 13 as the reference bright luminance value of the large detection area.

  Here, a specific example (FIG. 9) is shown and described. The large detection area is composed of, for example, vertical 128 × horizontal 128, a total of 16384 pixels. The luminance of 16384 pixels is arranged from the largest to the smallest, excluding the highest luminance to the tenth luminance, and the first order as The eleventh luminance to the 40th luminance as the 30th order are extracted and stored in the upper luminance register as a bright luminance group. Then, the average value of the brightness of the bright brightness group stored in the upper brightness register is obtained and set as the reference brightness value of the large detection area.

  In addition, when the pixels are arranged in the order of the luminance magnitude, the brightness / darkness luminance value calculating unit 28 counts from the lowest luminance to the larger direction and counts a predetermined number of luminances from the predetermined luminance to the luminance increasing direction. Are extracted and stored in the lower luminance register as a dark luminance group. Then, the average of the luminance of the dark luminance group stored in the lower luminance register is obtained and stored in the reference data storage unit 13 as the reference dark luminance value of the large detection area.

  Here, a specific example (FIG. 9) is shown and described. The large detection area is composed of, for example, a total of 16384 pixels, which is 128 × 128 in height. The luminance of the 16384 pixels is arranged from the largest to the smallest, and the smallest luminance to the tenth darkest luminance are excluded, and the first order Are extracted from the 11th darkest brightness to the 40th darkest brightness as the 30th order and stored in the lower brightness register as a dark brightness group. Then, an average value of the luminance of the dark luminance group stored in the lower luminance register is obtained and used as the reference dark luminance value of the large detection area. That is, in each of the bright luminance group and the dark luminance group, the luminance value data of 30 pixels from the 11th to the 40th pixels counted from the brightest luminance or the darkest luminance is extracted and the average value is stored.

  Further, the brightness / darkness luminance value calculation unit 28 repeats the same processing as that of the large detection region for each small detection region of the reference image data, extracts the data of the bright luminance group and the data of the dark luminance group for each small detection region, and The average value is stored as a bright luminance value and a dark luminance value. This calculation is performed for all large and small detection areas set as smoke areas.

  As described above, the brightness / darkness luminance value calculating unit 28 excludes the ten largest luminances from the luminance array in which the luminances of the reference image data of the detection regions are arranged in order of size for each detection region, and the ten luminance luminance values are calculated. The average of the 30 luminances following is taken as the bright luminance value. Then, the smallest 10 luminances are excluded, and an average of 30 luminances following the excluded 10 luminances is set as a dark luminance value. In this way, in an image, the brightness of the image is grasped by collecting luminance data without considering the pixel position (coordinates). For this reason, when imaging a situation where, for example, a water surface is included in a portion to be photographed and the light from the light source is reflected by a wave whose position changes with time, the reflected light may be excluded from false alarm factors. it can.

This will be described with reference to FIG. FIG. 10A is an image obtained by photographing the light reflected by the water surface. In the figure, the two parts shown in a cross shape are the reflected light parts, and have the highest luminance value among the photographed images. It is a big part. FIG. 10B is an image taken at the same place as in FIG. 10A at a time different from that in FIG. Also in FIG. 10B, the cross-shaped portion is a portion of reflected light.
As shown in FIG. 10, when a wave is generated on the water surface, the position of the wave reflected by the light from the light source changes with time, and as a result, the pixel with the highest luminance (cross-shaped portion) As shown in FIG. 10A and FIG. 10B, the inside of the image is changed.
Here, for example, if both are directly differentiated, in the difference image (FIG. 10C), two of the reflected light portions shown in a cross shape are respectively extracted as change areas, and all (four) of them are the differences. It appears as a change area in the image (FIG. 10C). For this reason, it is not suitable to perform difference processing on a captured image in an environment in which such reflected light can move one after another. This can be said not only in the case of the difference processing but also in the case where the transmittance is calculated for each detection region using the reference image and the latest image.

  By the way, even if the position of a bright region such as reflected light changes in such an image, the brightness itself of the entire image hardly changes. Therefore, when the brightness information of the image is arranged in the order of brightness by ignoring the position coordinates of the pixels by the brightness / darkness brightness calculation means 28, the brightness brightness value and dark brightness are substantially the same as in FIGS. 10 (a) and 10 (b). Will have a value. Therefore, by using the bright luminance value and the dark luminance value obtained by arranging in order of luminance, it is possible to calculate the correct transmittance in the reference image and the latest image.

