JP2023034503A - Visible smoke determination device, program, visible smoke determination system, and visible smoke determination method - Google Patents

Abstract

To reduce erroneous detection without increasing detection failure when determining whether visible smoke is generated from a facility being a monitored object or not by image analysis.SOLUTION: A visible smoke determination device is provided which comprises: an area setting unit which sets visible smoke determination areas within a captured image obtained by image capture in a direction of a facility being an object monitored with respect to visible smoke generation; an index value calculation unit which calculates index values being indexes of visible smoke generation on the basis of pixel information of individual pixels within the visible smoke determination areas; a candidate detection unit which detects visible smoke determination areas having index values equal to or higher than a prescribed reference value, as candidate areas being candidates for visible smoke generation; a time measuring unit which measures excess times during which the index values of the candidate areas being the reference value or higher are reduced to values lower than the reference value; and a determination unit which determines that visible smoke is generated in candidate areas of which the excess times are shorter than a prescribed threshold and is not generated in candidate areas of which the excess times are equal to or longer than the prescribed threshold.SELECTED DRAWING: Figure 4

Description

The present invention relates to a visible smoke determination device, a program, a visible smoke determination system, and a visible smoke determination method.

In steelworks, as part of measures to protect the atmospheric environment, exhaust gases emitted from facilities such as coke ovens and sintering machines are treated so as to satisfy emission standard values and compliance values before being discharged. In addition, from the viewpoint of harmony with nature and to meet the demands of local residents, operations are being carried out to suppress the generation of visible smoke. However, for example, if coke combustion failure occurs in a coke oven for some reason, black smoke may be generated when the coke is pushed out. Therefore, it is important to immediately confirm such an event and take measures to prevent its recurrence.

Therefore, efforts have been made to determine whether visible smoke is generated by automatically performing image analysis on a captured image in which the equipment to be monitored is taken as a subject. As a method for detecting smoke by applying image analysis technology to surveillance video, for example, Patent Document 1 discloses that a variable image that changes due to sunlight, rainfall, etc. is pixel-decomposed, and the pixels of several images are each moved and averaged. Using the standard image as the original image, the real-time variation image is subtracted pixel by pixel, the unchanged image portion that constitutes the background portion is removed, and the area and brightness difference of the dust and smoke image composed of the variation image portion are set to be constant. A smoke or dust detector is disclosed with means for issuing an alarm when a value is exceeded.

In addition, in Patent Document 2, the transition of the brightness of each pixel obtained by moving the image over time is differentiated with time, and the first derivative obtained by the time differentiation produces a peak, and the peak value Disclosed is a smoke or dust detector equipped with a means for estimating that the image is a smoke or dust image when the ratio of the number of pixels in which V exceeds a given value exceeds a given value.

In Patent Document 3, when estimating that it is a dust and smoke emission image by the same means as in Patent Document 2, the rate of change of the first differential coefficient between the analysis target area and the part other than that part is compared with the steady state value. A smoke or dust detector is disclosed which has a means for estimating that the image is a dust and smoke image when the ratio between the two rates of change is two or more.

In Patent Document 4, when the lightness difference for each hue between the background area of the image and the visible smoke determination area is equal to or greater than a preset threshold in at least one hue, the imaged smoke is determined to be visible smoke. A smoke determination device is disclosed.

JP-A-4-352077 JP-A-4-358285 JP-A-5-10738 JP 2015-169618 A

However, the visible smoke detection methods incorporated in the devices of Patent Documents 1 to 4 all set a threshold value for the detection index obtained from the captured image, and when the set value is exceeded, it is determined as visible smoke. It is a way to If the threshold value is set high, even if visible smoke of extra-large scale can be detected, it is difficult to detect visible smoke smaller than that. On the other hand, if the threshold is set low, the possibility of improving the detectability of visible smoke increases, but at the same time, false detection due to noise such as clouds increases. In this way, there is a trade-off between non-detection and false detection, and it has been difficult to reduce them at the same time.

Therefore, the present invention has been made in view of the above problems. An object of the present invention is to provide a visible smoke determination device, a program, a visible smoke determination system, and a visible smoke determination method capable of reducing false detections without increasing detection.

In order to solve the above problems, according to one aspect of the present invention, a region setting unit for setting a visible smoke determination region in a captured image obtained by capturing an image in the direction of a facility to be monitored for generation of visible smoke; , an index value calculation unit that calculates an index value as an index of visible smoke generation based on pixel information of each pixel in the visible smoke determination region; A candidate detection unit that detects a candidate area that is a candidate for visible smoke generation, a time measurement unit that measures the excess time from when the index value becomes equal to or greater than the reference value until it becomes less than the reference value for the candidate area, and the excess time a determination unit that determines that visible smoke is generated in a candidate region in which is less than a predetermined threshold, and determines that visible smoke is not generated in a candidate region in which the excess time is equal to or greater than the predetermined threshold , a visible smoke determination apparatus is provided.

The region setting unit further sets a background region in the captured image, and the index value calculation unit calculates the sum of pixel information of each pixel in the visible smoke determination region and the background region set in the captured image. The value obtained by dividing by the number of pixels in the area is calculated as the representative pixel information representing the area, and the index value is the difference value obtained by subtracting the representative pixel information of the visible smoke determination area from the representative pixel information of the background area. or a value obtained by subtracting the moving average value of the difference value from the difference value.

The region setting unit sets, as a visible smoke determination region, a region in which visible smoke generated from the equipment to be monitored is captured based on a plurality of captured images captured within a predetermined period of time in the past. A region that is not entirely covered may be set as a background region.

The captured image is an RGB color system image with 256 gradations, and the index value calculation unit calculates the pixel information of each pixel using a function represented by the red, green, and blue components of the pixel. good.

The pixel information may be information representing luminance or lightness.

The visible smoke determination device may include a threshold setting unit that sets at least one of the reference value of the index value and the threshold of the excess time. The threshold value setting unit determines the ratio of the number of visible smoke determined by the visible smoke determination device to the number of visible smoke that can be visually confirmed by a person using a plurality of captured images captured within a predetermined period of time in the past. The reproducibility is set to a predetermined ratio or more. The relevance rate, which is determined by the ratio of the number of visible smoke that can be visually confirmed by a person to the number of candidate areas detected by the visible smoke determination device, is set to a predetermined ratio or more. Alternatively, at least one of the reference value of the index value and the threshold value of the excess time may be set by maximizing the F value, which is the harmonic average of the recall rate and the precision rate.

The determination unit may, when determining that visible smoke is generated, instruct the alarm device to report that visible smoke is generated.

The monitored equipment may be a coke oven.

At this time, the visible smoke determination device may include a storage unit that associates and stores the coke extrusion time and the kiln number of the coke oven each time coke is pushed out in the coke oven. When the candidate detection unit detects a visible smoke determination region in which the index value is equal to or greater than the reference value, if the coke extrusion time that coincides with the detection time at which the index value becomes equal to or greater than the reference value is recorded in the storage unit records the detection time in the storage unit in association with the coke extrusion time and the kiln number of the coke oven, and when the determination unit determines that visible smoke is generated in the candidate area, to the alarm device , the generation of visible smoke, and the notification of the coke extrusion time and coke oven number recorded in the storage unit in association with the detection time of the candidate area may be instructed.

