KR102620980B1 - Earthquake Monitoring Guidance Method Using Image Analysis Techniques - Google Patents

Abstract

The embodiment includes an IP-based camera module 10 that collects real-time video information and transmits the collected video information; And a voice broadcasting unit 20 that analyzes the degree of shaking of the image information transmitted from the IP-based camera module 10, determines the occurrence of an earthquake according to the analysis result, and transmits an evacuation broadcast according to the judgment result. This is about an earthquake monitoring response guidance method using video analysis techniques that can be used as CCTV and automatically transmit announcements when an earthquake occurs.

Description

Earthquake Monitoring Guidance Method Using Image Analysis Techniques}

The content disclosed in this specification relates to an earthquake monitoring response guidance method using image analysis techniques. More specifically, it determines whether an earthquake has occurred through video analysis based on sculptures within the shooting range of the IP-based camera module and transmits an earthquake evacuation announcement.

Unless otherwise indicated herein, the material described in this section is not prior art to the claims of this application, and is not admitted to be prior art by inclusion in this section.

Unfortunately, there is increasing interest in earthquakes of both large and small magnitude, and there is also a goal to develop earthquake warnings and, for example, provide early warning services for large earthquakes, that is, earthquakes of magnitude 6.0 or higher.

On the other hand, it is also used for research observation purposes. These earthquake early warnings are installed in mountains and sparsely populated areas with low background noise, and use expensive sabo (separate force balance) accelerometers and high-performance seismic data recorders. Additionally, in these existing cases, the installation cost per unit is high.

Meanwhile, since the magnitude 5.8 earthquake occurred in the Gyeongju area in September 2016, more than 600 aftershocks have continued, proving that South Korea is no longer an earthquake-safe zone.

Republic of Korea Patent No. 10-2011-0110680, a prior art, proposes an earthquake monitoring system using an earthquake sensor and Ethernet communication. However, the prior art requires an additional seismic sensor for earthquake monitoring and has the problem of not being able to perform any other functions other than earthquake monitoring.

In addition, Republic of Korea Patent No. 10-2016-0151684, which is a prior art, proposes a remote control system for earthquake preparedness for water treatment facilities using sensors and camera devices. However, the prior art requires a gyro sensor or acceleration sensor to be installed in the facility, and earthquake detection analysis using camera images is conducted after shaking is detected by the gyro sensor or acceleration sensor, so it does not provide immediate earthquake guidance through image information analysis. There is a problem that I cannot do.

In addition, Taiwan's prior art, an Earthquake monitoring system for measuring strength of earthquake via monitor, uses two or more movable image capture devices located on an object or rail to capture an object. A system that determines whether an earthquake has occurred based on whether the image moves is presented. However, the prior art has a problem in that the image capture device only captures images of specific objects for earthquake detection and does not provide other functions.

1. Republic of Korea Patent Application Number: 10-2011-0110680 2. Republic of Korea Patent Application Number: 10-2016-0151684 3. Taiwan patent application number: TW097218168

The purpose of the disclosed content is to provide an intelligent earthquake monitoring response guidance method that determines whether an earthquake has occurred through image analysis based on sculptures within the shooting range of an IP-based camera module and transmits an earthquake evacuation announcement. .

In addition, video information captured by an IP-based camera module can be transmitted to the network, and the video information can be recorded and output to a PC or TV screen, so it can be used as CCTV in normal times and provides immediate earthquake guidance by analyzing CCTV video information when an earthquake occurs. The purpose is to provide broadcasting.