In the present embodiment, 10 from the largest and 10 from the smallest are excluded from the luminance array, but may be excluded from only one or the number to be excluded depending on the shooting location. What is necessary is just to set suitably.
Further, in the present embodiment, 30 elements from the eleventh to the fortyth element in the larger luminance array and 30 elements from the eleventh to the fortyth element in the smaller one are extracted, respectively, and the bright luminance value and Although the dark luminance value is calculated, the number to be extracted is not limited to this.

  Returning to FIG. 6 again, the initial setting means 21 includes reference data calculation means 29. The reference data calculation unit 29 obtains the reference average luminance and reference variance of the large or small detection area from the reference image data for each large or small detection area and stores it in the reference data storage unit 13.

Next, the configuration of the image storage unit 22 will be described. The image storage unit 22 includes a luminance correction unit 31 and a current image storage unit 32.
The luminance correction unit 31 corrects the luminance by performing inverse gamma correction on the image data input from the camera 2. The reason for the inverse gamma correction is that the image data captured and output by the camera 2 is subjected to gamma correction so as to match the human visibility when the image data is displayed on the monitor. This is because when the processed image data is processed as it is, distortion due to gamma correction is included in the processing result.

  The current image storage means 32 stores the image data whose luminance has been corrected for each predetermined capture period in the image storage unit 11. The current image storage means 32 extracts the image data whose brightness has been corrected by the brightness correction means 31 every 3 seconds, for example, stores it for 90 frames, and also deletes the old image data first when storing a new image. The current image data is updated according to the first-in first-out procedure.

Next, the configuration of the detection unit 23 will be described. The detection means is a general term for means for calculating feature quantities that are smoke detection determination elements for each detection region. As shown in FIG. 11, the detection unit 23 includes a luminance average variance calculation unit 41 and a determination unit 42.
The luminance average variance calculating means 41 calculates and calculates the average luminance and variance of the entire detection area for each set detection area size (large detection area, small detection area) each time an image is captured. The calculated average luminance and variance of the detection area are determined by the determination means 42. That is, the determination unit 42 determines whether the ratio between the calculated average luminance and the average luminance in the reference image is within a predetermined range for each detection region. Further, for each detection region, it is determined whether or not the calculated variance is within a predetermined range with respect to the variance of the reference image.

The detection unit 23 includes an average luminance frequency analysis unit 43.
Each time an image is captured, the average luminance frequency analysis means 43 calculates the average luminance of the latest image data in the detection area for each set detection area size (large detection area, small detection area), For example, the average luminance of the detection area of 63 pieces of image data stored in the image storage unit 11 is calculated, and the average luminance frequency spectrum is generated by frequency analysis of the 64 average luminances. And the determination means 42 determines whether the intensity | strength of an average luminance frequency spectrum is more than predetermined intensity | strength. Further, the determination unit 42 divides the average luminance frequency spectrum into a plurality of frequency bands, and determines whether or not the integrated value of the intensity in the low frequency band is larger.

  FIG. 12 shows an example of a frequency spectrum related to the time series data of the average brightness in the large or small detection area. When smoke does not exist in the detection area, the intensity does not exceed a predetermined threshold as shown in FIG. FIG. 12B shows a case where smoke is present in the detection region. Here, two frequencies (2 Hz and 8 Hz) are used for the frequency, and the frequency is divided into three frequency bands, low, medium and high. It can be seen that the integrated value of the intensity increases as the frequency is low. On the other hand, the artificial light source has a large middle or high frequency band, and by detecting the characteristics of this frequency, it is possible to distinguish the periodic fluctuation of the artificial light source from smoke.

The detection unit 23 includes a light / dark luminance value calculation unit 44, a transmittance calculation unit 45, and a convergence luminance value calculation unit 46. The brightness / darkness brightness value calculation means 44 has the same processing procedure as the brightness / darkness brightness value calculation means 28 of the initial setting means 21.
The difference is that the image to be processed is not the reference image data but the latest image data, and each time the latest image data is captured, the size of the detection area (large detection area, small detection area) is different. In addition, the average value of the bright luminance group (bright luminance value) and the average value of the dark luminance group (dark luminance value) are calculated.

  Each time the latest image data is captured, the transmittance calculation unit 45 calculates the transmittance by the equation (1) for each set detection area size (large detection area, small detection area).

Transmittance = (current bright luminance value−current dark luminance value) / (bright luminance value of reference image−dark luminance value of reference image) (1)
In the calculation of the transmittance, only one of the bright luminance value and the dark luminance value may be used, and the transmittance may be calculated from the ratio between the current and the reference image.

