KR20070028813A - Method and system for monitoring forest fire - Google Patents

Abstract

A forest fire monitoring method and system are provided to exactly detect a forest fire and to quickly cope with a forest fire by informing a manager's terminal of the forest fire. A forest fire monitoring and informing method is composed of steps of: storing sensing information transmitted from a sensor unit and image data received from an image obtaining unit through the wire or wireless network(610); judging whether the present time corresponds to a pre-set daytime range or not; determining that a forest fire breaks out if a grey region of the image data analyzed through RGB(Red,Green,Blue) analysis algorisms is over a predetermined critical regional area, when the present time corresponds to a pre-set daytime range and determining that a forest fire occurs if an ignition region of the image data analyzed through HSV(Hue Saturation Value) analysis algorisms is over a predetermined critical regional area, when the present time does not correspond to a pre-set daytime range; and generating emergency information by using the sensing information and transmitting the emergency information to a predetermined receiving terminal(625,630).

Description

Forest fire detection method and system {Method and system for monitoring forest fire}

1 is a schematic overall configuration diagram of a forest fire detection system according to an embodiment of the present invention.

Figure 2 is a block diagram showing the configuration of the basic information acquisition unit according to an embodiment of the present invention.

3 is a block diagram of an image acquisition unit according to an exemplary embodiment of the present invention.

Figure 4 is a block diagram of an image analysis and processing unit according to an embodiment of the present invention.

5 is a flow chart showing a forest fire detection method according to an embodiment of the present invention.

6 is a flowchart illustrating a method of using sensing information according to whether or not a fire occurs according to an exemplary embodiment of the present invention.

The present invention relates to a forest fire detection method and apparatus, and more particularly, to a forest fire detection method and apparatus using an image captured by a camera installed in a forest and sensing information by a sensor.

After Korea's liberation, more than 70% of the country's land consists of forests due to active forest greening policy. As the economy has improved since the 1970s, many people have moved to the mountains every holiday to enjoy idle time.

Accordingly, a lot of forest fires occur due to the carelessness of the hikers during the holiday, and forest fires are generated not only by the carelessness of the hikers but also by natural disasters such as rice paddy burning and lightning in rural areas. In addition, the causes of forest fires are diversifying.

Forest fires caused by various causes are gradually being enlarged through the forests, and it is not easy to extinguish when the initial suppression fails, which causes many lives and property damages. For example, 22 places of Naksansa Temple and cultural assets including forests were burned down by the forest fires in Goseong and Yangyang provinces, which occurred in April 2005. The damage is estimated at 100 billion won.

Therefore, the early detection and early suppression of forest fires should be considered as the top priority in terms of forest preservation, human life protection and property protection, and the demand for bold investments is increasing.

The conventional forest fire monitoring method is mainly based on visual monitoring and control of a manager. In other words, the forest monitoring post is installed at the elements of the forest where the human access is frequent or at the point where the forest fire is easy to monitor, and when the manager detects that the forest fire has occurred, the communication method is used to notify the central control station.

This forest fire monitoring method may be the most effective method within the manager's working hours and monitoring areas, but the area or time that the manager can monitor is limited, which makes it difficult to achieve the desired purpose. In addition, such a forest fire monitoring method is configured to have a plurality of managers resident or shift work for each monitoring area and forest monitoring post, causing a rise in labor costs and climbing fees.

In addition, the conventional forest fire monitoring system using a camera is a system that installs the camera on the main element of the mountain to detect the fire by detecting the visual image received through the camera by the operating personnel of the central control station.

However, in the conventional forest fire monitoring system, a small number of operating personnel cannot simultaneously monitor images received from a plurality of cameras installed on the main elements of the mountain. There was a problem that the detection was incorrect. In addition, the conventional forest fire monitoring system has a problem that it is impossible to quickly detect and deal with forest fire in the absence of the operating personnel.

Accordingly, an object of the present invention for solving the above problems is to provide a forest fire detection method and system that enables rapid follow-up of the forest fire can be quickly detected.

Another object of the present invention is to provide a forest fire detection method and system capable of early evolution by enabling rapid and accurate detection of a forest fire occurrence region and prediction of a forest fire path using a camera image and sensing information.