The earthquake monitoring response guidance method using video analysis techniques according to the embodiment is:

In the case of earthquake monitoring, a method of providing response guidance using image analysis techniques in the control unit of the management information processing device,

A first step of setting a reference point by selecting a sculpture or object within the shooting range of the IP-based camera module used in the video analysis technique;

A second step of confirming and storing the location coordinates of the sculpture or object;

A third step of capturing image information of the IP-based camera module and checking whether a change value in the positional coordinates of the sculpture or object in the image information is greater than a set change value;

As a result of the confirmation, a fourth step of performing a set general monitoring mode if the change value of the position coordinates of the sculpture or object in the image information is less than a set change value, and performing a set emergency monitoring mode if it is more than the set change value; and

A fifth step of providing general monitoring information to the registration management information processing device when the set general monitoring mode is performed, and starting and performing the following operations when the set emergency monitoring mode is performed; Including,

Here, the operation of the emergency monitoring mode is as follows:

A first process of continuously calculating changes and distribution of positional coordinates for the sculpture or object; and

A second process of confirming the movement pattern and corresponding type of the sculpture or object by the change and distribution of the position coordinates;

Compare the confirmed movement patterns and types with the specific movement patterns and types of the sculpture or object according to the format, identify which movement patterns and types it belongs to, and determine the seismic sensitivity corresponding to the identified movement patterns and types. Third process of extracting; and

The occurrence of an earthquake is notified differently to the registration management information processing device based on the extracted seismic sensitivity information,

When making the above notification, from a number of different areas and groups or channel information according to user settings, the registration management information processing device and the registration manager terminal belonging to the corresponding area, group or channel in the surrounding area, each corresponding to the seismic sensitivity. Fourth process of notifying by text or voice; Including,

The format is:

a) From the video signal, the movement pattern and corresponding type are set differently for a number of different earthquakes and the corresponding seismic sensitivity for each type,

By classifying and learning the location coordinate changes and distribution information for each movement pattern and its corresponding type, the format from which the specific movement pattern and its corresponding type and seismic sensitivity are derived is defined.

b) Data sets on location coordinate changes and distribution are extracted for each movement pattern and seismic sensitivity for each type,

In each of the above datasets, the location coordinate change and distribution information representing the characteristics of each movement pattern and corresponding type are detected, and the location coordinate change and distribution information are detected differently for each seismic sensitivity information, so that each information The properties for are determined,

c) By normalizing the information for which the above properties have been determined, the change and distribution of location coordinates representing the characteristics of each different movement pattern and corresponding type of seismic sensitivity are obtained,

Independent variables (specific movement patterns and seismic sensitivity representing the characteristics of each movement pattern and type) and dependent variables (position coordinates) for calculating movement patterns representing each characteristic of each movement pattern and corresponding seismic sensitivity for each type. By setting change and distribution information,

d) Characterized by learning and training such information.

According to embodiments, an intelligent earthquake monitoring response guidance method is provided that determines whether an earthquake has occurred through image analysis based on sculptures within the shooting range of an IP-based camera module and transmits an earthquake evacuation announcement.

In other words, the IP-based camera module collects image information, sets the location coordinates of a sculpture or specific object within the image information shooting range as a standard, calculates the change in the location coordinates through image analysis, and calculates the change in location coordinates. Accordingly, it is effective in providing an intelligent earthquake monitoring response guidance system that determines whether an earthquake has occurred and transmits earthquake evacuation announcements.

In addition, video information captured by an IP-based camera module can be transmitted to the network, and the video information can be recorded and output to a PC or TV screen, so it can be used as CCTV in normal times and provides immediate earthquake guidance by analyzing CCTV video information when an earthquake occurs. It has the effect of providing broadcasting.

Figure 1 is a diagram showing a system applying an earthquake monitoring response guidance method using an image analysis technique according to an embodiment.
Figure 2 is a diagram showing images for CCTV and earthquake judgment functions in an earthquake monitoring response guidance method using a video analysis technique according to an embodiment.
Figure 3 is a schematic flowchart sequentially showing an earthquake monitoring response guidance method using an image analysis technique according to an embodiment.
Figure 4 is a diagram showing images for CCTV and earthquake judgment functions in an earthquake monitoring response guidance method using a video analysis technique according to another embodiment.

Figure 1 is a diagram showing a system applying an earthquake monitoring response guidance method using an image analysis technique according to an embodiment.

Referring to Figure 1, the guidance system according to a preferred embodiment is an intelligent system that first determines whether an earthquake has occurred through image analysis based on sculptures within the shooting range of the IP-based camera module and transmits an earthquake evacuation guidance broadcast. Provides guidance on earthquake monitoring and response.