  Every time the latest image data is fetched, the convergence luminance value calculation means 46 calculates a convergence luminance value by the equation (2) for each set detection area size (large detection area, small detection area).

    Convergence luminance value = {current bright luminance value− (reference bright luminance value × transmittance) / (reference bright luminance value−reference dark luminance value) (2)

The transmittance represented by the formula (1) becomes small when smoke enters the detection region. This is because the difference between the bright luminance value and the dark luminance value when smoke enters is smaller than that before the smoke enters, that is, the difference between the bright luminance value and the dark luminance value of the reference image. However, the decrease in transmittance may also occur when the illumination is dark.
The convergence brightness value represented by the expression (2) increases or decreases with respect to the convergence brightness value before smoke enters, that is, the reference image, when smoke other than black smoke, which is extremely black, enters the detection region. It does not decrease below. On the other hand, when the illumination becomes dark, it decreases below a predetermined threshold. By utilizing this characteristic, it is possible to identify smoke that is close to white and that the illumination has become dark.

  The determination unit 42 determines whether or not the calculated transmittance is equal to or less than a predetermined value. The determination unit 42 determines whether or not the calculated convergence luminance value is equal to or greater than a predetermined threshold value.

In the determination means 42 of the detection means 23, the ratio of the average luminance of the latest image to the average luminance of the reference image is within a predetermined range, and the variance is within a predetermined range relative to the variance of the reference image, and It is determined that the possibility of smoke generation is high when the intensity of the average luminance frequency spectrum is equal to or higher than a predetermined intensity or the integrated value of the intensity of the low frequency band is large.
Or, when the intensity of the average luminance frequency spectrum is higher than the predetermined intensity or the integral value of the intensity of the low frequency band, the transmittance is lower than the predetermined threshold, and the convergence luminance value is higher than the predetermined threshold, the generation of smoke Judged as likely.

As described above, in the present embodiment, a plurality of calculation means 41 to 46 for detecting smoke are provided. In particular, the smoke detection determination element includes a transmittance, a convergence luminance value, a frequency spectrum, an average value of average luminance, and a variance of average luminance, and the calculated values thereof are feature quantities of the smoke detection determination element.
These calculation means use all, and when all the calculation means satisfy the condition of the determination means 42, it is determined that smoke has been generated in the detection region. It becomes possible. However, these two computing means can be used in combination of two or more as appropriate, and can sufficiently improve the accuracy of smoke detection compared to a single computing process, such as a change in the amount of light of an artificial light source or a moving object. Can be mistakenly recognized as smoke.

  Returning again to FIG. 6, the smoke determination unit 24 determines whether or not there is smoke in the image based on the number of detection areas with smoke stored in the detection number storage unit 16. The calculation means 41 to 46 of the detection means 23 determine the presence or absence of smoke for each detection area by the determination means 42, and the number of detection areas in one screen determined to be smoke generation is stored in the detection number storage unit 16. ing. Then, the smoke discrimination means 24 reads out the number of detection areas determined to have smoke in the previously captured image from the detection number storage unit 16 as time series data, and the number of detections for a certain period exceeds a predetermined upper limit value. When it continues for a predetermined number of times, in other words, when the density of the detection area determined to have smoke in the image becomes a predetermined value or more, it is determined as smoke.

In FIG. 6, a reference image update unit 25 is a unit for periodically updating reference image data serving as a reference. This means reads the latest image data stored in the image storage unit 11 and the reference image data stored in the reference image storage unit 12, and uses these both images to obtain the following equation (3). Based on this, update reference image data is created.
When the reference image data is updated or replaced, the reference image update unit 25 calculates a certain evaluation value, and determines whether to update the reference image data or replace it with the latest image based on the evaluation value. It has a discrimination means. That is, the reference image update unit 25 has an evaluation value calculation unit (not shown), and the evaluation value calculation unit will be described.
The evaluation value calculation means calculates two evaluation values (first evaluation value and second evaluation value). The first evaluation value is obtained from the difference between the reference image data and the latest image data by the difference means and the total amount of luminance difference of the entire screen. In addition, after reading the reference image data and the latest image data, the image is divided into two regions (zones) up and down by a dividing unit that divides the screen into an upper region and a lower region. For each region, the total amount of difference in luminance between the latest image data and the reference image data is calculated, and the value is used for the calculation of the second evaluation value. That is, the ratio of the total amount of luminance difference in the upper region to the total amount of luminance difference in the lower region is calculated as the second evaluation value.
Note that the first evaluation value can be calculated by calculating the total amount of differences between the latest image and the reference image for each upper and lower region and calculating the total amount of differences on the entire screen from the sum of the total amounts of these differences. Good. In this way, the evaluation value calculation means calculates the difference from the reference image for each of the upper and lower areas divided by the dividing means, and calculates the total amount of differences for each area as an evaluation value.

