KR20100058180A - System and method for detecting disaster occurrence - Google Patents

Abstract

PURPOSE: A system and a method for detecting disaster occurrence are provided to detect a disaster by applying an image processing corresponding to the classification standard of category. CONSTITUTION: An image input part(110) receives the image of a structure. An image processing unit(120) reconstructs the image of the structure based on the classification standard of the structure state category. The image processing compares a gray value of the structure which is normal state with the gray value of the current structure. An image managing unit(130) transmits an abnormal structure image and imaging processing result data to the outside.

Description

System and method for detecting disaster occurrence

The present invention relates to a disaster detection system and method, and more particularly, to detect the presence or absence of a disaster by applying image processing corresponding to a classification criterion of a category to a captured image of a managed structure to detect a disaster occurrence. The present invention relates to a disaster detection system and method.

Recently, interest in the prevention and management of accidents caused by various disasters is increasing rapidly. The annual accident rate of disasters is increasing by 5%, while the preventive and safeguards against such disasters use the old methods.

In other words, due to the development of electronic and measurement technology, many fields are using automated measurement management system using advanced devices and methods.However, natural disasters such as typhoon and flood, as well as disasters related to various structures such as fire and snowfall in tunnels and bridges The field still maintains a system that is directly supervised by managers.

As such, direct supervision by managers is not always complete due to the fact that it is not always on its own and not supported by a large number of people. Therefore, a system that can monitor 24 hours of disasters without resorting to manpower is required.

Therefore, many systems have been studied to develop such a system, and as a result, a method for detecting a disaster of a structure through quantitative analysis of data measured using a sensor has been developed. However, the disaster detection system using such a sensor has a problem that it takes a lot of time to analyze the measurement data and is not efficient when dealing with emergency disasters.

Therefore, it is possible to check the disaster situation with the naked eye before analyzing the data for quick initial response in case of disaster, and it is possible to detect the abnormality automatically and send it to the relevant organizations in real time according to the type to enable effective response to the disaster. Development of the system is urgent.

The present invention has been made to solve such a conventional problem, by transmitting the captured image of the structure in real time to check the presence of disaster in real time, and apply the image processing corresponding to the classification criteria of the category to the captured image The purpose of this is to automatically detect the presence of a disaster and enable effective coping with disasters.

In order to achieve the above object, a disaster detection system according to the present invention includes an image input unit, an image processing unit, and an image manager.

The image input unit receives a structure image photographing the structure to be managed, and the image processing unit reconstructs the structure image based on the classification criteria of a preset structure state category, and then, the image is input to the standard image gray value of the normal state that was previously learned. The image processing unit compares the gray values of the images, and the image manager divides the abnormal structure image that does not match the standard image gray value and the image processing result data of the abnormal structure image by the structure state category among the image processing results. Save and send to the outside.

In this case, the structure state category may include one or more of categories representing a plurality of structure states such as a water level state category, a lost state category, a fire state category, and a traffic state category.

In addition, the image processing unit may perform image processing based on the position of the ruler displayed on the structure in the structure image so as to correspond to the classification criteria of the water level state category.

In addition, the image processing unit may perform image processing based on the outer shape of the structure in the structure image to correspond to the classification criteria of the lost state category.

In addition, the image processing performing unit may perform image processing around a predetermined area of the upper portion of the structure in the structure image so as to correspond to the classification criteria of the fire state category.

In addition, the image processing performing unit may perform image processing around an arbitrary fixed area of the vehicle passing through the upper road of the structure in the structure image so as to correspond to the classification criteria of the traffic condition category.

In addition, the disaster detection system according to the present invention may further include an image collection unit for collecting the structure image photographed to include the entire image of the structure to be managed to transmit to the image input unit.

In addition, the disaster occurrence detection system according to the present invention stores the abnormal structure image received from the image management unit to correspond to the internal database configured according to the structure status category, and further includes a central management unit for determining whether the disaster from the abnormal structure image can do.

The central management unit may generate an alarm when the abnormal structure image is determined to be a disaster and transmit the abnormal structure image and the disaster determination result data to the manager.

A disaster detection method according to the present invention includes an image input step, an image processing step, and an image management step.

