JPH07245757A - Fault detector having visible ray camera and infrared ray camera - Google Patents

Abstract

PURPOSE:To surely and quickly detect a fault by combining the visible image system and the thermal image system and using an image processing algorithm in common to compensate draw-backs of both systems. CONSTITUTION:A visual monitor image from a visible ray camera 1 is processed by an image processing unit 3, and a fault occurrence location in a supervised image is stored in a memory 21 as fault information together with a fault content such as a fire or smoke breakout and similarly a thermal supervisory image from an infrared ray camera 2 is processed by the image processing unit 3 and a fault occurrence location in a supervised image is stored in the memory 21 as fault information together with a fault content such as a fire or smoke breakout and a CPU 22 is used to discriminate the fault by using a majority decision theory based on the fault alarm obtained from the two different supervised images and when a fault of the same content is sensed in the same location, the fault is detected with high reliability and the visual ray image system and the thermal image system are combined and an image processing algorithm in common to compensate defects of both systems to detect the fault surely and quickly.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an abnormality detecting device for detecting an abnormality in an area to be detected, such as an oil leak, smoke emission, fire in a fuel storage facility of a thermal power plant and an intruder.

[0002]

2. Description of the Related Art Oil leaks, smoke, fires and intruders may occur in fuel storage facilities of thermal power plants.
These abnormalities are currently detected by thermal fire detectors, ion fire detectors, smoke fire detectors, flame fire detectors, visual image abnormality detectors, and thermal image abnormality detectors, either alone or in combination. I am doing it. A thermal fire detector detects a fire using a mechanical contact (bimetal etc.) operated by heat. The ion-type fire detector detects a fire by using an electrical contact operated by ions contained in smoke particles. The smoke-type fire detector detects smoke and indirectly detects a fire depending on the magnitude of smoke concentration in the air. The flame type fire detector detects a fire by detecting ultraviolet rays generated from the flame. The visible image type abnormality detector detects a change in a visible image captured by a visible camera by an image processing device to detect fire, smoke, or an intruder. The thermal image type abnormality detector detects a change in a thermal image captured by an infrared camera by an image processing device and detects a fire, smoke, or an intruder.

[0003]

However, there are the following problems in the conventional oil leak, smoke generation, fire and abnormality detection of an intruder. (1) It is difficult to say that it is always possible to detect anomalies at the present time because there is a time delay in detecting anomalies in thermal fire detectors, ion fire detectors, and smoke fire detectors. In addition, it cannot be used in large areas or outdoors. Furthermore, since the purpose is to detect a single anomaly, a plurality of detectors must be used together to detect a plurality of anomalies. (2) Although flame-type detectors respond quickly to abnormalities, they aim to detect a single abnormality, so multiple detectors must be used together to detect multiple abnormalities. (3) The visible image anomaly detector can overcome the problems of (1) and (2), but it is difficult to detect anomalies at night due to oil leaks, smoke, and intruders' dark surroundings. In the early days of the daytime (fire), it is difficult to detect abnormalities due to the bright surroundings. (4) Thermal image type abnormality detector, (1) in the same manner as (3) above,
Although the problem of (2) can be overcome, it is not possible to detect smoke when the heat generating part is invisible.

The present invention has been made in view of the above situation. In addition to combining a visible image system and a thermal image system, an image processing algorithm for compensating for each other's defects is also used, and abnormality detection can be performed accurately and quickly. An object is to provide an abnormality detection device.

[0005]

To achieve the above object, the structure of the present invention comprises: a visible camera for obtaining a visual image of a detected area; an infrared camera for obtaining an infrared image of the detected area;
A storage device that digitizes and stores the image signals of the visible camera and the infrared camera, an image processing device that performs image processing calculation of information in the storage device, a central processing device that controls image processing calculation by the image processing device, and an image processing signal And a control terminal for operating the central processing unit to grasp the abnormal state of the detected area.

[0006]

[Function] The visible monitor image from the visible camera is processed by the image processing device, and the location of the abnormality in the monitor image is stored in the storage device as an abnormality alarm together with the details of the abnormality such as fire and smoke
Similarly, the thermal monitoring image from the infrared camera is processed by the image processing device, and the location of the abnormality in the monitoring image is stored as an abnormality alarm in the storage device together with the details of the abnormality such as fire and smoke.
Next, based on the abnormality warnings obtained from these two different monitoring images, the majority judgment theory is used to make an abnormality judgment, and when abnormalities of the same content are detected at the same location, a highly reliable abnormality warning is given. If an abnormality is detected only in the monitoring image of, the abnormality alarm is added to that effect, or the monitoring image is processed again and then the abnormality alarm is issued.

[0007]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a block diagram showing the configuration of an anomaly detection apparatus for both visible and infrared cameras according to an embodiment of the present invention.
2 and 3 show flowcharts of the image processing apparatus.