The reference image replacement unit 26 determines that the camera 2 has shifted when the first evaluation value is equal to or greater than a predetermined threshold and the second evaluation value is within a predetermined range. Then, the reference image replacement means 26 replaces the reference image data with the latest image data and stores it in the reference image storage unit 12.
Further, the determining means determines that the camera 2 has not shifted when the first evaluation value is less than the predetermined threshold value or the second evaluation value is outside the predetermined range. Then, the reference image update unit 25 updates the reference image data with the image data calculated by the equation (3) using the reference image data and the latest image data, and stores the updated reference image data in the reference image storage unit 12.

    Update reference image data = reference image data × α + latest image data (1−α) (3)

Hereinafter, update or replacement of the reference image will be described using a specific example.
In order to reflect changes due to the incidence of light from the window or the like and the degree of illumination, the reference image is updated by the IIR filter every time the reference image is updated, for example, every few seconds to several tens of seconds. In this update, since the latest image data is slightly different from the reference image data, the update coefficient is α and the latest image data is reflected in the reference image data by Expression (3). α is about 0.05 to 0.5.

FIG. 13A shows the latest image captured when the camera 2 is misaligned and superimposed on the reference image.
The reference image data updated by the IIR filter can be compared with the latest image data as long as the light gradually changes such as incident light and illumination brightness. However, when a shift occurs in the camera 2 by mistake, as shown in FIG. 13A, there are many areas that have changed with respect to the luminance of the reference image data and the latest image data. That is, if the difference between the two images is obtained by the difference means, the difference amount is large, and although there is no smoke generated, there is a possibility of erroneous detection and accurate smoke detection cannot be performed.

When the camera 2 is mistakenly shifted, the total difference in luminance between the reference image data and the latest image data is increased. However, the total difference in luminance between the reference image data and the latest image data is May be larger.
Here, FIG. 14A is an example in which the total amount of differences regarding luminance increases when there is no shift in the camera 2, and in particular, a case where a change occurs in the upper region of the image. FIG. 14B is another example in which the total amount of difference regarding luminance increases when there is no deviation in the camera 2, and in particular, a case where a change occurs in the lower region of the image.
FIG. 14A shows an image when smoke is generated, and the generated smoke drifts toward the upper side of the screen. At this time, the evaluation value is calculated by dividing the image vertically.
First, the first evaluation value, which is the total amount of luminance difference of the entire screen due to the difference between the latest image data and the reference image data, is larger than a predetermined threshold value. However, since only the total difference in the brightness of the upper area where the smoke drifts is large and there is almost no difference in the difference in the lower area, the total difference in the brightness of the upper area relative to the total difference of the brightness in the lower area The ratio (second evaluation value) takes a large value and falls outside the predetermined range.

  FIG. 14B is an image when a person enters, and the person is leaning toward the lower side of the screen. At this time, the evaluation value is calculated by dividing the image into upper and lower parts. First, the first evaluation value, which is the total amount of luminance difference of the entire screen due to the difference between the latest image data and the reference image data, is larger than a predetermined threshold value. However, since only the total amount of difference in brightness in the lower area where people entered is large and there is almost no difference in the difference in the upper area, the difference in brightness in the upper area relative to the total difference in brightness in the lower area The ratio of the total amount (second evaluation value) takes a small value and falls outside the predetermined range.

On the other hand, when a shift occurs in the camera 2 by mistake, the shift occurs on the entire screen as shown in FIG. That is, there are regions that change in both the upper area in FIG. 13B and the lower area in FIG. 13C that are divided vertically by the screen dividing means. Therefore, as in the case of FIG. 14, the first evaluation value, which is the total amount of luminance difference of the entire screen due to the difference between the latest image data and the reference image data, is larger than a predetermined threshold value.
The ratio of the total difference in the brightness of the upper area to the total difference in the brightness of the lower area (second evaluation value) is such that the total difference in the upper area and the lower area is increased. Within a predetermined range.

  In this way, every time it comes time to update the reference image, the image is divided into two upper and lower regions, and the total amount of difference in brightness between the latest image data and the reference image data is calculated for each zone, The first evaluation value is the sum of the total differences related to the brightness of the upper and lower areas (= total difference of the entire screen), and the ratio of the total brightness difference of the upper area to the total difference of the brightness of the lower area is 2 Calculated as an evaluation value. Then, when the first evaluation value is equal to or greater than a predetermined threshold and the second evaluation value is within a predetermined range, it is determined that the camera 2 is displaced, and the reference image data is replaced with the latest image data. In other words, the latest image data is used as the reference image as it is without using the previous reference image. For this reason, even if the camera 2 is displaced, it is possible to prevent erroneous detection and to perform accurate smoke detection.