Still another object of the present invention is to provide a forest fire detection method and system capable of quickly and optimal forest fire detection with minimal system construction cost.

Still another object of the present invention is to provide a forest fire detection method and system that can promptly respond to a manager terminal when detecting a forest fire so as to enable rapid response, and can also induce a safe evacuation of mountaineers by notification to a disaster prevention system.

Other objects of the present invention will become more apparent through the preferred embodiments described below.

According to an aspect of the present invention for achieving the above object, there is provided a forest fire detection method and / or a recording medium that enables the method.

Forest fire detection and notification method according to an embodiment of the present invention comprises the steps of (a) storing the image data received from the image acquisition unit and the sensing information received from the sensor unit via a wired or wireless network, respectively; (b) determining whether the current time falls within a preset weekly range; (c) If the gray range of the image data analyzed using the RGB analysis algorithm is more than a predetermined critical area area when the day range is within the weekly range, it is determined that a wildfire has occurred. Determining that a fire has occurred if the ignition region of the analyzed image data is greater than or equal to a predetermined critical region area; And (d) generating emergency information by using the sensing information and transmitting the emergency information to a predetermined receiver.

The sensing information may include one or more of temperature, humidity, wind direction, and wind speed, and the like. The emergency information may include one or more of a fire ignition point, a forest fire spread direction, a forest fire spread rate, current damage area information, and hourly damage prediction area information. It may include. The emergency information may further include the image data determined to have been determined to be a fire.

The forest fire ignition point may be a position coordinate calculated by projecting the image data determined to have been determined to have a forest fire on a GIS drawing.

Performing m-1 calculation of the amount of change between two consecutive image data on the stored m (natural numbers of two or more) image data before the step (b) in the forest fire detection and notification method; And determining whether each change amount calculated in m-1 times is maintained at a predetermined size or more, and if step (b) is performed when each change amount is maintained at a predetermined size or more. It may be.

The ignition region may be region information within a range of HSV values preset to have a yellow color value.

The forest fire detection and notification method may include generating control information for controlling a shooting cycle or an operation mode of the image acquisition unit when it is determined that a forest fire has occurred; And transmitting the generated control information to the image acquisition unit through a wired or wireless network.

Hereinafter, a forest fire detection method and apparatus according to an embodiment of the present invention will be described in detail with reference to the accompanying drawings. In the description with reference to the accompanying drawings, the same or corresponding components are designated by the same reference numerals regardless of the reference numerals and duplicate description thereof will be omitted.

1 is a schematic overall configuration diagram of a forest fire detection system according to an embodiment of the present invention, Figure 2 is a block diagram showing the configuration of the basic information acquisition unit according to an embodiment of the present invention, Figure 3 4 is a block diagram of an image acquisition unit according to an embodiment of the present invention, and FIG. 4 is a block diagram of an image analysis and processing unit according to an embodiment of the present invention.

Referring to FIG. 1, the forest fire detection system according to the present invention includes n (natural numbers) basic information acquisition units 110-1, 110-2,..., 110-n-hereinafter referred to as '110'. do.

The basic information acquisition unit 110 transmits the captured image and sensing information to the forest fire detection and processing unit 120 through a wired or wireless network in order to enable forest fire detection at a remote location.

The basic information acquisition unit 110 may be separately provided at a position capable of acquiring a wide range of images or main elements of a mountain for forest fire monitoring.

As illustrated in FIG. 2, the basic information obtaining unit 110 may include an image obtaining unit 210 providing image information, a sensing information obtaining unit providing sensing information (ie, the gateway 220 and n sensors). Sections 230-1, 230-2, ..., 230-n-hereinafter referred to as '230'. Although FIG. 1 illustrates that only one image acquisition unit 210 is included in one basic information acquisition unit 110, the number of image acquisition units 210 may be increased or decreased as necessary.

The image acquisition unit 210 is individually provided at a position (for example, a mountain peak) where major elements of the mountain or a wide range of images can be acquired for forest fire monitoring, and the captured image is a forest fire through a wired or wireless network. It is delivered to the sensing and processing unit 120.

As illustrated in FIG. 3, the image acquisition unit 210 may include a camera unit 310, a data converter 320, a transceiver 330, a power supply unit 340, and a controller 350.