The image input step receives a structure image photographing a structure to be managed, and the image processing step reconstructs a structure image based on a classification criteria of a preset structure state category, and then learns a normal image gray value of a normal state previously learned. Image processing is performed to compare the gray values of the structure image to the image management step, and the image processing result data of the abnormal structure image and the image processing result data for the abnormal structure image that does not match the standard image gray value among the image processing results Save by category and send to the outside.

In addition, the structure state category may include one or more of categories representing a plurality of structure states such as a water level state category, a lost state category, a fire state category, and a traffic state category.

In the image processing step, the image processing may be performed based on the position of the ruler displayed on the structure in the structure image so as to correspond to the classification criteria of the water level state category.

In the image processing step, the image processing may be performed based on the outer shape of the structure in the structure image so as to correspond to the classification criteria of the lost state category.

In addition, in the image processing step, the image processing may be performed around a predetermined area of the upper part of the structure in the structure image so as to correspond to the classification criteria of the fire state category.

In addition, in the performing of the image processing, the image processing may be performed around a fixed area of the vehicle that passes through the upper road of the structure in the structure image so as to correspond to the classification criteria of the traffic condition category.

In addition, the disaster occurrence detection method according to the present invention may further include an image collection step of collecting and transmitting the structure image photographed to include the entire image of the structure to be managed before the image input step.

In addition, the disaster occurrence detection method according to the present invention stores the abnormal structure image to correspond to the internal database configured according to the structure status category after the image management step, and further comprises a central management step of determining whether the disaster from the abnormal structure image It may include.

In addition, in the central management step, when an abnormal structure image is determined to be a disaster, an alarm is generated, and the abnormal structure image and the disaster determination result data may be transmitted to the manager.

According to the present invention, image processing is performed on an image photographing a managed object structure to detect a disaster occurrence according to a category classification criterion, and the presence or absence of a disaster occurrence of the managed object structure is detected by comparing the photographed data to the steady state data. This can enable effective coping with disasters.

In addition, by transmitting the captured image of the management target structure to the administrator in real time to check the presence of a disaster in real time.

Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings. In order to more clearly understand the present invention, the same reference numerals are used for the same components in different drawings.

1 is a block diagram schematically showing an embodiment of a disaster detection system configuration according to the present invention.

The disaster detection system for performing image processing based on classification criteria of a category to detect the occurrence of a disaster of a managed object structure includes an image collector 200, an image processing module 100, and a central management unit 300. The image processing module 100 may include an image input unit 110, an image processing unit 120, and an image manager 130.

The image collection unit 200 may photograph the structure image to include the entire image of the structure to be managed, and includes an image of a disaster occurrence site such as a flood, heavy snowfall, collapse, and fire installed at a point that is difficult to access. Collected structure image is transmitted to the image input unit 110 in real time.

The structure to be managed may be a structure that may be damaged due to a disaster such as a building, a bridge, a road, a tunnel, or a dam. Hereinafter, a bridge installed in a demonstration area is an example of a representative structure to be managed. To illustrate the invention.

The image collecting unit 200 may correspond to a device suitable for capturing an image. In the present invention, the image collecting unit 200 includes a camera, a lens, a pan tilt, and the like. In this case, the user may set the camera to include the entire image of the structure to be managed.

In order to automatically detect a structure to be managed and take a structure image without setting a camera from a user, the image collection unit 200 may use a 'VISION sensor' that may be connected to a camera.

'VISION sensor' is an optical sensor that recognizes an object in the image input from the camera by using image sensing technique, and is mainly used for robotics.

Here, the image sensing technique is an advanced image recognition technique capable of determining the presence or absence of an object in an image having a small number of pixels by recognizing the outline of an object, etc., and is capable of recognizing an object with minimum illumination even in a dark place. It is a technique that can recognize a structure regardless of the change in the photographing angle of the target structure.

The 'VISION Sensor' is not affected by 360 degree rotation and scale (zoom in or out), and can recognize the object even when the camera angle is 45 degrees, and even if part of the object is hidden or changed. It can be recognized, and a plurality of objects can be recognized simultaneously.

In addition, it has a fast recognition speed and is independent of the format of the recognized image and image file, and has little characteristic of light.