  In the present embodiment, the total amount of difference regarding luminance is calculated by using the upper part and lower part of the image as they are as two areas, but a part of the upper part and the lower part is defined as two areas. You may calculate the total amount of the difference regarding a brightness | luminance.

  Next, a fire detection procedure performed by the smoke detection device will be described with reference to FIGS. When the fire detection procedure starts, as shown in FIG. 15, an image storage procedure (S101) is executed, followed by a smoke detection procedure (S102), a fire discrimination procedure (S103), and a reference image update procedure (S104). Thus, one fire detection procedure is completed, and the fire detection procedure is repeatedly executed. Here, the fire is determined by detecting the smoke generated at the time of the fire.

First, the image storage procedure will be described with reference to the flowchart of FIG.
When the image storage procedure is started, it is determined in step S201 whether or not image data has been input from the camera 2. If not, step S201 is repeated, and if it has been input, the process proceeds to step S202.
In step S202, it is determined whether or not the image capture time has arrived. If it has arrived, the process proceeds to step S203, and if not, the process returns to step S201.
In step S203, the input image data is set as the latest image data, and the latest image data is subjected to luminance correction by inverse gamma correction.
In step S204, the latest image data whose luminance has been corrected is stored in the current image storage unit 12 by FIFO.

Next, the smoke detection procedure will be described with reference to the flowcharts of FIGS.
In the smoke detection procedure, the procedure shown in the flowcharts of FIGS. 17 and 18 is performed for all detection regions. That is, first, in the small detection area, based on the calculation means 41 to 47, the average brightness, variance, frequency analysis, bright / dark brightness value, transmittance, and convergence brightness value of the detection area are calculated, and the determination means 42 Thus, the presence or absence of smoke in the detection area is determined.
If it is determined that there is smoke in each determination element, the corresponding flag is set to 1, and if it is determined that there is no smoke, the flag is set to 0. This procedure is sequentially performed for all the small detection areas, and when the procedure is completed for all the small detection areas, the size of the detection area is increased and the same procedure is performed for the large detection areas. In other words, in the large detection area, based on the calculation means 41 to 47, calculation is performed on the average luminance, variance, frequency analysis, bright / dark luminance value, transmittance, and convergence luminance value of the detection area, and the determination means 42. Thus, the presence / absence of smoke in the detection region is determined. However, in this case, since the large detection area is composed of four small detection areas, the calculation is performed using the calculation value calculated in the small detection area. For example, the average luminance value in a certain large detection area is obtained by adding the average luminance values of four small detection areas and dividing by four. As described above, by calculating the small detection area in advance, using the calculated value to obtain the calculation value of the determination element in the large detection area, the calculation process becomes simple and the processing time is short. The calculation can be completed.
When the smoke detection procedure is started, in step S301, a small detection area having a small area size is set as a detection area, and the detection area flag is set to 1. It also resets the smoke detection register.
In step S302, the setting of various flags that leave the determination result is set to zero.
In step S303, one detection area that has not yet undergone smoke detection processing is designated based on the information stored in the detection area storage unit 14 of the detection area corresponding to the set detection area flag value.

In step S <b> 304, the latest image data is read from the image storage unit 11, the average luminance and variance of the designated detection area are calculated, and stored in the luminance average variance storage unit 15.
In step S305, the luminance average and variance stored in each of the most recent multiple smoke detection procedures including the latest data of the specified detection area from the luminance average variance storage unit 15 are read, and the average luminance moving average and Calculate the moving average of the variances.
In step S306, when the moving average of the average brightness varies within a predetermined range with respect to the average of the reference image, 1 is set to the average flag.
In step S307, the dispersion flag is set to 1 when the moving average of the luminance dispersion decreases within a predetermined range with respect to the dispersion of the reference image.

In step S308, the image data stored in each of the most recent image storage procedures including the latest image data of the designated detection area from the image storage unit 11 is read out, the average of the respective luminances is calculated, and the average A frequency spectrum is generated by frequency analysis of time series data relating to luminance.
In step S309, the frequency spectrum is corrected by excluding components around 0 Hz and 10 Hz (0 Hz when the shooting location is 60 Hz commercial power and 10 Hz when 50 Hz) from the frequency spectrum.
In step S310, the frequency spectrum is divided into three frequency band spectra by two frequencies of 2 Hz and 8 Hz, and when the integrated value of the intensity of each frequency band is larger as the lower frequency band, 1 is set in the frequency flag.