The camera 310 captures an image by capturing an external image every predetermined period (for example, one frame every 10 seconds before a fire occurs, and one frame every second after detecting a fire) under the control of the controller 350. Generate data.

The camera 310 may control a shooting mode such as pan, tilt, and zoom in / out under the control of the controller 350.

The camera unit 310 may be formed on an upper end of a support (not shown) having a predetermined height. This is to enable the imaging of a wide area, and also to be able to quickly and accurately measure the occurrence point of the fire by using the images provided from the plurality of image acquisition unit 210.

The data converter 320 converts the image captured by the camera 310 into a form that can be transmitted through a wired or wireless network.

The transmitter / receiver 330 transmits the image data captured by the camera unit 310 and converted by the data converter 320 to the forest fire detection and processing unit 120 through a wired or wireless network. The controller 120 receives a control signal (for example, a photographing period change command, a photographing mode control command, etc.) from the control signal to the control unit 350.

When the transmission / reception unit 330 transmits the converted image data to the forest fire detection and processing unit 120, unique information (eg, a serial number, installation, etc.) assigned to the image acquisition unit 210 under the control of the control unit 350. At least one of a local name, an installation address, etc.) may be transmitted together. The forest fire detection and processing unit 120 may use the received unique information to recognize which image data is received from which image acquisition unit 210 at which position, and capture a mode of the camera unit 310 (eg, For example, by analyzing the photographing angle, etc., it is possible to recognize which region the image is.

The power supply unit 340 provides operation power of each component included in the image acquisition unit 210. The power supply unit 340 may be in the form of a battery or the like, or in the form of a self-generating device (eg, a solar heat collecting plate, wind power generation, etc.).

The control unit 350 generates image data corresponding to the control information (for example, a shooting cycle, a shooting mode, etc.) received from the wildfire detection and processing unit 120 or through predetermined wires, and detects and processes the wildfire detection and processing unit through a wired or wireless network. The camera unit 310, the data converter 320, the transmitter / receiver 330, and the power supply 340 are controlled to be transmitted to the 120.

In addition, when the power supply unit 340 is in the form of a battery, the controller 350 may transmit a battery replacement request to the wildfire detection and processing unit 120 through the transceiver 330 when the remaining power is below a predetermined level.

Although not shown in FIG. 3, the image acquirer 210 may further include a storage. The storage unit may store an operation program of the image acquisition unit 210 and unique information given to the image acquisition unit 210.

Referring back to FIG. 2, the sensing information acquisition unit includes n sensor units 230 and a gateway 220.

Each sensor unit 230 is separately provided as necessary in the required portion of the mountain top and / or the slope, the Ad-hoc network is formed between each sensor unit 230.

Ad-hoc network method is a method of connecting and connecting all devices within a radius that can be recognized by each terminal device independently in space. In an ad-hoc network, each device can be a server, a hub, or even a client.

Sensing information by each sensor unit 230 is transmitted to the gateway 220 through the Ad-hoc network and to the forest fire detection and processing unit 120 through a wired or wireless network. As an example, the IEEE 802.15.4 Wireless Personal Area Network (WPAN) protocol and Zigbee, which are low speed and low power wireless interfaces, may be used for communication between the sensor units 230. Since a method of configuring an Ad-hoc network is formed between the sensor units 230, it will be apparent to those skilled in the art, and thus a detailed description thereof will be omitted.

The sensor unit 230 may be referred to as a small computer system having a built-in MCU, and is provided with an operating system (OS) for sensing processing and communication with each node (that is, the sensor unit 230 and / or the gateway 220). . Specifically, the sensor unit 230 includes a plurality of sensors for sensing external environmental information, a transceiver for wireless communication, a power supply unit for operating power supply (for example, a battery, a solar panel, wind power generation, etc.) and software. It may include a processing unit mounted. The sensor included in the sensor unit 230 may include a wind direction sensor, a wind speed sensor, a temperature sensor, a humidity sensor, and the like, and may include various sensors at the same time. The sensor unit 230 may transmit unique information (eg, at least one of a serial number, an installation position, etc.) of the sensor unit 230 in transmitting the sensing information to the gateway 220. The unique information may be given to correspond to a sensor that detects the corresponding sensing information. In addition, the sensor unit 230 may further include a GPS unit that generates and provides location coordinates in order to recognize the installation position of the sensor unit 230.