Moreover, in order to solve the above problems, according to another aspect of the present invention, there is provided a program that causes a computer to function as the visible smoke determination device described above.

Furthermore, in order to solve the above problems, according to another aspect of the present invention, an imaging device that acquires a captured image by capturing an image in the direction of a facility that is a target for monitoring generation of visible smoke, and based on the captured image, a visible smoke determination device that determines whether visible smoke is generated from a facility to be monitored, wherein the visible smoke determination device includes an area setting unit that sets a visible smoke determination area in a captured image; An index value calculation unit that calculates an index value as an index of visible smoke generation based on pixel information of each pixel in the region; A candidate detection unit that detects a candidate area as a candidate, a time measurement unit that measures the excess time from when the index value becomes equal to or greater than the reference value until the candidate area becomes less than the reference value, and the excess time is a predetermined threshold value. a determination unit that determines that visible smoke is generated in a candidate area that is less than the specified threshold, and determines that visible smoke is not generated in a candidate area that is equal to or greater than a predetermined threshold. A system is provided.

Further, in order to solve the above problems, according to another aspect of the present invention, an area for setting a visible smoke determination area in a captured image obtained by capturing an image in the direction of a facility to be monitored for generation of visible smoke a setting step; an index value calculating step of calculating an index value serving as an index of visible smoke generation based on pixel information of each pixel in the visible smoke determination area; and a visible smoke determination step in which the index value is equal to or greater than a predetermined reference value. a candidate detection step of detecting an area as a candidate area for generating visible smoke; and a time measurement step of measuring the excess time from when the index value of the candidate area becomes greater than or equal to the reference value to when it becomes less than the reference value. a determination step of determining that visible smoke is generated in a candidate region in which the excess time is less than a predetermined threshold, and that no visible smoke is generated in a candidate region in which the excess time is equal to or greater than the predetermined threshold; A visible smoke determination method is provided, comprising:

As described above, according to the present invention, when the index value, which is an index of visible smoke generation, exceeds the reference value and the time period during which the index value exceeds the reference value is less than the predetermined threshold value, the detection target of visible smoke. As a result, it is possible to reduce erroneous detection without increasing undetected when judging whether or not visible smoke is generated from the equipment to be monitored by image analysis.

FIG. 3 is a schematic diagram showing an example of a captured image of the sky above a facility to be monitored for generation of visible smoke. FIG. 10 is a schematic diagram showing an example of a result of performing visible smoke determination on a captured image using a brightness difference between inside and outside of a visible smoke determination region in the captured image as an index value; FIG. 10 is a schematic diagram showing an example of a result of performing visible smoke determination on a captured image using an index value representing temporal change in luminance difference between inside and outside of a visible smoke determination region in the captured image; 1 is a functional block diagram showing a configuration example of a visible smoke determination system including a visible smoke determination device according to an embodiment of the present invention; FIG. FIG. 4 is a schematic diagram showing an example of setting a visible smoke determination region and a background region; It is a flowchart which shows the visible smoke determination method which concerns on the same embodiment. It is a block diagram which shows an example of the hardware constitutions of the information processing apparatus which functions as a visible smoke determination apparatus which concerns on the same embodiment. FIG. 10 is a schematic diagram showing an example of transition of an index value representing temporal change in brightness difference between inside and outside of a visible smoke determination region in a captured image and an example of visible smoke detection results in a certain monitoring time zone in the embodiment.

Preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings. In the present specification and drawings, constituent elements having substantially the same functional configuration are denoted by the same reference numerals, thereby omitting redundant description.

[1. overview]
First, based on FIGS. 1 to 3, the outline of visible smoke determination by the visible smoke determination device according to one embodiment of the present invention will be described. FIG. 1 is a schematic diagram showing an example of a captured image of the sky above a facility to be monitored for generation of visible smoke. FIG. 2 is a schematic diagram showing an example of the result of performing visible smoke determination on a captured image using the brightness difference between the inside and outside of the visible smoke determination region in the captured image as an index value. FIG. 3 is a schematic diagram showing an example of the result of performing visible smoke determination on a captured image using an index value representing a temporal change in luminance difference between the inside and outside of the visible smoke determination region in the captured image.

The visible smoke determination apparatus according to the present embodiment focuses on the difference in the time change of smoke that changes in a relatively short time with respect to clouds that change slowly, and the index value, which is the index of visible smoke generation, exceeds the reference value. Visible smoke to be detected is determined when the overtime period during which the index value exceeds the reference value is less than a predetermined threshold value.

The captured image is an image obtained by capturing the direction of the equipment (equipment or above the equipment, etc.) that is the monitoring target for the generation of visible smoke. For example, the captured image 10 shown in FIG. 1 is an image of a factory equipped with various monitored equipment 5 taken from a distance, and includes the monitored equipment 5 (including a plurality of chimneys 3) behind the trees 7, The sky above it is shown. Although not directly visible in the captured image 10 , there are also a plurality of monitoring target facilities 5 hidden behind trees 7 . When visible smoke is generated from the monitoring target equipment 5 in the factory, the visible smoke 9 is also captured in the captured image 10 as shown in FIG. Note that the visible smoke 9 shown in FIG. 1 is an example of smoke generated from the monitored equipment 5 hidden behind the trees 7 .

When detecting visible smoke 9, which is a detection target, by performing image processing on a captured image 10 as shown in FIG. Using the indicated index value, visible smoke is detected when the index value exceeds a reference value. At this time, if the reference value is lowered in order to reduce undetected visible smoke, noise such as clouds may be erroneously detected as visible smoke, resulting in lower detectability.

Here the smoke changes in a short time relative to the slowly changing cloud. Therefore, the visible smoke determination device according to the present embodiment detects that the index value exceeds the reference value, and measures the time (exceeding time) from when the index value exceeds the reference value to when it becomes less than the reference value. Distinguish between smoke and clouds. As a result, even when the reference value is set to be low, it is possible to detect smoke mixed with clouds and to reduce erroneous detection due to clouds.

For example, FIG. 2 shows an example of the result of performing visible smoke determination on a captured image using the luminance difference between the inside and outside of the visible smoke determination region in the captured image as an index value (hereinafter referred to as “index value A”). . Here, the visible smoke determination area is an area set for determining whether or not visible smoke is generated from the facility to be monitored. FIG. 2 shows the index value A (luminance difference) and the reference value. However, the reference value may be a moving average of luminance differences inside and outside the visible smoke determination region. It should be noted that, among the captured images image-processed at each time, some of the captured images include thin clouds that do not completely cover the sky (background) (hereinafter also referred to as “clouds a”), rain clouds and the sky ( background) (hereinafter also referred to as “cloud b”), large smoke such as black smoke (hereinafter also referred to as “smoke a”) or thin small smoke (hereinafter also referred to as “smoke b” say.) is in the picture.