Due to these features, when the 'VISION sensor' is used in the image collecting unit 200, the user may set the camera to more accurately detect the structure to be managed and photograph the structure image than to photograph the structure to be managed.

The image processing module 100 is an image processing unit based on a category classification criterion based on the image collected by the image collection unit 200, and may be located integrally with or adjacent to the image collection unit 200. A typical configuration may include a camera control controller, an industrial PC, and a control box, and the system configuration may include the image input unit 110, the image processing unit 120, and the image manager 130 as described above. It may include.

The image input unit 110 receives a structure image photographing a structure to be managed from the image collection unit 200, and the image processing unit 120 reconstructs the structure image based on a classification criteria of a preset structure state category. In addition, image processing is performed to compare the gray values of the structure image to the standard image gray values of the steady state trained in advance.

In addition, the structure state category may include one or more of categories representing the state of a plurality of structures such as a water level state category, a lost state category, a fire state category, and a traffic state category.

At this time, the classification criteria of the state category is a portion designated as an object for comparing gray values in the structure image. That is, if a ruler is designated as a comparison target in the bridge structure image, the corresponding structure image is classified into a water level state category. If the exterior shape of the structure is designated as a comparison object, the corresponding structure image is classified into a lost state category.

In addition, if a predetermined area of the upper part of the structure is designated as a comparison object in the structure image, it is classified into a fire state category, and if a certain fixed area of a road on the structure is designated as a comparison object, it is classified into a traffic state category.

The most important factor in using the image sensing technique in the above-described image collector 100 is to select a factor that does not affect its own image sensing even when the external environment is changed by analyzing the shape of the object to be detected. In the present invention, a method of analyzing a change in gray value by performing image processing on a structure image undergoing image sensing as an important factor is applied to the sensing level for measuring a change in gray value. , Lost conditions, fire conditions, and traffic conditions.

In addition, the image processing using the image is to determine the degree of large-scale shape change of the object by using the information of the pixels constituting the image, in the present invention, the image acquisition unit 200 applies a high resolution black and white camera As such, the image processing module 100 may measure the change in the gray value, thereby measuring the change in the structure to be managed.

Hereinafter, an image classified according to structure state categories will be described in detail with reference to FIGS. 2 to 5, and as described above, FIGS. 2 to 5 show bridges as embodiments of a structure to be managed.

In addition, the reconstruction of the structure image referred to below means that the structure image is appropriately reconstructed into a category to be classified through editing such as enlargement, reduction, rotation, and selective cropping.

2 is a diagram illustrating an embodiment of performing image processing based on classification criteria of a water level state category.

 FIG. 2 shows that the image is reconstructed around a ruler in meters (meters) marked with fluorescent material on the bridge (to detect changes in the water level at night), and the water level at the bridge is changed. In this case, since the degree of the ruler appearing on the surface changes, image processing can analyze the amount of gray value change in the ruler image to determine whether a disaster has occurred.

3 is a diagram illustrating an embodiment of performing image processing based on classification criteria of a lost state category.

3 shows that the image is reconstructed based on the shape of the bridge set as the exterior image, such as a dotted line. When the bridge is lost, the exterior image of the bridge is changed, so that the gray value of the exterior image is processed through image processing. The amount of change can be analyzed to determine if a disaster has occurred.

4 is a diagram illustrating an embodiment of an image that performs image processing based on classification criteria of a fire state category.

4 shows that the image is reconstructed around a predetermined area of the upper part of the bridge indicated by a dotted line, and in case of a fire, an image change of a brighter and wider shape than the headlight of the vehicle appears in the dotted area. The amount of gray value change in the internal image may be analyzed to determine whether a disaster has occurred.

FIG. 5 is a diagram illustrating an embodiment of performing image processing based on classification criteria of a traffic condition category.

FIG. 5 shows that the image is reconstructed around an arbitrary fixed area of the vehicle that passes through the upper road of the bridge, such as the area indicated by the dotted line, and the gray value change of the image inside the dotted area is analyzed through image processing. The number of passes per unit time of the vehicle or the traffic can be determined.