In FIG. 1, 1 is a visible camera, 2 is an infrared camera, 3 is an image processing device, 4 is a television display as a display device, 5 is an operation terminal device, and 6 is a fire source-oriented water discharge having a rotating device. Gun, 7 sprinkler device, 8 acoustic device (alarm), 9 oil leak detection device, 1
Reference numeral 0 is an intrusion detection device.

The television display 4 has a visible camera 1
Alternatively, the image signal of the infrared camera 2 is input via the image processing device 3. Inside the image processing device 3, a memory 21 as a storage device for digitizing and storing the image signal of the visible camera 1 or the infrared camera 2 and a central processing unit (CPU) 22 for controlling image processing calculation are provided. ing. The CPU 22 performs various kinds of arithmetic processing according to a preset image processing algorithm (flow chart).

The function of the CPU 22 will be described with reference to FIGS. 2 and 3. When the image processing is started and the thermal image original image needs to be changed (described later), the thermal image is taken from the infrared camera 2 and stored in the memory A21a. Further, when it is necessary to change the visible image original image (described later), the visible image from the visible camera 1 is captured and stored in the memory B21b. Next, the thermal image from the infrared camera 2 is temporarily stored in the memory X21x as an inspection image. After the storage, the contents of the memory A21a stored previously are compared with the contents of the memory X21x.
As a result of the comparison, if an abnormality (fire / smoke (with heat generating part) / leakage (with temperature) / intruder: detection method described later in thermal image) is detected, the details of the abnormality and position information in the thermal image Is stored in the memory C21c. Similarly, the visible image from the visible camera 1 is used as the inspection image in the memory X21x.
Temporarily store in. After storage, the memory B21b previously stored
And the contents of the memory X21x are compared. As a result of comparison, anomaly {Fire, smoke, oil leak, intruder by visible image (each has a difference in brightness from the surroundings): Detection method will be described later}
When is detected, the abnormality content and position information in the visible image are stored in the memory D21d. Finally, as a mutual correction between the thermal image and the visible image, the memory C21c and the memory D
The contents of 21d are compared and it is confirmed whether or not there is the same abnormality at the same position. As a result, if there is an abnormality, the place of the abnormality is displayed on the television display 4 as an alarm process. At this time, the display screen may be a composite with a visible image or a thermal image. Further, the alarm by the alarm 8 may be used together with the display on the television display 4. If the infrared camera 2 and the visible camera 1 have different visual fields, the positions may be corrected so that the visual fields are the same. When the abnormality content at the same position is different, the display form (content) different from the above is used as the alarm processing, the abnormality detection processing is performed from the start again, and then the visible image / thermal image mutual correction and the alarm processing are performed. FIG. 4 shows a schematic block diagram in detection of fire and smoke.

A case where the thermal image original image or the visible image original image needs to be changed will be described. The flowcharts of FIGS. 2 and 3 are periodically executed by a program incorporated in the image processing apparatus 3, and the inspection image changes with the passage of time, but the thermal image original image or the visible image original image to be compared with the inspection image is changed. Does not change because is not changed. To detect abnormalities, the difference between the inspection image and the thermal image original image or the visible image original image is captured as a change.Therefore, when the thermal image original image or visible image original image is fixed, it is detected when the change occurs. Can not. Therefore, the thermal image original image or the visible image original image is changed at appropriate time intervals corresponding to the detected abnormality. In addition, depending on the monitoring target or the environment in which this device is placed, the monitoring image may change at intervals shorter than the above-mentioned time interval, so the appropriate time interval is variable. That is, when the change is fast, the time interval is set short, and when the change is slow, the time interval is set long.

Next, a method of detecting a fire, smoke, oil leak, or intruder by using a thermal image or a visible image will be described. There are various detection methods, but the main one is extraction of change characteristics.

The detection of a fire (the same applies to the detection of smoke) from a thermal image from the infrared camera 2 will be described with reference to FIG.
FIG. 5 shows an explanatory diagram of fire detection. Generally, an object having an absolute temperature or higher emits infrared rays according to the temperature. As the temperature of an object rises, infrared energy increases sharply. Fire detection is performed under the following conditions. There is a high-temperature fixed fire point (preset threshold value), there is a high-temperature part with an area change above the fire point and its center of gravity fluctuates, and the area of the high-temperature part above the fire point is the time Growing up with. That is, when there is an infrared level higher than the alarm set level, there is fluctuation due to temporal change, and there is no shift of the center of gravity and there is a time-series change in the generated area, it is detected as a fire. If the infrared level is lower than the alarm setting level, there is no fluctuation due to temporal changes, and there is no change in the time-series occurrence area due to movement of the center of gravity, it is processed as a disturbance.