In the conventional method, visible smoke is determined when the index value A (luminance difference) exceeds the reference value, so cloud a, cloud b, and smoke a are determined as visible smoke. On the other hand, according to the method according to the present embodiment, among the clouds a, clouds b, and smoke a, only the clouds a and smoke a for which the time over which the index value exceeds the threshold value are determined to be visible smoke. is not determined as visible smoke. That is, the method according to the present embodiment can eliminate erroneous detection of the cloud b by the conventional method. The smoke b is not detected by both the conventional method and the method according to the present embodiment. As described above, according to the method according to the present embodiment, it is possible to detect smoke mixed with clouds even if the reference value is set low, and to reduce false detections due to clouds. Reduce false positives without increasing

As another example, FIG. 3 shows the visible smoke determination of the captured image as an index value (hereinafter referred to as "index value B") representing the temporal change in the luminance difference between the inside and outside of the visible smoke determination region in the captured image. An example of the result of implementation is shown. Here, the index value B is a value obtained by subtracting the moving average of the luminance difference from the luminance difference. FIG. 3 shows an index value B (luminance difference-moving average of luminance difference) and a reference value. The captured image used for the visible smoke determination in FIG. 3 is the same as the captured image used for the visible smoke determination in FIG. Here, the moving average is a value obtained by smoothing time-series data, and includes, for example, a simple moving average, a weighted moving average, and an exponential moving average.

In the conventional method, visible smoke is determined when the index value B (luminance difference-moving average of luminance difference) exceeds a reference value, so cloud b, smoke a, and smoke b are determined as visible smoke. On the other hand, according to the method according to the present embodiment, among the clouds b, the smoke a, and the smoke b, only the smoke a and the smoke b for which the time over which the index value B exceeds the threshold is determined to be visible smoke. A cloud a in which the value B does not exceed the reference value and a cloud b in which the time over which the index value B exceeds the threshold is not determined as visible smoke. As described above, when the index value B is used in the method according to the present embodiment, visible smoke can be determined with higher accuracy than when the index value A is used. can be further reduced.

The visible smoke determination method according to this embodiment will be described in detail below.

[2. Visible Smoke Judgment System]
First, based on FIG. 4, the visible smoke determination system 1 including the visible smoke determination device 100 according to the present embodiment will be described. FIG. 4 is a functional block diagram showing a configuration example of the visible smoke determination system 1 including the visible smoke determination device 100 according to this embodiment.

The visible smoke determination system 1 is composed of an imaging device 50, a visible smoke determination device 100, and an alarm device 200, as shown in FIG. Further, the visible smoke determination system 1 may be configured such that the information processing device 300 and the monitored object management device 400 can transmit and receive information to and from the visible smoke determination device 100 .

[2-1. Imaging device]
The image capturing device 50 captures an image of the direction of the equipment that is the target of monitoring for the generation of visible smoke. Note that the imaging range of the imaging device 50 does not necessarily include the equipment to be monitored, and includes an area where visible smoke appears when visible smoke is generated from the equipment to be monitored (for example, above the equipment). as long as it is In other words, the captured image does not need to show the facility to be monitored so that it can be visually confirmed, but only needs to show visible smoke generated from the facility. The imaging range of the imaging device 50 may include only one facility to be monitored, or may include a plurality of facilities to be monitored.

The imaging device 50 continuously captures images of the imaging range to acquire captured images. The imaging device 50 may, for example, capture the imaging range as a moving image. In this case, each frame of the moving image becomes a captured image subjected to image processing by the visible smoke determination device 100 . Alternatively, the imaging device 50 may continuously capture the imaging range at predetermined intervals. For example, the imaging device 50 may acquire one still image every three seconds. The imaging device 50 may be a color camera that acquires color images, or a monochrome camera that acquires monochrome images. The captured image of the imaging device 50 is output to the visible smoke determination device 100 .

[2-2. Visible smoke determination device]
The visible smoke determination device 100 performs image processing on the image captured by the imaging device 50 and determines whether visible smoke is generated from the equipment to be monitored. Visible smoke determination device 100 includes region setting section 110 , index value calculation section 120 , candidate detection section 130 , determination section 140 , time measurement section 150 , threshold setting section 160 , and storage section 170 .

(Area setting part)
The region setting unit 110 sets a visible smoke determination region within the captured image. The visible smoke determination area is an area set for determining whether or not visible smoke is generated from the equipment to be monitored, and is set for each equipment to be monitored, between equipment, or above the equipment. For example, as shown in FIG. 5, the area setting unit 110 sets visible smoke determination areas A to M between and above the equipment 5 (including the chimney 3) to be monitored in the captured image 10. good too. When the equipment 5 to be monitored is a coke oven, the visible smoke determination region is set above the coke oven equipment, for example. Although the visible smoke determination area is a rectangular area in FIG. 5, the present invention is not limited to such an example, and other shapes may be used. Also, when setting a plurality of visible smoke determination regions, the shape and size of each visible smoke determination region may be the same or different.

Also, the area setting section 110 may further set a background area X in the captured image. The background area X is an area that is not entirely covered with the generated visible smoke even if visible smoke is generated from the equipment to be monitored. In order to improve the accuracy of determination of visible smoke generation by the visible smoke determination device 100, the background region X is desirably wider than the visible smoke determination region. For example, the area setting unit 110 may set a background area X above the visible smoke determination areas A to M in the captured image 10, as shown in FIG.

The shape, size, position and number of the visible smoke determination areas A to M and the background area X are empirically set based on the positions at which visible smoke appeared in captured images acquired in the past. For example, the area setting unit 110 does not necessarily need to set the visible smoke determination area above the equipment 5 (including the chimney 3). A region in which visible smoke generated from the facility has been imaged may be set as the visible smoke determination region, and a region not entirely covered by visible smoke may be set as the background region.

After setting the visible smoke determination region in the captured image, the region setting unit 110 outputs information about the region set in the captured image (position information of each region in the captured image) to the index value calculation unit 120 .

(Index value calculation unit)
The index value calculation unit 120 calculates an index value as an index of visible smoke generation based on the pixel information of each pixel within the visible smoke determination region.

Pixel information is, for example, information representing luminance or lightness. When the captured image is a grayscale monochrome image, that is, a multi-gradation monochromatic image, the gradation of the pixel becomes the image information of the pixel as it is. When the captured image is a color image, the index value calculator 120 calculates pixel information of each pixel using a function represented by the red, green, and blue components of the pixel. For example, when the captured image is an RGB color system image with 256 gradations, pixel information based on luminance and pixel information based on brightness of each pixel can be expressed as follows.