That is, when there is no change in gray value for a predetermined time, it can be determined as a stagnant state, and the number of times that the gray value has changed (vehicle passes) for a predetermined time or more can be determined as the number of passing vehicles. .

According to the structure status category, the order of disaster occurrence can be changed according to the structure to be managed. For example, when the structure to be managed is a bridge, it is lost, level, fire, and traffic in order. It may be determined sequentially, if the structure to be managed is a dam may be determined in the order of lost state, the water level, and if the structure to be managed is a building, it may be determined in the order of lost state, fire.

Referring again to FIG. 1, the image manager 130 may perform image processing result data on the abnormal structure image and the abnormal structure image that do not match the standard image gray value among the image processing performance results analyzed as described above. The structure is classified by category and stored and transmitted to the central management unit (300).

In this case, as a transmission method, a high speed / capacity communication device such as HSPDA, Wibro, or an existing Internet network may be used.

The central management unit 300 stores the abnormal structure image and the image processing result data of the abnormal structure image received from the image management unit 130 to correspond to an internal database configured according to the structure state category, and the abnormal structure image and the abnormal structure It is possible to determine whether a disaster has occurred from the image processing result data for the image.

That is, when the result of comparing the gray values does not correspond to the standard image gray value, the image processing result data may not be included in the safety range, or may be determined as the occurrence of a disaster when included in the danger range.

The central management unit 300 may generate an alarm when the abnormal structure image is determined to be a disaster and transmit the abnormal structure image and the disaster determination result data to the manager.

In other words, the change of the water level is compared and analyzed by comparing the transmitted ruler analysis result and data related to the danger level range (ruler range safe for flooding or ruler range during flooding) stored in the database of the central management unit 300. / Save the analysis results in the form of images (graphs or charts) and text, and when the ruler is included in the danger level range, it is recognized as a disaster situation (overflow) and an alarm is generated. Information and level change rate) and the abnormal structure image received from the image management unit 130 may be provided to the operator or manager.

The operator or manager can visually detect the real-time disaster situation through the abnormal structure image and can predict the occurrence of flood by using the database of comparison / analysis results for annual statistics management, so take effective measures to minimize the damage. Can be taken.

In addition, whether or not the bridge is lost is compared and analyzed by the external image analysis results of the transmitted bridges and the lost safety range (safety state in which there is a bridge loss but no risk of collapse) stored in the database. Graph or chart) and text format, and when the result of the external image analysis is not included in the lost safety range data, it is recognized as a disaster (loss of the bridge) and an alarm is generated, and the comparison / analysis result (inconsistent data) Numerical information of) and the abnormal structure image received from the image management unit 130 may be provided to the operator or administrator.

The operator or manager can visually detect the real-time disaster situation through the abnormal structure image, and through the comparison / analysis results, it is possible to accurately identify the loss or damage of some sections of the bridge, and to effectively reduce the damage. Can be taken.

In addition, the fire of the structure is analyzed by the analysis result of the upper area of the bridge and the fire-free range stored in the database (the state in which there is no fire in the bridge, when the vehicle headlight appears, in the rain or in the fog). Compare / analyze data and save comparison / analysis results in image form (graphs or charts) and text form.

If the upper area analysis result does not match the fire-free range data (large gray value brighter and wider than the headlights), it is recognized as a disaster (fire occurrence) and an alarm is generated. Numerical information of the data that does not match) and the abnormal structure image transmitted from the image management unit 130 may be provided to the operator or manager.

Operators or managers can visually detect real-time disasters through abnormal structure images, and compare / analyze results to accurately detect fire and smoke caused by vehicle accidents or other accidents on bridges. Effective measures can be taken to minimize them.

In addition, the traffic condition on the structure compares / analyzes the analysis result of the fixed area of the vehicle passing the bridge upper road and the vehicle location state data stored in the database, so that the analysis result of the fixed area for the unit time is The number of times that match the position status data is detected by the number of passes of the vehicle.

At this time, the fixed arbitrary area is an area in which one vehicle can be located, and the data stored in the database includes gray value data and vehicle position state data in case of fog or rain other than passing through the vehicle. The vehicle position state data refers to gray value data when various vehicles such as passenger cars (small, medium and large), trucks, buses, SUVs, etc. pass through a fixed area.