In step S311, as shown in FIG. 9, the luminances of the latest image data in the designated detection area are arranged according to the size, the bright luminance group is stored in the upper luminance register, and the dark luminance group is stored in the lower luminance. Store in register.
In step S312, the average of the brightness of the bright brightness group stored in the upper brightness register is obtained and used as the current bright brightness value, and the average of the brightness of the dark brightness group stored in the lower brightness register is obtained. Then, it is set as the current dark luminance value, and the process proceeds to step S313 via the node A.
In step S313, the reference bright luminance value and the reference dark luminance value of the specified detection area are read from the reference data storage unit 13, and the transmittance is calculated according to the equation (1). Further, the convergence luminance value is calculated according to the equation (2).
In step S314, when the transmittance is equal to or less than a predetermined value, 1 is set in the transmittance flag.
In step S315, when the convergence luminance value is greater than or equal to a predetermined threshold, 1 is set in the convergence luminance flag.
In step S316, it is determined whether 1 is set to the average flag, the dispersion flag, and the frequency flag, or 1 is set to the frequency flag, the transmittance flag, and the convergence luminance flag, and the average flag, the dispersion flag, and the frequency flag are set. If 1 is set, or if 1 is set in the frequency flag, the transmittance flag, and the convergence luminance flag, the process proceeds to step S317; otherwise, the process proceeds to step S318.
In step S317, the smoke detection register is incremented assuming that smoke may be generated, and the process proceeds to step S320.
In step S318, it is determined whether or not 1 is set in the frequency flag and the transmittance flag. If 1 is set in the frequency flag and the transmittance flag, the process proceeds to step S319. Otherwise, step S320 is performed. Proceed to
In step S319, a warning signal is issued to warn of the possibility of the occurrence of black smoke, and the process proceeds to step S320.

In step S320, it is determined whether or not all of the detection areas corresponding to the set detection area flag have been specified. If not, the process returns to step S303 via the node B, and if specified, step S321. Proceed to In other words, since a plurality of detection areas where smoke should be detected is set in the image, smoke detection is sequentially performed on these detection areas one by one.
In step S321, it is determined whether or not 1 is set in the detection area flag. When 1 is set in the detection area flag, the process proceeds to step S322. In step S322, the detection area flag is set to 0 and the node is set. The process returns to step S303 via B. At step 321, when the detection area flag is set to 0, the smoke detection procedure is terminated.
However, even if the smoke detection procedure for the small detection region is completed, the smoke detection procedure for the large detection region is not yet performed, so this time, in step S301, set large instead of small, select the large detection region, In the same procedure as described above, the presence / absence of smoke is determined for all large detection areas.

In the description so far, the steps (procedures) for determining whether or not there is smoke in one detection region in the image have been described. Next, a procedure for determining whether or not there is smoke in the entire image composed of a plurality of detection areas will be described.
Hereinafter, the fire discrimination procedure in the smoke discrimination means 24 will be described with reference to FIG.
In FIG. 20, it is assumed that a plurality of, for example, a plurality of nine detection areas A to I are set in the image. In this figure, at time t ₀ ~t _5, it shows how going smoke filled the room to be photographed by the camera.
The fire discrimination threshold in the smoke detection device is set to 6/9, for example. “9” indicates the number of detection areas, and “6” indicates the number of detection areas in the detection areas where smoke is detected by the smoke detection means and it is determined that there is smoke.

Now, at time t _2, only the detection region E is determined that there is smoke. Here, when determining fire, the total value of the number of detection areas determined to have smoke for three frames is used. That is, it is “2” because it is the total value of the detection areas with smoke for three frames from t _{0 to} t ₂ . Since “2” is smaller than the threshold value 6, it is not determined as a fire.
If the same calculation is made sequentially, “5” is obtained at t ₃ , “5” is obtained at t ₄ , and “6” is obtained at t ₅ (= 3 + 1 + 2). That is, in the case of such smoke generation would t ₅ Oite, fire is determined.
This fire discrimination method considers the number of detection areas with smoke in a certain time and the number of detection areas with smoke in a predetermined time, in other words, detection of presence of smoke in an image. When the density of the region (temporal and spatial density) exceeds a predetermined value, the generation of smoke is detected. If smoke always because its size (area) is changed, for example, it is determined that there is smoke in the detection region H at time t _1, at time t _2, the smoke is moving in the detection area E Thus, the detection region H may disappear. In addition, smoke always spreads spatially over time, and the number of detected areas tends to increase. Therefore, such a discrimination method is an effective discrimination method for detecting smoke. Conceivable.