The sensor unit 230 may perform a sensing operation every predetermined period or whenever a sensing command is received from the wildfire detection and processing unit 120 through the gateway 220. While the sensor unit 230 does not perform the sensing operation, The power consumption may be minimized by switching to the sleep mode by the control.

The gateway 220 transmits sensing information received from each sensor unit 230 to the forest fire detection and processing unit 120 through a wired or wireless network. In addition, the gateway 220 transmits a control signal for each sensor unit 230 from the wildfire detection and processing unit 120 to the corresponding sensor unit 230.

Each sensor unit 230 and the gateway 220 may be implemented to be compatible with IPv6.

As shown in FIG. 2, the information transmission and reception path between the image acquisition unit 210 and the forest fire detection and processing unit 120 and the information transmission and reception path between the gateway 220 and the forest fire detection and processing unit 120 may be configured separately. Obviously, it may be integrated into one information transmission / reception path as needed.

Referring back to FIG. 1, the forest fire detection and processing unit 120 of the forest fire detection system according to the present invention uses the image data received from each image acquisition unit 210 separately provided in the main elements of the mountain to determine whether the forest fire occurs. The controller determines an alarm, performs an alarm when a fire occurs, and records an image received if necessary. The forest fire detection and processing unit 120 generates forecast information on the likelihood of occurrence of forest fire using the sensing information received from the sensing information acquisition unit, and generates emergency information on forest fire progress prediction after the forest fire occurs. When a fire occurs, the forest fire detection and processing unit 120 transmits a video image and emergency information of a forest fire situation to the manager terminal 130 and / or the disaster prevention system 140.

When the forest fire detection and processing unit 120 detects that a forest fire has occurred by image data received from an image acquisition unit 210, the lens angle of the surrounding image acquisition unit 210 is adjusted (that is, the photographing mode). By transmitting the control command, it is also possible to collect images of various angles for the corresponding area (that is, the area recognized as having a fire). Similarly, the forest fire detection and processing unit 120 may transmit a sensing control command to the sensing information acquisition unit according to whether a forest fire occurs.

In general, the forest fire detection and processing unit 120 may be installed in a central control station.

Referring to FIG. 4, the configuration of the forest fire detection and processing unit 120, the forest fire detection and processing unit 120 includes a transceiver 410, a data converter 420, an image analyzer 430, and a sensing information analyzer. 440, a display unit 450, an alarm generator 460, a storage unit 470, a message generator 480, and a controller 490.

The transceiver 410 transmits a corresponding control signal (eg, a photographing period changing command, a photographing mode control command, a sensing period changing command, etc.) to each of the image obtaining unit 210 and / or the sensing information obtaining unit, The image data and / or sensing information captured by each image acquisition unit 210 and / or the sensing information acquisition unit are received and transmitted to the data conversion unit 420. As described above, each image acquisition unit 210 and the sensing information acquisition unit may transmit pre-allocated unique information together in transmitting the captured image data and / or the sensing information.

The data converter 420 converts the image data and the sensing information received through the transceiver 410 into data of a form that can be recognized by the forest fire detection and processing unit 120.

The image analyzer 430 determines whether a forest fire occurs at an arbitrary position corresponding to the specific image acquirer 210 by using the image data converted by the data converter 420. When a forest fire occurs, the image analyzer 430 calculates a forest fire ignition point by using unique information of the image acquisition unit 210 and the photographing angle of the image acquisition unit 210 that transmitted the corresponding image. In order to accurately calculate a fire ignition point, an image image photographing the same point from a plurality of image acquisition units may be used. The image analyzer 430 detects whether a fire occurs by using an analysis algorithm designated to correspond to the reception time of the image data. The time range for selecting the analysis algorithm may be specified by dividing the day range, the night range, and the like. The daytime range and nighttime range may be determined according to the time adjustment of the operating personnel, for example, the daytime range may be specified from sunrise time to sunset time. An image analysis process by the image analyzer 430 will be described in detail later with reference to a flowchart.