The vehicle position status data includes gray values for a fixed area, but if there is no change, it is determined that the vehicle is stationary (if a large number of vehicles pass through a certain area, a large gray value (vehicle) and a small gray value ( Space) appears repeatedly, but when the vehicle is congested and the vehicle is located in a certain area, a large gray value is continuously displayed), a comparison / analysis result (vehicle passing information or congestion state information) and the image management unit 130 The abnormal structure image received from the operator can be provided to the operator or manager.

The operator or manager can visually detect the real-time traffic situation through the abnormal structure image (the format is preferably a video), and can accurately analyze the traffic situation on the road on the bridge through the comparison / analysis results. It can provide drivers with useful traffic information.

6 is a flowchart schematically showing an embodiment of a disaster occurrence detection method according to the present invention, and will be described by taking a bridge as an example of a structure to be managed.

The structure image is photographed so that the structure itself is included from the image collection unit 200 installed at the position where the entire bridge to be managed may be photographed (S100). The photographed structure image is input to the image processing module 100 and reconstructed based on the classification criteria of the preset structure state category (S200).

The reconstruction of the image is reconstructed based on different classification criteria for each category, and the structure image classified into the water level state category is reconstructed based on the ruler displayed on the pier where the degree of change occurs as the level changes. The image to be analyzed is reconstructed around the entire outer image form of the bridge.

Also, the image to analyze the fire condition of the bridge is reconstructed centering on a predetermined area of the upper part of the bridge, and the image to analyze the traffic condition is centered on a fixed predetermined area of the vehicle that passes the road on the upper road of the bridge. Reconstruct

 Next, image processing is performed on the reconstructed structure image (S300), and as a result of performing image processing, it is checked whether the standard image gray value and the structure image gray value match (S400).

If the two gray values coincide with each other, it is determined that no disaster has occurred and the process is terminated as it is. If the structure image gray value is different from the standard image gray value and the image processing result does not match, the gray value different from the standard image gray value is The structure image having a structure (hereinafter referred to as an abnormal structure image) and the image processing result data for the abnormal structure image is transmitted to the central management unit 300 through a communication network.

The central management unit 300 compares / analyzes the transmitted abnormal data and the image processing result data of the abnormal structure image with safety range data or unsafe range data to determine whether a disaster has occurred (S500).

If the transmitted results are included in the safety range data or not included in the unsafe range data, it is determined that a disaster has not occurred and is terminated as it is, and the transmitted results are not included in the safe range data, or the unsafe range data. If it is included in the disaster is determined to have occurred (S600), generates a warning, and transmits the abnormal structure image and disaster occurrence determination result to the manager and operator (for example, the Ministry of Construction and Transportation) (S700).

Managers and operators can visually determine the disaster situation through the original structure, and the exact result of the disaster can be detected through the result of the disaster occurrence judgment. Therefore, this data can be managed systematically for disaster prediction and disaster case management. Can be.

In general, data transmission to managers and operators is performed using a communication network, but in case of emergency, text and video transmission can be performed through a mobile phone.

If the present invention is applied to a large number of CCTVs, etc. installed in various facilities in the country, to have an automatic detection or determination function, it will be possible to transmit the disaster image in real time when various disasters occur.

In addition, in the exemplary embodiment of the present invention, one original structure image is photographed and reconstructed, and image processing is performed for each category. However, the structure is photographed so as to conform to the category selected when the original structure image is captured by setting (for example, a ruler). Only part of the image or only a predetermined area on the bridge), image processing can be performed immediately without reconstruction.

So far I looked at the center of the preferred embodiment for the present invention. Those skilled in the art will appreciate that the present invention can be implemented in a modified form without departing from the essential features of the present invention. Therefore, the disclosed embodiments should be considered in descriptive sense only and not for purposes of limitation. The scope of the present invention is shown in the claims rather than the foregoing description, and all differences within the scope will be construed as being included in the present invention.

1 is a block diagram schematically showing an embodiment of a disaster detection system configuration according to the present invention.

2 is a diagram illustrating an embodiment of performing image processing based on classification criteria of a water level state category.