  In addition, from the viewpoint of this temporal and spatial density, the discrimination method is used not only for the final fire discrimination but also for each step used in the smoke detection means, for example, judgment of transmittance, average value and threshold value. You may make it do. For example, instead of simply setting a flag when the transmittance falls below a predetermined value, the detected data for a plurality of, for example, four times, when three or more times are detected below the predetermined value, that is, there is a continuous decrease. When it is seen, the flag may be set for the first time.

Next, the fire determination procedure will be described with reference to the flowchart of FIG.
When the fire determination procedure is started, in step S401, the number set in the smoke detection register, that is, the number of detection areas determined to have smoke (hereinafter referred to as smoke detection number) is read and stored in the detection number storage unit 16. Remember.
In step S402, it is determined whether or not the latest smoke detection number is equal to or greater than a predetermined threshold (number). When the latest smoke detection number is equal to or greater than the predetermined threshold, the process proceeds to step S403. The determination procedure ends.
In step S403, the number of smoke detections stored in the detection number storage unit 16 is read for each of the most recent fire discrimination procedures.
In step S404, it is determined whether or not the value stored in the most recent predetermined number of fire determination procedures relating to the latest number of smoke detections greater than or equal to the predetermined number is continuously greater than or equal to a predetermined threshold value. If it is equal to or greater than the threshold value, the process proceeds to step S405, and otherwise, the fire determination procedure is terminated.
In this way, it is possible to determine whether there is a fire at a certain point in time when the number of detection areas determined as having smoke is greater than or equal to a predetermined value and spatially determining that there is smoke. The number of detection areas determined to be equal to or greater than a predetermined number of times in a continuous time, and when it can be determined that smoke is present even in time, that is, the number of detection areas with smoke within a predetermined period is high. Sometimes it is determined that a fire has occurred.
In step S405, it is determined that a fire has occurred, an alarm is issued, and the fire determination procedure is terminated.

Next, the reference image update procedure will be described with reference to the flowchart shown in FIG.
When the reference image update procedure is started, it is determined in step S501 whether or not the reference image update time has arrived. When the reference image update procedure has not arrived, the reference image update procedure is terminated, and when it has arrived, the process proceeds to step S502.
In step S502, the latest image data from the image storage unit 11 and the reference image data from the reference image storage unit 13 are read.
In step S503, the image is divided into two upper and lower areas, and the total amount of the difference in luminance between the latest image data and the reference image data is calculated for each area.
In step S504, the sum of the total difference regarding the luminance of the upper and lower areas is calculated as a first evaluation value, and the ratio of the total difference of the luminance of the upper area to the total difference regarding the luminance of the lower area is calculated as the second evaluation value. .
In step S505, it is determined whether or not the first evaluation value is greater than or equal to a predetermined threshold and the second evaluation value is within a predetermined range, and the first evaluation value is greater than or equal to the predetermined threshold and the second evaluation value is within the predetermined range. When the first evaluation value is less than a predetermined threshold value or the second evaluation value is outside the predetermined range, the process proceeds to step S507.
In step S506, the reference image replacement unit 26 replaces the reference image data with the latest image data and stores the reference image data in the reference image storage unit 13.
In step S507, the reference image update unit 25 updates the reference image data with the image data calculated by the expression (3) using the reference image data and the latest image data, and stores the updated reference image data in the reference image storage unit 13.

In step S508, a small detection area having a small area size is set as the detection area, and the detection area flag is set to 1.
In step S509, one detection region for which the brightness / darkness value is not calculated is designated based on the information stored in the detection region storage unit 14 of the detection region corresponding to the set detection region flag value.

In step S510, the reference image data of the designated detection area is read from the reference image storage unit 12, arranged in the order of the luminance levels of the pixels of the reference image data, and the bright luminance group is stored in the upper luminance register. The luminance group is stored in the lower luminance register.
In step S511, the average brightness of the bright brightness group stored in the upper brightness register is obtained and stored in the reference data storage unit 13 as a reference bright brightness value. In addition, the average of the luminance of the dark luminance group stored in the lower luminance register is obtained and stored in the reference data storage unit 13 as a reference dark luminance value.
In step S512, it is determined whether or not all the detection areas corresponding to the set detection area flag have been specified. If not, the process returns to step S509, and if specified, the process proceeds to step S513.
In step S513, it is determined whether or not 1 is set in the detection area flag. When 1 is set in the detection area flag, the process proceeds to step S514. In step S514, the detection area flag is set to 0 and step S509 is performed. Return to. If the detection area flag is set to 0 in step S513, it is determined whether or not a large detection area has been designated (step S515). If not, the process returns to step S508 to select the selected detection. The large detection area is selected by changing the area from small to large. Then, the procedure from step S508 to step S514 is repeated in the same manner. Finally, if the detection area flag is 1 in step S513, this time the reference image update procedure is terminated.