The sensing information analyzer 440 may use environment information (for example, wind direction, wind speed, temperature, and the like) at an arbitrary position corresponding to the specific sensor unit 230 by using the sensing information converted by the data converter 420. Humidity, etc.). The sensing information analyzer 440 generates forecast information or emergency information by using environment information. Environmental information may be used to generate forecast information (eg, dry advisory, forest fire probability information, etc.) before a fire, and emergency information (eg, fire ignition point, wildfire spread direction, Wildfire spreading rate, etc.). The sensing information analyzer 440 may include the forest fire ignition point information and the sensing information (for example, wind speed and wind direction) recognized by the image analyzer 430 so that the forest fire direction and progress speed may be included in the emergency information. can do.

The display unit 450 outputs image data, an alarm signal, and the like received from the image acquisition unit 210 in the form of visual information. The display unit 450 may further output forecast information or emergency information generated by the sensing information analyzer 440 in the form of visual information. The display unit 450 may be, for example, a CRT monitor, a liquid crystal display, or the like. The display unit 450 may be divided into a plurality of divided regions and displayed, and may include image data received from the plurality of image acquisition units 210 or information generated by the sensing information analyzer 440 (that is, environment information). , At least one of forecast information, emergency information, etc.) may be displayed in each divided area. If it is determined that a forest fire has occurred or selection information for one of the divided images is input, an image displayed on the divided region is displayed on the entire screen under the control of the controller 490.

The alarm generator 460 outputs a predetermined alarm signal when it is determined that the wildfire has occurred by image analysis of the image analyzer 430. The alarm signal may be output in a sound form through the speaker or through the display unit 450 as visual information (eg, text, graphics, etc.).

The storage unit 470 may include an operation program, unique information corresponding to each image acquisition unit 210 and / or a sensor unit 230, image data received from each image acquisition unit 210, and each sensor unit 230. Received sensing information, image (or video) data received from the corresponding image acquisition unit 210 when detecting a forest fire and the like. Before the forest fire occurs, the previous image data except for the minimum image data (for example, m (natural number) image data) required for analysis by the image analyzer 430 may be deleted by the control of the controller 490. Can be.

The message generator 480 generates a message of a predetermined type according to whether or not a fire occurs under the control of the controller 490 and transmits the message to a predetermined receiver. For example, before a fire occurs, the message generator 480 generates the forecast information generated by the sensing information analyzer 440 in the form of a message (for example, an SMS message) in a predetermined format and generates a manager terminal ( 130 and / or to the disaster prevention system 140. In addition, after the fire occurs, the message generator 480 generates a message including the image data received from the corresponding image acquisition unit 210 and the emergency information generated by the sensing information analyzer 440 and the manager terminal ( 130 and / or to the disaster prevention system 140. Obviously, the message generated by the message generator 480 after the forest fire is not included in the image data itself, and may include a URL for accessing the image data.

Of course, the communication between the message generator 480 and the manager terminal 130 may not be performed in a one-way message transmission mode, but may be performed in a bidirectional data communication mode between the message generator 480 and a specific application in the manager terminal 130. have. For example, when only the emergency code is transmitted from the message generator 480 to the manager terminal 130, an application preloaded in the manager terminal 130 is operated to be connected to the message generator 480, and the manager is driven. Emergency information can also be checked through the execution screen of the application.

The controller 490 may determine whether a forest fire occurs using image data received from each image acquisition unit 210, perform an alarm when a forest fire occurs, and record a received image if necessary. Transceiver 410, data converter 420, image analyzer 430, sensing information analyzer 440 to generate emergency information and / or forecast information using the sensing information received from the sensing information acquisition unit. ), The display unit 450, the alarm generator 460, the storage unit 470, and the message generator 480 are controlled. In addition, the controller 490 may control the image data received from the corresponding image acquisition unit 210 to be displayed on the entire screen of the display unit 450 when the forest analysis is detected by the image analyzer 430. In addition, the controller 490 generates a control signal for each image acquisition unit 210 to be transmitted through the transceiver 410.

Although not shown in FIG. 3, the forest fire detection and processing unit 120 may further include an input unit (not shown) for receiving a control command (for example, changing a shooting cycle, changing a screen display, etc.) by an operating agent. have.

Referring back to FIG. 1, the manager terminal 130 is a user device that receives at least one of forecast information, emergency information, and a forest fire generation image from the forest fire detection and processing unit 120. For example, the manager terminal 130 may be implemented in the form of a mobile communication terminal, PDA, etc. having a communication function.