Pixel information based on luminance Red component x 0.21 + green component x 0.72 + blue component x 0.07
Image information based on brightness = {(max (red component, green component, blue component)) + (min (red component, green component, blue component))}/2

The index value calculator 120 may generate representative pixel information representing the area based on the pixel information of each pixel in the area set in the captured image. The representative pixel information is representative pixel information representing characteristics of pixel information of each pixel in the area. When the luminance value is pixel information, the representative pixel information may be, for example, a value obtained by dividing the total sum of luminance values of pixels in an area by the number of pixels in the area (that is, an average luminance value). Also, when the brightness is pixel information, the representative pixel information may be, for example, a value obtained by dividing the sum of the brightness of each pixel in an area by the number of pixels in the area (that is, the average brightness).

Whether to use the luminance value or the lightness as the pixel information may be determined according to the characteristics such as the color of the smoke to be detected. In this embodiment, a case of using pixel information based on luminance will be described.

The index value calculation unit 120 calculates an index value, which is an index of visible smoke generation, using pixel information within a region set in the captured image. For example, the index value calculator 120 calculates the above-described index values A, B, etc. as indices of visible smoke generation.

For example, the index value calculator 120 may set a value obtained by multiplying the representative pixel information of the visible smoke determination region by -1 as the index value of the visible smoke determination region. In this case, the lower the representative pixel information of the visible smoke determination region, the higher the index value.

Further, the index value calculation unit 120 may use the time change value of the value obtained by multiplying the representative pixel information of the visible smoke determination region by -1 as the index value of the visible smoke determination region. The index value increases when the value obtained by multiplying the representative pixel information of the visible smoke determination area by -1 increases sharply.

Alternatively, the index value calculation unit 120 may use a value obtained by subtracting the moving average value of the value obtained by multiplying the representative pixel information of the visible smoke determination region by -1 as the index value of the visible smoke determination region.

In this way, when the index value is calculated only from the representative pixel information of the visible smoke determination area, the index value is influenced and changed by clouds or the like in the background of the visible smoke determination area. Therefore, in order to reduce the influence, there is also a method of subtracting the influence of the background area from the visible smoke determination area. For example, the index value calculation unit 120 may use the difference value obtained by subtracting the representative pixel information of the visible smoke determination region from the representative pixel information of the background region as the index value of the visible smoke determination region. In this case, if the representative pixel information of the visible smoke determination area is lower than the representative pixel information difference of the background area, the value of the index value will be high. As for the representative pixel information of the background area, the average luminance value, the average brightness, etc. can be used as in the case of the visible smoke determination area.

Further, the index value calculation unit 120 may use the time change value of the difference value obtained by subtracting the representative pixel information of the visible smoke determination region from the representative pixel information of the background region as the index value of the visible smoke determination region. The index value increases as the difference value increases rapidly.

Alternatively, the index value calculation unit 120 calculates the value obtained by subtracting the moving average value of the difference value from the difference value obtained by subtracting the representative pixel information of the visible smoke determination region from the representative pixel information of the background region, as the index value of the visible smoke determination region. may be For example, a value obtained by subtracting the moving average value of the difference values in the past one minute from the difference value at the determination time is used as the index value. At this time, when the pixel information difference value increases in several seconds, which is a short time shorter than one minute, the value of the index value increases. In this embodiment, a case will be described in which a value obtained by subtracting a moving average value of the difference value from the difference value is used as the index value. If the representative pixel information is not used, the same number of pixels are selected (for example, randomly selected) from the background region and the visible smoke determination region, respectively, and the sum of the difference values of each pixel (the sum of the absolute values) etc. may be used as index values.

Index value calculation section 120 outputs the index value calculated for each visible smoke determination region to candidate detection section 130 .

(candidate detector)
The candidate detection unit 130 detects, from the captured image, a visible smoke determination region having an index value equal to or greater than a predetermined reference value as a candidate region for generating visible smoke. The reference value may be a value set by a threshold value setting unit 160, which will be described later, or a value set by an observer via the information processing apparatus 300 based on past determination results. Details of the setting of the reference value by the threshold setting unit 160 will be described later.

The candidate detection unit 130 compares the index value of each visible smoke determination region calculated by the index value calculation unit 120 with a reference value, and detects visible smoke determination regions with an index value equal to or greater than a predetermined reference value as visible smoke generation. is detected as a candidate region that is a candidate for Upon detecting the candidate area, candidate detection section 130 notifies determination section 140 that the candidate area has been detected.

(Judgment part)
The determination unit 140 determines whether visible smoke is generated in the candidate area detected by the candidate detection unit 130 . When the candidate detection unit 130 notifies the determination unit 140 that the candidate area has been detected, the determination unit 140 informs the time measurement unit 150 (to be described later) that the index value of the candidate area becomes equal to or greater than the reference value and then becomes less than the reference value. Instruct to measure the time exceeding the standard value (the time during which the reference value is exceeded). Then, when the excess time during which the index value of the candidate region exceeds the reference value is input from the time measurement unit 150, the determination unit 140 compares the excess time with a predetermined threshold, and determines that the excess time is less than the threshold. Then, it is determined that visible smoke is generated in the candidate area. The threshold for the excess time may be a value set by the threshold setting unit 160, which will be described later, or may be a value set by the supervisor via the information processing device 300 based on past determination results. The details of the setting of the threshold for the excess time by the threshold setting unit 160 will be described later.

When the determination unit 140 determines that visible smoke is generated in the candidate area, the determination unit 140 instructs the notification device 200 to report the occurrence of visible smoke. Also, the determination unit 140 may record the determination result in the storage unit 170 .

(Time measurement part)
The time measurement unit 150 measures the excess time from when the index value becomes equal to or greater than the reference value to when it becomes less than the reference value for the candidate area detected by the candidate detection unit 130 . After starting to measure the excess time, the time measurement unit 150 acquires the index value of the candidate area calculated by the index value calculation unit 120 every time the captured image is input from the imaging device 50 to the visible smoke determination device 100, The index value and the reference value are compared. Then, when the index value of the candidate region for which the overtime is being measured becomes less than the reference value, the time measurement unit 150 outputs the measured overtime to the determination unit 140 . However, the time measurement unit 150 does not need to actually measure the elapsed time, and may obtain the excess time from information indicating the time when the input captured image was captured.

(Threshold setting unit)
The threshold setting unit 160 sets various thresholds based on past determination results. For example, the threshold setting unit 160 sets a reference value for the index value, a threshold for the excess time, and the like.

For example, the reference value of the index value is a value set to detect a candidate area where visible smoke is highly likely to occur based on the index value of the visible smoke determination area. The threshold setting unit 160 can set the reference value using past determination results recorded in the storage unit 170 . The storage unit 170 records time-series data of index values of each visible smoke determination region determined by the visible smoke determination device 100 for a plurality of captured images captured within a predetermined period of time in the past. In addition, the storage unit 170 records visible smoke determination areas and occurrence times of the confirmed visible smoke generated by the observer visually confirming visible smoke in these captured images.

For example, the threshold setting unit 160 sets the reproduction rate determined by the ratio of the number of visible smoke determined by the visible smoke determination device 100 to the number of visible smoke that can be visually confirmed by the observer so that the reproducibility is equal to or greater than a predetermined ratio. , a reference value for the index value may be set. This is a method of setting a reference value based on the detection rate of visible smoke that can be visually confirmed. In general, the lower the reference value, the higher the recall.