  In the above-described embodiment, the smoke detection device has been described as detecting smoke generated at the time of a fire. However, smoke detects smoke generated from, for example, a chimney, piping, plant equipment, or electronic equipment. You may do it.

It is a figure explaining the principle of smoke detection in this invention. It is a figure explaining the principle of smoke detection in this invention. It is a figure explaining the principle of smoke detection in this invention. It is a block diagram of the smoke detection apparatus which concerns on embodiment of this invention. It is various memory | storage parts ensured by the smoke detection apparatus which concerns on embodiment. It is a functional block diagram of the smoke detection apparatus which concerns on embodiment. It is a figure explaining setting two types of detection areas from which size differs. It is the image which image | photographed the smoke which generate | occur | produced in the location with a depth. It is a figure which shows a mode that the brightness | luminance value of a pixel in a detection area is arranged in order of a magnitude | size, and a bright luminance value and a dark luminance value are calculated | required from it. It is the image which image | photographed the water surface which reflects the light from the light source which the wave has occurred. It is a functional block diagram of the smoke detection means which concerns on embodiment. It is a graph of the frequency spectrum obtained by frequency-analyzing the time series data of the luminance average of the detection region. It is the figure which superimposed the reference image on the image when the shift | offset | difference occurred in the camera. These are two examples in which the total amount of difference regarding luminance increases when there is no shift in the camera. It is a flowchart which shows the fire detection procedure which concerns on embodiment. It is a flowchart which shows the image storage procedure which concerns on embodiment. It is a flowchart which shows the first half of the smoke detection procedure which concerns on embodiment. It is a flowchart which shows the second half of the smoke detection procedure which concerns on embodiment. It is a flowchart which shows the fire discrimination | determination procedure which concerns on embodiment. It is drawing explaining the method of the fire discrimination which concerns on embodiment. It is a flowchart which shows the reference | standard image update procedure which concerns on embodiment.

Explanation of symbols

  2 camera, 3 smoke detection device, 4 central processing unit (CPU), 5 ROM, 6 RAM, 7 built-in timer, 8 input / output interface (I / O), 9 frame grabber, 10 external storage device, 11 image storage unit, 12 reference image storage unit, 13 reference data storage unit, 14 detection area storage unit, 15 luminance average dispersion storage unit, 16 detection number storage unit, 21 initial setting unit, 22 image storage unit, 23 detection unit, 24 smoke discrimination unit, 25 reference image update means, 26 reference image replacement means, 27 detection area setting means, 28 brightness / darkness brightness value calculation means, 29 reference data calculation means, 31 brightness correction means, 32 current image storage means, 41 brightness average variance calculation means, 42 Determination means, 43 Average luminance frequency analysis means, 44 Bright / dark luminance value calculation means, 45 Transmittance calculation means, 46 Convergent luminance value Calculation means.

Claims (2)

The image obtained by photographing when it was confirmed that there was no smoke has a reference image storage unit for storing the image as a reference image, and was taken by performing image processing on the image obtained by photographing with the camera In the smoke detection device that detects the generation of smoke within the range,
A detection area setting means for setting a predetermined detection area in the latest image taken and in the reference image;
Bright brightness value calculation means for arranging the pixels in the detection region in order of brightness value, and extracting and averaging the brightness values of a predetermined number of pixels from the larger brightness value,
A transmittance calculating means for calculating a transmittance from a ratio between the bright luminance value of the reference image and the bright luminance value of the latest image;
A smoke detection apparatus comprising: a determination unit that determines the presence or absence of smoke in the detection region based on the transmittance. A dark luminance value calculating unit that arranges the pixels in the detection area in order of the luminance value, extracts a luminance value of a predetermined number of pixels from the smaller luminance value, and calculates an average as a dark luminance value,
The transmittance calculating means calculates a transmittance from a difference between the bright luminance value and the dark luminance value in the reference image and a ratio between the bright luminance value and the dark luminance value in the latest image. 2. The smoke detection device according to claim 1.

Images