As described above, the forest fire occurrence image, forecast information, emergency information, etc. may be received in the form of a one-way message transmission from the forest fire detection and processing unit 120 as data of a predetermined format.

In addition, data transmission and reception may be performed between the forest fire detection and processing unit 120 and the manager terminal 130 in a bidirectional communication method. For example, when only the emergency code is transmitted from the message generator 480 to the manager terminal 130, an application preloaded in the manager terminal 130 is operated to be connected to the message generator 480, and the manager is driven. Emergency information can also be checked through the execution screen of the application. The execution screen can be equipped with a function button that can be selected by the administrator (for example, view the current damage area, view the estimated damage area, view the direction of fire progress and speed, etc.), and is operated according to the function button selected by the administrator. The application may calculate and display the corresponding information or display data provided from the forest fire detection and processing unit 120.

The disaster prevention system 140 may be a system of an institution for disaster prevention purposes, such as a 119 information system, a disaster countermeasure headquarters, and the like. When emergency information is received from the forest fire detection and processing unit 120, the disaster prevention system 140 may convert the information into a predetermined format and transmit an SMS text message to a communication terminal of a general citizen in real time. In this case, the location information of each communication terminal already recognized by the mobile communication service system may be used to transmit only to communication terminals located within a limited distance (for example, a radius of 1 km) from the fire ignition point.

5 is a flowchart showing a forest fire detection method according to an embodiment of the present invention.

5 relates to a method of detecting whether a fire occurs using image data received from one image acquisition unit 210, the forest fire detection and processing unit 120 is an image received from a plurality of image acquisition unit 210 Obviously, the data can be used to individually determine whether a forest fire occurred.

Referring to FIG. 5, in operation 510, the storage unit 470 stores image data (that is, the n (natural number) received image data) from the image acquisition unit 210 through a wireless or wired network.

In the storage unit 470, m (natural numbers) image data necessary for the image analyzer 430 to recognize whether a fire occurs should be stored. The image acquisition unit 210 detects a forest fire through wired or wireless network image data obtained by capturing an external image at a predetermined time period (for example, a time period designated as a default, a time period corresponding to a shooting cycle change command, etc.). And it can be transmitted to the processing unit 120 as described above.

In operation 515, the image analyzer 430 determines whether the current time (that is, the reception time of the n-th time image data) is a weekly range in order to select an algorithm for determining whether or not a fire occurs. For example, the daytime range may be specified after sunrise but before sunset.

If the current time falls within the weekly range, the image analyzer 430 proceeds to step 520 to perform analysis using a predetermined RGB algorithm.

In general, during the daytime, forest fires can be identified using smoke that rises. Since a large amount of smoke is generated in the early stages of wildfire generation, it is easier and more accurate to detect smoke by detecting smoke than light caused by wildfire.

The smoke detection may use a gray area determined by the R, G, and B values of the received and stored image. For example, in the case of dark gray, values of R (red), G (green), and B (blue) represent values of about 50 in the range of 0 to 255, respectively. Therefore, after setting a range of R, G, and B values corresponding to the gray area representing the smoke in advance, the range of pixels belonging to the range may be identified as the gray area (that is, the area representing the smoke).

It is possible to determine whether a forest fire occurs by the existence of an area determined as a gray area and / or whether the gray area is distributed wider than a predetermined threshold area.

However, in order to prevent a cloud floating in the sky from being incorrectly identified as smoke, the installation angle or the photographing angle of the camera 310 may be limited to be photographed below the skyline.

However, if the current time is in the night range, the image analyzer 430 proceeds to step 525 to perform analysis using a predetermined HSV algorithm.

In general, the distribution of red color can be seen at night when viewed from a distance. Therefore, forest fires can be identified using the red color distribution at night without using smoke, and the color distribution is HSV (H: color value, S: saturation, V: brightness) rather than RGB color space. Using color spaces is more effective.

In case of wildfire, the occurrence of wildfire can be detected by using the characteristic that light of yellow system is mainly generated at the initial occurrence. In this case, an area from red to yellow (for example, 0 <H <60 or 300 <H <360) can be detected as a forest fire generation area. A method of converting pixel information of an image into an HSV color space (for example, a method of converting between an RGB color space and an HSV color space) is obvious to a person skilled in the art, and a lot of software therefor is also omitted.