The detection of oil leakage by the thermal image from the infrared camera 2 will be described. The oil leak is detected under the following conditions. In the monitoring place, there is a fixed oil leak source having a temperature slightly higher than the surroundings (about several degrees), and the area growing around the oil leak source with time.

Detection of an intruder by a thermal image from the infrared camera 2 will be described. The detection of an intruder is made under the following conditions. In the monitoring place, there is an object of about body temperature and there is no change in area, and that object moves with time.

The detection of fire by the visible image from the visible camera 1 will be described. Fire detection is based on the following conditions. There is a fixed high-intensity fire point (preset threshold value), there is a change in the area of the mid-intensity area corresponding to the external flame area above the fire point, and its center of gravity fluctuates, and above the fire point That the area of the middle brightness part grows with time,
Is.

Detection of smoke emission from a visible image from the visible camera 1 will be described with reference to FIG. FIG. 6 shows an illustration of smoke detection. The smoke detection is performed under the following conditions. There is a fixed low-intensity smoke source, there is a low-intensity area change above the smoke source, and its center of gravity fluctuates, and a low-intensity area above the smoke source grows over time. . That is, if there is a change in shape and area in the difference image and the occurrence locations are the same, it is detected as smoke. Also, if there is no change in shape and area in the difference image and there is no identity in the place of occurrence, it is processed as a disturbance.

The detection of oil leakage by the visible image from the visible camera 1 will be described. The oil leak is detected under the following conditions. There is a fixed source of oil leakage with a lower intensity than the surveillance location,
And that the area grows over time, centered on the oil spill source.

Detection of an intruder by a visible image from the visible camera 1 will be described. Intruder detection is based on the following conditions. That is, the area does not change at a constant brightness, and the center of gravity moves with time.

When a fire, smoke, oil leakage, or an intruder abnormality is detected by the above-described method, the operation terminal 5, the fire-source-directed water gun 6 and the sprinkler device 7 are combined to perform automatic or manual operation. Take immediate action such as extinguishing fire.

Therefore, the abnormality can be detected regardless of day and night, and unmanned operation can be performed substantially continuously, and labor for monitoring can be reduced. In addition, it is possible to detect abnormalities early, and it is possible to minimize damage. Further, it is possible to improve the responsiveness of the judgment and the treatment instruction by making the abnormal state grasped properly. Also, by combining with a fire extinguisher, it becomes possible to enhance the responsiveness of fire fighting activities.

[0022]

EFFECTS OF THE INVENTION The abnormality detection apparatus for combined use of a visible camera and an infrared camera according to the present invention processes a visible surveillance image from the visible camera by an image processing device, and the location of the abnormality in the surveillance image together with the content of the abnormality such as fire or smoke. Is stored in the storage device as an abnormality alarm, and similarly, the thermal monitoring image from the infrared camera is processed by the image processing device, and the location of the abnormality in the monitoring image is stored as an abnormality alarm in the storage device along with the details of the abnormality such as fire and smoke. However, based on the abnormality alarms obtained from these two different monitoring images, the majority judgment theory is used to make an abnormality determination, so if an abnormality of the same content is detected at the same location,
It is possible to perform highly reliable abnormality detection. As a result, it becomes possible to combine the visible image type and the thermal image type together with the image processing algorithm for compensating for each other's defects, and accurately and quickly detect the abnormality.

[Brief description of drawings]

FIG. 1 is a block diagram of a configuration of a visible camera / infrared camera combined-type abnormality detection device according to an embodiment of the present invention.

FIG. 2 is a flowchart of the image processing apparatus.

FIG. 3 is a flowchart of the image processing apparatus.

FIG. 4 is a schematic block diagram for detecting fire and smoke.

FIG. 5 is an explanatory diagram of fire detection.

FIG. 6 is an explanatory diagram of smoke detection.

[Explanation of symbols]

 1 Visible Camera 2 Infrared Camera 3 Image Processing Device 4 Television Display 5 Operation Terminal Device 6 Fire Source Oriented Water Spray Gun 7 Sprinkler Device 8 Acoustic Device (Alarm) 9 Oil Leakage Detection Device Device 10 Intrusion Detection Device 21 Memory 22 Central Processing Unit ( CPU)

Claims (1)

[Claims] 1. A visible camera for obtaining a visual image of a detected area, an infrared camera for obtaining an infrared image of a detected area, a storage device for digitizing and storing image signals of the visible camera and the infrared camera, and a storage device. An image processing apparatus for performing image processing of information, a central processing unit for controlling the image processing calculation by the image processing apparatus, a display unit for displaying an image processing signal, and a central processing unit for operating an abnormal state of a detected area. An anomaly detection device that combines a visible camera and an infrared camera, and is composed of an operating terminal for grasping.