3 illustrates an embodiment of performing image processing based on classification criteria of a lost state category.

4 illustrates an embodiment of performing image processing around classification criteria of a fire condition category.

5 is a diagram illustrating an embodiment of performing image processing based on classification criteria of a traffic condition category.

6 is a flow chart schematically showing an embodiment of a disaster occurrence detection method according to the present invention.

Claims (18)

An image input unit configured to receive a structure image photographing a structure to be managed; An image processing unit configured to perform image processing by reconstructing the structure image based on a classification criterion of a preset structure state category, and comparing the gray value of the structure image to a previously learned normal state gray value; And An image management unit for storing the abnormal structure image which does not match the standard image gray value and the image processing result data of the abnormal structure image by dividing the image by the structure state category and transmitting it to the outside; Disaster occurrence detection system comprising a. The method of claim 1, The structure state category, Disaster occurrence detection system, characterized in that it comprises one or more of the categories representing the status of a plurality of structures, such as water level status category, lost status category, fire status category, traffic status category. 3. The method of claim 2, The image processing unit, And the image processing centering on the position of the ruler displayed on the structure in the structure image so as to correspond to the classification criteria of the water level state category. 3. The method of claim 2, The image processing unit, And the image processing is performed around the outer shape of the structure in the structure image so as to correspond to the classification criteria of the lost state category. 3. The method of claim 2, The image processing unit, And the image processing is performed around a predetermined area of an upper portion of the structure in the structure image so as to correspond to the classification criteria of the fire condition category. 3. The method of claim 2, The image processing unit, And the image processing is performed around an arbitrary fixed area of the vehicle passing through the upper road of the structure in the structure image so as to correspond to the classification criteria of the traffic condition category. The method of claim 1, An image collecting unit collecting the structure image photographed to include the entire image of the structure to be managed and transmitting the image to the image input unit; Disaster detection system further comprises a. The method of claim 1, A central manager configured to store the abnormal structure image received from the image manager so as to correspond to an internal database configured according to the structure state category, and determine whether or not a disaster is generated from the abnormal structure image; Disaster detection system further comprises a. The method of claim 8, The central management unit, And generating an alarm when the abnormal structure image is determined to be a disaster, and transmitting the abnormal structure image and the disaster determination result data to an administrator. An image input step of receiving a structure image photographing a structure to be managed; Performing image processing for reconstructing the structure image based on a classification criterion of a preset structure state category, and then performing image processing to compare the gray value of the structure image with a previously learned normal state gray value; And An image management step of dividing the abnormal structure image which does not match the standard image gray value and the image processing result data of the abnormal structure image by the structure state category and storing the image; Disaster occurrence detection method comprising a. The method of claim 10, The structure state category, Disaster occurrence detection method characterized in that it comprises one or more of the categories representing a plurality of structural conditions, such as water level status category, lost status category, fire status category, traffic status category. The method of claim 11, In the image processing step, And the image processing based on the position of the ruler displayed on the structure in the structure image so as to correspond to the classification criteria of the water level state category. The method of claim 11, In the image processing step, Disaster occurrence detection method characterized in that for performing the image processing around the outer shape of the structure in the structure image, so as to correspond to the classification criteria of the lost state category. The method of claim 11, In the image processing step, Disaster occurrence detection method, characterized in that for performing the image processing around a predetermined area of the upper portion of the structure in the image to correspond to the classification criteria of the fire condition category. The method of claim 11, In the image processing step, And the image processing is performed around an arbitrary fixed area of the vehicle passing through the upper road of the structure in the structure image so as to correspond to the classification criteria of the traffic condition category. The method of claim 10, Before the image input step, An image collection step of collecting and transmitting a structure image photographed to include the entire image of the structure to be managed; Disaster occurrence detection method further comprising a. The method of claim 10, After the image management step, A central management step of storing the abnormal structure image to correspond to an internal database configured according to the structure state category, and determining whether or not a disaster is generated from the abnormal structure image; Disaster occurrence detection method further comprising a. The method of claim 17, In the central management step, And generating an alarm when the abnormal structure image is determined to be a disaster, and transmitting the abnormal structure image and the disaster determination result data to an administrator.

Images