A fire can be determined by the presence or absence of a ignition region (eg, from red to yellow when using the HSV color space) and / or whether the ignition region is distributed wider than a predetermined threshold region. have.

In operation 530, the image analyzer 430 determines whether a forest fire is determined by the determination of operation 520 or 525.

If it is determined that a wildfire has occurred, the flow proceeds to step 535 to perform an alarm (for example, outputting sound through a speaker and outputting visual information through a display unit), and receiving image data received from the corresponding image acquisition unit 210. Record video of forest fire area.

As described above, in step 535, the control unit 490 transmits a control signal (for example, a shooting cycle change command, a shooting mode change command, etc.) to the corresponding image acquisition unit 210 and / or the image acquisition unit 210 installed in the vicinity. ), So that the image corresponding to the forest fire area can be provided in real time or at a very short time period.

When it is determined that the forest fire occurs as a result of the determination in step 530, the image analyzer 430 may analyze the location of the ignition point by projecting the area indicated by the video image on the GIS map. This is because the fire detection and processing unit 120 may recognize the installation position and the photographing angle of the corresponding image acquisition unit 210. When the area where wildfire is generated on the video image is widely recognized, the wildfire may be recognized at the center coordinates. For more accurate analysis of the ignition point, it is apparent that the image analyzer 430 may simultaneously analyze a plurality of images photographing the area where the fire occurred.

However, if it is determined that no wildfire has occurred, the wildfire detection and processing unit 120 proceeds to step 510 to store image data newly received from the image acquisition unit 210 (that is, the n + 1th received image data). Proceed again.

The forest fire detection method using the image data according to the present invention has been described with reference to FIG. 5. In addition, the number of image data used for detecting a fire is described based on the case of one.

However, it is obvious that a plurality of image data used in detecting a fire may be applied. That is, after repeatedly performing steps 510 to 525 of the m images received in FIG. 5, if a wildfire is commonly recognized, it may be determined that a wildfire has occurred.

In addition, the process of determining the rate of change of color over time may be preceded before step 515. In general, the camera unit 310 for monitoring the forest fire target area is taken from a long distance over a large area, so in the case of two images taken at short time intervals after fixing the lens direction in one direction color histogram, brightness Because of the very similarities, shapes, and so on, the rate of change over time is very small. However, there is a certain rate of change in the case of movement of the sun and clouds, rainfall and snowfall, changes in the shape of trees caused by the wind, and other objects such as mountain climbers. Other individuals such as mountain climbers affect only a short time when judging the amount of change comparing a plurality of images, and other cases may have a long change but have a small change rate. That is, in the case of a forest fire or the area left by the forest fire, the color and size change not only lasts a long time but also has a large change rate. May be limited to proceed only to step 415.

In addition, although only a case of using a camera for capturing an image so far has been described, an infrared camera may be added or replaced with an infrared camera. When using an infrared camera, k (natural numbers) images are acquired at intervals of 1 or 2 seconds. After the obtained image is processed in black and white, two or more consecutive images are subjected to XOR operation (i.e., operation to determine whether they match in units of pixels and output corresponding results). ), The rate of change for the k images can be detected. In the case of smoke and natural phenomena occurring in private houses, there is no big difference in the rate of change by XOR operation.However, in case of smoke of forest fire, the amount of change increases with the passage of time.

6 is a flowchart illustrating a method of using sensing information according to whether or not a fire occurs according to an exemplary embodiment of the present invention.

Referring to FIG. 6, in operation 610, the storage unit 470 stores sensing information (ie, n-th sensing information received from the plurality of sensing information acquisition units) through a wireless or wired network.

In the storage unit 470, the sensing information quantity necessary for the sensing information analyzer 440 to generate forecast information or emergency information may be stored. The forest fire detection and processing unit detects the sensing information detected by each sensor unit 230 at a predetermined time period (for example, a time period designated as a default, a time period corresponding to a sensing period change command, etc.) through a wired or wireless network. 120 may be transmitted as described above. The sensing information analyzer 440 may also calculate an amount of change (for example, a change in the wind direction) of the environmental information when the plurality of pieces of sensing information sensed by the same sensor are stored in the storage 470.