In addition, the threshold setting unit 160 sets the matching rate, which is determined by the ratio of the number of visible smoke that can be visually confirmed by the observer to the number of candidate areas detected by the visible smoke determination device 100, so that it is equal to or greater than a predetermined ratio. , a reference value for the index value may be set. This is a method of setting a reference value based on the ratio of what is determined to be visible smoke to actually be visible smoke. In general, the higher the reference value, the higher the matching rate.

Alternatively, the threshold setting unit 160 may set the reference value of the index value by maximizing the F value, which is the harmonic average of the recall and the precision. Since there is a trade-off relationship between the recall rate and the precision rate, if it is desired to set a reference value that is intermediate between them, the reference value may be set by such a method.

As described above, there are a plurality of methods for setting the reference value of the index value, and the method to be used may be selected according to the purpose. For example, the distance from the installation position of the imaging device 50 to the smoke generation source may differ in each visible smoke determination region. When the distance is long, compared to when the distance is short, even if the actual size of the smoke is the same, it appears relatively small within the visible smoke determination area of the captured image, so the change in the index value is also slow. Become. For example, if the reference value is set based on the reproducibility, the reference value when the distance from the installation position of the imaging device 50 to the smoke source is long is often smaller than the reference value when the distance is short. Therefore, it is preferable to set different reference values for each visible smoke determination region, but the same reference value may be set for all visible smoke determination regions.

The threshold setting unit 160 can also set the threshold for the excess time in the same manner as the reference value for the index value. For example, the threshold value setting unit 160 may set the threshold value of the excess time so that the recall determined by the ratio of the number of visible smoke determined by the visible smoke determination device 100 is equal to or greater than a predetermined ratio. In this method, the threshold value is set based on the detection rate of visible smoke that can be visually confirmed by an observer in the visible smoke determination area detected as a candidate area for generating visible smoke.

In addition, the threshold setting unit 160 sets the matching rate, which is determined by the ratio of the number of visible smoke that can be visually confirmed by the observer to the number of candidate areas detected by the visible smoke determination device 100, so that it is equal to or greater than a predetermined ratio. , a threshold for excess time may be set. This is a method of setting a threshold value based on the ratio of visible smoke determined by the visible smoke determining apparatus 100 to be actually visible smoke.

Alternatively, the threshold setting unit 160 may set the threshold for the excess time by maximizing the F value, which is the harmonic average of the recall and the precision. If the overtime threshold is set longer, the number of undetected visible smoke that originates from the candidate area decreases, resulting in a higher recall rate. do. On the other hand, if the overtime threshold is set shorter, the number of undetected visible smoke that originates from the candidate area increases, resulting in a lower recall rate. becomes higher. Thus, there is a trade-off between the recall of visible smoke and the relevance of visible smoke. Therefore, if it is desired to set a threshold that is intermediate between the recall rate and the precision rate, such a method should be used.

As described above, there are a plurality of methods for setting the threshold value of the excess time, and the method to be used may be selected depending on the purpose. As described above, the distance from the installation position of the imaging device 50 to the smoke generation source may differ in each visible smoke determination region. When the distance is long, compared to when the distance is short, even if the actual size of the smoke is the same, it appears relatively small within the visible smoke determination area of the captured image, so the overtime threshold is also short. Become. For example, if the threshold is set based on the reproducibility, the threshold when the distance from the installation position of the imaging device 50 to the smoke source is long is often shorter than the threshold when the distance is short. Therefore, it is preferable to set different threshold values for each visible smoke determination region, but the same threshold value may be set for all visible smoke determination regions.

The threshold value setting unit 160 outputs to the candidate detection unit 130 when the reference value of the index value is set, and outputs to the determination unit 140 when the threshold value of the excess time is set.

(storage unit)
The storage unit 170 is a storage unit that stores various types of information used by the visible smoke determination device 100 . The storage unit 170 stores, for example, time-series data of index values of each visible smoke determination region determined by the visible smoke determination device 100 for a plurality of captured images captured in the past, and the monitor's information about these captured images. Visually confirmed visible smoke determination areas where visible smoke is generated, time of generation, etc. are recorded.

Further, the storage unit 170 may record operation information input from the monitored object management device 400 that manages the operation of the equipment to be monitored. For example, when the equipment to be monitored is a coke oven, the monitoring object management device 400 outputs the coke extrusion time and the kiln number of the coke oven to the visible smoke determination device 100 every time coke is pushed out in the coke oven. and may be recorded in the storage unit 170 . At this time, when the candidate detection unit 130 detects a visible smoke determination region in which the index value is equal to or greater than the reference value, the candidate detection unit 130 records in the storage unit 170 the coke extrusion time that coincides with the detection time at which the index value becomes equal to or greater than the reference value. If so, the detection time may be recorded in the storage unit 170 in association with the coke extrusion time and the kiln number of the coke oven. Then, when determining unit 140 determines that visible smoke is generated in the candidate area, the occurrence of visible smoke is recorded in storage unit 170 in association with the detection time of the candidate area. You may instruct the notification of the coke extrusion time and the kiln number of the coke oven. As a result, the observer who receives the notification of visible smoke generation from the alarm device 200 can grasp that visible smoke is generated in the coke oven, and can identify the kiln that is highly likely to generate visible smoke. can grasp.

[2-3. alarm device]
The notification device 200 receives a notification from the visible smoke determination device 100 and notifies a monitoring person of the equipment, a factory, or the like that visible smoke is generated from the equipment to be monitored. The alarm device 200 is, for example, an information processing terminal such as a computer, a tablet terminal, or a mobile communication terminal, and includes a display unit for displaying information, such as a display. The alarm device 200 may include an audio output unit that outputs audio, such as a speaker. Upon receiving the notification from the visible smoke determination device 100, the alarm device 200 may notify the observer of the generation of visible smoke by displaying a message on the display unit, or notify the concerned parties of the factory or the like by e-mail. You may At this time, an image of smoke determined to be visible smoke may be displayed on the display unit or attached to the e-mail.

In addition, when the operation information of the equipment to be monitored is recorded in the storage unit 170, the alarm device 200 notifies information about the equipment to be monitored, which is highly related to the generation of visible smoke, together with the generation of visible smoke. You may For example, if the coke oven is being taken out at the same time as the visible smoke is generated, the kiln number may be notified.

[2-4. Information processing device]
The information processing device 300 is a terminal capable of transmitting and receiving information to and from the visible smoke determination device 100. For example, an observer or the like inputs to the visible smoke determination device 100 a reference value for an index value, a threshold value for excess time, and the like. used in some cases. The information processing device 300 may also be used as the alarm device 200, for example.

[2-5. Monitoring target management device]
The monitored object management device 400 is a device that manages the operation of the equipment to be monitored, and may be configured to be able to transmit and receive information to and from the visible smoke determination device 100 . The monitored object management device 400 may output, for example, operational information related to the generation of visible smoke in the facility to the visible smoke determination device 100 . For example, when the equipment to be monitored is a coke oven, the monitoring object management device 400 outputs the coke extrusion time and the kiln number of the coke oven to the visible smoke determination device 100 every time coke is pushed out in the coke oven. You may

The configuration of the visible smoke determination system 1 according to the present embodiment has been described above.

[3. Visible smoke determination method]
Next, the visible smoke determination method according to this embodiment will be described with reference to FIG. FIG. 6 is a flowchart showing a visible smoke determination method according to this embodiment.

As shown in FIG. 6, first, the imaging device 50 captures an image of a facility to be monitored for generation of visible smoke as a subject to obtain a captured image (S101). The imaging device 50 continuously captures images, and outputs the captured image to the visible smoke determination device 100 each time it acquires a captured image.

When a captured image is input from the imaging device 50, the visible smoke determination device 100 processes the captured image of the imaging device 50 and determines whether visible smoke is generated from the equipment to be monitored. First, the region setting unit 110 sets a visible smoke determination region within the captured image (S103). For example, the area setting unit 110 sets visible smoke determination areas A to M above the facility to be monitored in the captured image 10, as shown in FIG. The shape, size, position, and number of visible smoke determination regions set for the captured image are appropriately set, for example, based on the positions at which visible smoke appeared in previously acquired captured images. At this time, the area setting section 110 may further set a background area within the captured image. For example, the area setting unit 110 may set a background area X above the visible smoke determination areas A to M in the captured image 10, as shown in FIG.

After setting the visible smoke determination region in the captured image, the region setting unit 110 outputs information about the region set in the captured image (position information of each region in the captured image) to the index value calculation unit 120 .

Next, the index value calculation unit 120 calculates an index value, which is an index of visible smoke generation, based on the pixel information of each pixel within the visible smoke determination region (S105). For example, assume that the captured image is an RGB color system image with 256 gradations, and the visible smoke determination region and the background region are set. When information representing luminance is used as image information, the index value calculation unit 120 uses a function represented by the above-described red, green, and blue components of the pixels for each region set in the captured image, Obtain pixel information based on luminance. Next, the index value calculator 120 generates representative pixel information representing the area based on the pixel information of each pixel in the area set in the captured image. For example, the index value calculation unit 120 obtains, as representative pixel information, a value obtained by dividing the sum of luminance values of pixels in a region by the number of pixels in the region (that is, an average luminance value).

Then, the index value calculation unit 120 calculates an index value, which is an index of visible smoke generation, using pixel information in the area set in the captured image. For example, the index value calculation unit 120 calculates the value obtained by subtracting the moving average value of the difference value from the difference value obtained by subtracting the representative pixel information of the visible smoke determination region from the representative pixel information of the background region, as the index value of the visible smoke determination region. may be The index value calculation section 120 outputs the index value calculated for each visible smoke determination region to the candidate detection section 130 and the time measurement section 150 .

When the index value of the visible smoke determination region is input from the index value calculation unit 120, the candidate detection unit 130 or the time measurement unit 150 to which the index value is input receives the index value of the visible smoke determination region calculated in step S105. is compared with the reference value (S107). The processing of step S107 is performed by different functional units depending on whether or not the visible smoke determination region of the input index value has already been detected as a candidate region for the generation of visible smoke. If the visible smoke determination area is not a candidate area, the candidate detection unit 130 performs the determination, and if the visible smoke determination area is the candidate area, the time measurement unit 150 performs the determination. In order to make the description of the flowchart easier to understand, the following processing will be described with reference to the flow from when a certain visible smoke determination region is detected as a candidate region and the measurement of the excess time starts to the end of the measurement of the excess time.

First, after completing the processing up to step S105 for the first captured image, the candidate detection unit 130 compares the index value of the visible smoke determination region calculated in step S105 with the reference value (S107). When the candidate detection unit 130 determines that the index value of the visible smoke determination region is less than the reference value (S107: YES), the determination unit 140 determines that the time measuring unit 150 has already measured the excess time for the visible smoke determination region. It is determined whether or not there is (S109). If the determination unit 140 determines that the measurement of the excess time has not been performed for the visible smoke determination region (S109: NO), the determination unit 140 determines that the visible smoke determination region is not a visible smoke generation candidate region. Returning to the process, the process for the next captured image is started.

Returning to step S107, when the candidate detection unit 130 determines that the index value of the visible smoke determination region is equal to or greater than the reference value (S107: NO), the time measurement unit 150 has already measured the excess time for the visible smoke determination region. It is determined whether or not it is performed (S111). When the candidate detection unit 130 determines that the overtime measurement has not been performed for the visible smoke determination region (S111: NO), the candidate detection unit 130 determines that the visible smoke determination region is a visible smoke generation candidate region. Then, candidate detection section 130 notifies determination section 140 that a candidate region has been detected. The determination unit 140 instructs the time measurement unit 150 to start measuring the time exceeding the reference value for the detected candidate area (S113). After that, the process returns to step S101, and the process for the next captured image is started.

Returning to step S107 again, the time measurement unit 150 compares the index value of the candidate area and the reference value. When the time measurement unit 150 determines that the index value of the candidate area is continuously equal to or greater than the reference value (S107: NO), the candidate area is in a state of monitoring when the index value falls below the reference value. , the process returns to step S101 (S111: YES), and the process for the next captured image is started. Further, in step S107, the time measurement unit 150 compares the index value of the candidate area with the reference value, and determines that the index value of the candidate area is less than the reference value (S107: YES). If it is determined that the overtime is being measured (S109: YES), the index value that had exceeded the reference value in the candidate region has fallen below the reference value for the first time. The measurement of the excess time ends (S115). The time measurement unit 150 then outputs the measured excess time to the determination unit 140 .

When the measured overtime is input from the time measuring unit 150, the determining unit 140 compares the overtime with a threshold value (S117). When the excess time is equal to or greater than the threshold (S117: NO), the determination unit 140 determines that no visible smoke is generated in the candidate area (S119), terminates the processing of FIG. The process of FIG. 6 is restarted for the captured image. On the other hand, when the excess time is less than the threshold (S117: YES), the determination unit 140 determines that visible smoke is generated in the candidate area (S121). Then, the determination unit 140 instructs the notification device 200 to report the generation of visible smoke, and the notification device 200 notifies the observer of the generation of visible smoke (S123). At this time, if the operation information of the equipment to be monitored is recorded in the storage unit 170, information regarding the equipment to be monitored that is highly relevant to the generation of visible smoke may be notified along with the generation of visible smoke. good.

The visible smoke determination method according to the present embodiment has been described above. According to the visible smoke determination method according to the present embodiment, when the index value, which is an index of visible smoke generation, exceeds the reference value and the time period during which the index value exceeds the reference value is less than the predetermined threshold, It is determined that the visible smoke to be detected is generated in the visible smoke determination area. As a result, smoke mixed with clouds can be detected even if the reference value is set low, and erroneous detection due to clouds can be reduced, so erroneous detection can be reduced without increasing undetected visible smoke.