In operation 615, the sensing information analyzer 440 determines whether a forest fire is detected by the analysis of the image analyzer 430 (see description of FIG. 5).

If it is determined that the forest fire has not occurred, the flow proceeds to step 620 to generate forecast information (eg, dry advisory, forest fire occurrence probability information, etc.) using the received and stored sensing information. For example, the sensing information analysis unit 440 may use temperature and humidity information in generating forest fire probability information. In other words, if the humidity is high, the probability of occurrence of forest fire is relatively low, and if the humidity is low, the probability of forest fire is relatively high. Similarly, strong wind speeds can produce relatively high levels of fire caution.

However, if it is determined that a fire occurs, the flow proceeds to step 625 to generate emergency information (eg, a fire ignition point, a fire spread direction, a fire spread rate, etc.) using the stored sensing information. The emergency information may further include current damage area information, estimated hourly damage area information considering wind direction and wind speed. Current damage area information and estimated damage area information can be displayed using a GIS drawing, which makes it easier to select an evolution point.

In operation 630, the message generator 480 generates a message using the image image generated by the image acquirer 210 and the sensing information generated by the sensing information analyzer 440 according to a predetermined condition. For example, to the administrator terminal 130 and / or disaster prevention system 140, etc.). As described above, the message generator 480 may use a unidirectional transmission method or a bidirectional transmission / reception method in transmitting a message generated by the receiver.

Step 630 may be repeatedly performed at predetermined intervals, and the manager or the disaster prevention system 140 may determine the degree of forest fire extinguishing, etc. in real time using a repeatedly received message, and may appropriately cope with each time point. Become.

As described above, the method and system for detecting wildfire according to the present invention enables rapid detection of wildfire occurrence, thereby enabling rapid follow-up.

In addition, the present invention has the effect that it is possible to quickly detect the forest fire occurrence area and predict the progress path of the forest fire by using the camera image and sensing information early evolution.

In addition, the present invention has the effect that it is possible to detect the rapid and optimal forest fire with a minimum system construction cost.

In addition, the present invention has the effect that can be promptly responded to the administrator terminal when a forest fire is detected, and can be promptly responded to, and also to notify the disaster prevention system to induce safe evacuation of mountaineers.

Although the above has been described with reference to a preferred embodiment of the present invention, those skilled in the art to which the present invention pertains without departing from the spirit and scope of the invention as set forth in the claims below It will be appreciated that modifications and variations can be made.

Claims (7)

In the forest fire detection and notification method, (a) storing image data received from the image acquisition unit and sensing information received from the sensor unit, respectively, via a wired or wireless network; (b) determining whether the current time falls within a preset weekly range; (c) If the gray range of the image data analyzed using the RGB analysis algorithm is more than a predetermined critical area area when the day range is within the weekly range, it is determined that a wildfire has occurred. Determining that a fire occurs when the fired area of the analyzed image data is equal to or larger than a predetermined critical area area; And and (d) generating emergency information using the sensing information and transmitting the emergency information to a predetermined receiver. The method of claim 1, The sensing information includes one or more of temperature, humidity, wind direction, and wind speed, and the emergency information includes one or more of a forest fire ignition point, a forest fire spread direction, a forest fire spread rate, current damage area information, and damage prediction area information per hour. Forest fire detection and notification method characterized in that. The method of claim 2, And the emergency information further comprises the image data determined to have been determined to be a fire. The method of claim 2, The forest fire ignition point is a forest fire detection and notification method, characterized in that the position coordinates calculated by projecting the image data determined to determine the forest fire on a GIS drawing. The method of claim 1, Before step (b) above, Performing m-1 calculations of the amount of change between two consecutive image data on the stored m (natural numbers of two or more) image data; And Determining whether or not each change amount calculated in m-1 times is maintained above a predetermined size is preceded. The fire detection and notification method, characterized in that the step (b) is performed when the amount of each change is maintained above a predetermined size. The method of claim 1, The fire detection region is a forest fire detection and notification method, characterized in that the area information within the preset HSV value range to have a yellow color value. The method of claim 1, Generating control information for controlling a shooting cycle or an operation mode of the image acquisition unit when it is determined that a forest fire has occurred; And And transmitting the generated control information to the image acquisition unit through a wired or wireless network.

Images