[4. Hardware configuration]
The hardware configuration of the visible smoke determination device 100 according to this embodiment will be described based on FIG. FIG. 7 is a block diagram showing an example of the hardware configuration of an information processing device 900 that functions as the visible smoke determination device 100 according to this embodiment.

Information processing apparatus 900 includes a processor (CPU 901 in FIG. 7), ROM 903 and RAM 905 . The information processing device 900 also includes a bus 907 , an input I/F 909 , an output I/F 911 , a storage device 913 , a drive 915 , a connection port 917 and a communication device 919 .

The CPU 901 functions as an arithmetic processing device and a control device. The CPU 901 controls all or part of the operations in the information processing apparatus 900 according to various programs recorded in the ROM 903 , RAM 905 , storage device 913 , or removable recording medium 925 . A ROM 903 stores programs used by the CPU 901, calculation parameters, and the like. The RAM 905 temporarily stores programs used by the CPU 901, parameters that change as appropriate during execution of the programs, and the like. These are interconnected by a bus 907 comprising an internal bus such as a CPU bus.

A bus 907 is connected to an external bus such as a PCI (Peripheral Component Interconnect/Interface) bus via a bridge.

The input I/F 909 is an interface that receives input from an input device 921, such as a mouse, keyboard, touch panel, button, switch, and lever, which is operation means operated by the user. The input I/F 909 is configured as, for example, an input control circuit that generates an input signal based on information input by the user using the input device 921 and outputs the signal to the CPU 901 . The input device 921 may be, for example, a remote control device using infrared rays or other radio waves, or an external device 927 such as a PDA corresponding to the operation of the information processing device 900 . A user of the information processing apparatus 900 can operate the input device 921 to input various data to the information processing apparatus 900 and instruct processing operations.

The output I/F 911 is an interface that outputs input information to an output device 923 that can notify the user visually or audibly. The output device 923 may be, for example, a display device such as a CRT display device, a liquid crystal display device, a plasma display device, an EL display device and a lamp. Alternatively, the output device 923 may be an audio output device such as speakers and headphones, a printer, a mobile communication terminal, a facsimile machine, or the like. The output I/F 911 instructs the output device 923 to output, for example, processing results obtained by various types of processing executed by the information processing device 900 . Specifically, the output I/F 911 instructs the display device to display the result of processing by the information processing device 900 as text or an image. In addition, the output I/F 911 instructs the audio output device to convert an audio signal such as audio data for which a reproduction instruction has been received into an analog signal and output the analog signal.

The storage device 913 is one of the storage units of the information processing device 900 and is a device for storing data. The storage device 913 is composed of, for example, a magnetic storage device such as a HDD (Hard Disk Drive), a semiconductor storage device, an optical storage device, a magneto-optical storage device, or the like. The storage device 913 stores programs executed by the CPU 901, various data generated by executing the programs, various data obtained from the outside, and the like.

The drive 915 is a recording medium reader/writer and is built in or externally attached to the information processing apparatus 900 . The drive 915 reads information recorded on the attached removable recording medium 925 and outputs it to the RAM 905 . The drive 915 can also write information to the attached removable recording medium 925 . The removable recording medium 925 is, for example, a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory. Specifically, the removable recording medium 925 includes CD media, DVD media, Blu-ray (registered trademark) media, CompactFlash (registered trademark) (CompactFlash: CF), flash memory, and SD memory cards (Secure Digital memory cards). etc. Also, the removable recording medium 925 may be, for example, an IC card (Integrated Circuit card) equipped with a contactless IC chip, an electronic device, or the like.

A connection port 917 is a port for directly connecting a device to the information processing apparatus 900 . The connection port 917 is, for example, a USB (Universal Serial Bus) port, IEEE1394 port, SCSI (Small Computer System Interface) port, RS-232C port, or the like. The information processing apparatus 900 can directly acquire various data from the external device 927 connected to the connection port 917 and provide various data to the external device 927 .

The communication device 919 is, for example, a communication interface configured with a communication device or the like for connecting to the communication network 929 . The communication device 919 is, for example, a communication card for wired or wireless LAN (Local Area Network), Bluetooth (registered trademark), or WUSB (Wireless USB). Further, the communication device 919 may be a router for optical communication, a router for ADSL (Asymmetric Digital Subscriber Line), a modem for various types of communication, or the like. The communication device 919 can transmit and receive signals and the like to and from the Internet and other communication devices, for example, according to a predetermined protocol such as TCP/IP. A communication network 929 connected to the communication device 919 is configured by a wired or wireless network or the like. For example, the communication network 929 is the Internet, home LAN, infrared communication, radio wave communication, satellite communication, or the like.

An example of the hardware configuration of the information processing apparatus 900 has been described above. Each component described above may be configured using general-purpose members, or may be configured by hardware specialized for the function of each component. The hardware configuration of the information processing device 900 can be changed as appropriate according to the technical level at which the present embodiment is implemented.

In order to verify the effects of the present invention, a coke oven was used as a facility to be monitored, and an image captured by an imaging device was image-processed to detect generation of black smoke in the coke oven. The implementation period was 5 days, and the monitoring time was daytime (6:00 to 18:00). The captured image is an RGB color system image, and visible smoke determination regions A to M and a background region X are set in the captured image, as shown in FIG. The image information is information representing luminance, and pixel information based on luminance is obtained using the function represented by the above-described pixel red, green, and blue components. A value (that is, an average luminance value) obtained by dividing by the number of pixels within is used as representative pixel information.

As the index value, the index value B used in FIG. 3 (that is, the value obtained by subtracting the moving average of the luminance difference from the luminance difference) was used. The moving average was a simple moving average for the past 10 minutes. The reference value for the index value was set to 6, and the overtime threshold was set to 45 seconds. The reference value was set so that 80% or more of the visible smoke that can be visually confirmed by the observer is detected as the candidate area. In other words, the reference value was set so that the reproducibility determined by the ratio of the number of visible smoke determined by the visible smoke determination device to the number of visible smoke that the observer could visually confirm was 0.8 or more. . In addition, the threshold for the excess time is set so that all visible smoke detected as candidate areas is determined as visible smoke. In other words, the threshold for the excess time was set so that the conformity rate, which is determined by the ratio of the number of visible smoke that can be visually confirmed by the observer to the number of candidate areas detected by the visible smoke determination device, is 1. Therefore, as shown in Table 1 below, the non-detection rates of Comparative Examples and Examples are the same.

Under these conditions, as an example, the visible smoke determination method based on the flowchart of FIG. 6 was performed. As a comparative example, the conventional method of judging the occurrence of visible smoke only by judging the index value without judging the excess time was carried out. Table 1 shows the results. 8 shows time-series data of the index values of the visible smoke determination regions A, C, and M from 12:00 to 15:00 on the third day of Table 1. As shown in FIG.

According to Table 1, in the rainy and cloudy weather on the first day, erroneous detection due to clouds occurred 49 times in the comparative example, but decreased to 18 times in the example. On the second to fifth days, the weather was sunny or cloudy and then sunny. In the comparative example, erroneous detection due to clouds occurred 3 to 5 times, but in the example, there were 0 times. A five-day average reduced false positives due to clouds by 72%.

Moreover, from FIG. 8, two cases (P1 and P2) visually confirmed as black smoke were detected in both the example and the comparative example. In these cases, the over-threshold time for which the index value exceeded the baseline was less than 45 seconds. On the other hand, in the two cases (P3, P4) in which the reference value was exceeded due to cloud influence, the threshold exceeding time during which the index value exceeded the reference value was 45 seconds or longer. Therefore, black smoke was erroneously detected in the comparative example, but was not detected as black smoke in the example.

Although the preferred embodiments of the present invention have been described in detail above with reference to the accompanying drawings, the present invention is not limited to such examples. It is obvious that a person having ordinary knowledge in the technical field to which the present invention belongs can conceive of various modifications or modifications within the scope of the technical idea described in the claims. It is understood that these also naturally belong to the technical scope of the present invention.

1 visible smoke determination system 10 captured image 50 imaging device 100 visible smoke determination device 110 region setting unit 120 index value calculation unit 130 candidate detection unit 140 determination unit 150 time measurement unit 160 threshold value setting unit 170 storage unit 200 alarm device 300 information processing Device 400 Monitoring Target Management Device 900 Information Processing Device 907 Bus 913 Storage Device 915 Drive 917 Connection Port 919 Communication Device 921 Input Device 923 Output Device 925 Removable Recording Medium 927 External Device 929 Communication Network

Claims (12)

an area setting unit for setting a visible smoke determination area in a captured image obtained by imaging the direction of the equipment to be monitored for visible smoke generation;
an index value calculation unit that calculates an index value as an index of visible smoke generation based on pixel information of each pixel in the visible smoke determination region;
a candidate detection unit that detects a visible smoke determination region in which the index value is equal to or greater than a predetermined reference value as a candidate region that is a candidate for generation of visible smoke;
a time measuring unit that measures an excess time from when the index value becomes equal to or greater than the reference value to when it becomes less than the reference value for the candidate region;
determining that visible smoke is generated in the candidate region for which the excess time is less than a predetermined threshold, and determining that no visible smoke is generated for the candidate region for which the excess time is equal to or greater than the predetermined threshold; a determination unit for
A visible smoke determination device comprising: The area setting unit further sets a background area within the captured image,
The index value calculation unit
For the visible smoke determination area and the background area set in the captured image, a value obtained by dividing the total sum of pixel information of each pixel in the area by the number of pixels in the area is used as representative pixel information representing the area. calculate,
As the index value,
a difference value obtained by subtracting the representative pixel information of the visible smoke determination region from the representative pixel information of the background region;
A time change value of the difference value, or
2. The visible smoke determination device according to claim 1, wherein either a value obtained by subtracting a moving average value of said difference value from said difference value is calculated. The region setting unit, based on a plurality of captured images captured within a predetermined period of time in the past,
setting an area in which visible smoke generated from the equipment to be monitored is imaged as the visible smoke determination area;
3. The visible smoke determination device according to claim 2, wherein an area that is not entirely covered with the visible smoke is set as the background area. The captured image is a 256-gradation RGB color system image,
The visible light according to any one of claims 1 to 3, wherein the index value calculation unit calculates pixel information of each pixel using a function represented by a red component, a green component, and a blue component of the pixel. Smoke detector. The visible smoke determination device according to any one of claims 1 to 4, wherein said pixel information is information representing luminance or lightness. a threshold setting unit that sets at least one of a reference value of the index value and a threshold of the excess time;
The threshold setting unit uses a plurality of captured images captured within a predetermined period of time in the past,
The recall determined by the ratio of the number of visible smoke determined by the visible smoke determination device to the number of visible smoke that can be visually confirmed by a person is set to a predetermined ratio or more,
A matching rate determined by the ratio of the number of visible smoke that can be visually confirmed by a person to the number of candidate areas detected by the visible smoke determination device is set to a predetermined ratio or more, or
Any one of claims 1 to 5, wherein at least one of the reference value of the index value and the threshold value of the excess time is set by maximizing an F value that is a harmonic average of the recall and the precision. The visible smoke determination device according to item 1. The visible smoke determination according to any one of claims 1 to 6, wherein the determination unit instructs a notification device to notify the generation of visible smoke when it determines that visible smoke is generated. Device. The visible smoke determination device according to any one of claims 1 to 7, wherein the equipment to be monitored is a coke oven. A storage unit that associates and stores the coke extrusion time and the kiln number of the coke oven each time coke is pushed out in the coke oven,
When the candidate detection unit detects a visible smoke determination region in which the index value is equal to or greater than the reference value, the coke extrusion time that coincides with the detection time at which the index value becomes equal to or greater than the reference value is stored in the storage unit. , the detection time is recorded in the storage unit in association with the coke extrusion time and the kiln number of the coke oven,
When the determination unit determines that visible smoke is generated in the candidate area, the determination unit instructs the alarm device to record the generation of visible smoke in association with the detection time of the candidate area in the storage unit. 9. The visible smoke determination device according to claim 8, which instructs notification of the coke extrusion time and the kiln number of the coke oven. A program that causes a computer to function as the visible smoke determination device according to any one of claims 1 to 9. an imaging device that captures the direction of a facility that is a monitoring target for visible smoke generation and acquires a captured image;
a visible smoke determination device that determines, based on the captured image, that visible smoke is generated from the equipment to be monitored;
including
The visible smoke determination device is
an area setting unit that sets a visible smoke determination area in the captured image;
an index value calculation unit that calculates an index value as an index of visible smoke generation based on pixel information of each pixel in the visible smoke determination region;
a candidate detection unit that detects a visible smoke determination region in which the index value is equal to or greater than a predetermined reference value as a candidate region that is a candidate for generation of visible smoke;
a time measuring unit that measures an excess time from when the index value becomes equal to or greater than the reference value to when it becomes less than the reference value for the candidate region;
determining that visible smoke is generated in the candidate region for which the excess time is less than a predetermined threshold, and determining that no visible smoke is generated for the candidate region for which the excess time is equal to or greater than the predetermined threshold; a determination unit for
A visible smoke determination system, comprising: a region setting step of setting a visible smoke determination region in a captured image obtained by capturing the direction of a facility to be monitored for generation of visible smoke;
an index value calculation step of calculating an index value as an index of visible smoke generation based on pixel information of each pixel in the visible smoke determination region;
a candidate detection step of detecting a visible smoke determination region in which the index value is equal to or greater than a predetermined reference value as a candidate region that is a candidate for generation of visible smoke;
a time measurement step of measuring the excess time from when the index value becomes equal to or greater than the reference value to when it becomes less than the reference value for the candidate region;
determining that visible smoke is generated in the candidate region for which the excess time is less than a predetermined threshold, and determining that no visible smoke is generated for the candidate region for which the excess time is equal to or greater than the predetermined threshold; a determination step to
A visible smoke determination method, comprising:

Images