In other words, image information is acquired through an IP-based camera module, the location coordinates of a sculpture or specific object within the image information shooting range are taken as a standard, the change in the location coordinates is calculated through image analysis, and the location coordinate change is calculated. It provides an intelligent earthquake monitoring response guidance method that determines whether an earthquake has occurred based on the value and transmits an earthquake evacuation announcement.

To this end, this configuration includes an IP (Internet Protocol)-based camera module 10, a voice broadcasting unit 20, a network 30, and a display unit 40.

The IP-based camera module 10 includes a camera 11 and an encoder 12.

The camera 11 captures video information, and the encoder 12 converts the video information so that it can be transmitted through a TCP/IP network and transmits it to the audio broadcasting unit 20 and the network 30. .

Additionally, the camera is an infrared camera and is capable of shooting at night.

The audio broadcasting unit 20 analyzes image information captured and transmitted by the IP-based camera module 10, determines whether an earthquake has occurred, and transmits an earthquake information broadcast according to the earthquake occurrence determination result.

The earthquake announcement is delivered to people near the CCTV by connecting the IP-based camera module 10 or the voice broadcasting unit 20 with a speaker, or is transmitted to the entire building or people near the building by connecting to the broadcasting equipment of the building. It is possible to deliver.

Additionally, the earthquake announcement may be a pre-recorded human guidance voice, a siren, or a fragmentary warning sound.

In the video information analysis method, the audio broadcasting unit 20 sets a reference point by selecting a sculpture or a specific object within a shooting range when installing the IP-based camera module 10; Confirming and storing the location coordinates of the sculpture or specific object; Capturing image information and continuously calculating changes in the location coordinates; and a step of notifying the occurrence of an earthquake when a change in the positional coordinates of the sculpture or a specific object exceeds a certain value due to an earthquake and the change in the positional coordinates continues for more than a certain period of time.

In addition, the IP-based camera module 10 can adjust the angle, so when resetting the reference point, the shooting area can be changed.

The network 30 can receive and store image information transmitted from the IP-based camera module 10 and transmit it to the display unit 40.

The display unit 40 includes a PC, TV, or smartphone, and displays image information transmitted from the network 30.

Figure 2 is a diagram showing images for CCTV and earthquake judgment functions in an earthquake monitoring response guidance method using a video analysis technique according to an embodiment.

Referring to Figure 2, specific objects are attached to the four corners within the camera radius and their position coordinates are matched. If an earthquake occurs during CCTV filming, the image shakes and the location coordinates of the specific object within the recording radius change. At this time, if the change value of the location coordinates increases beyond a certain value and continues for more than a certain time, it is determined that an earthquake has occurred, and the stored earthquake evacuation broadcast is immediately transmitted.

The specific object can be designated by using existing facilities such as existing statues, flower pots, or fire extinguishers, by attaching stickers to buildings or places, or by marking using paint or magic.

The marking, regardless of its shape, such as a circle, triangle, square, asterisk, or

Figure 3 is a schematic flowchart sequentially showing an earthquake monitoring response guidance method using an image analysis technique according to an embodiment.

Referring to FIG. 3, an IP-based camera module is installed at a location where filming is desired, and a reference point for earthquake detection image analysis is set within the filming radius (100).

Images are collected from the IP-based camera, and changes in the location coordinates of the reference point are monitored (101).

Due to the occurrence of an earthquake, the location coordinates change more than a certain value (102_YES), and if the change in the location coordinates continues for more than a certain time (103_YES), an earthquake evacuation announcement is sent (104) and earthquake monitoring is performed using video analysis techniques. The response guidance system ends.

If the change in the location coordinates is insignificant and does not change by more than a certain value (102_NO), image collection and change monitoring continue (101).

If the position coordinates change by more than a certain value (102_YES), and the change in the position coordinates does not last more than a certain time (103_NO), image collection and change monitoring continue (101).

Therefore, through this, one embodiment provides an intelligent earthquake monitoring response guidance method that determines whether an earthquake has occurred through image analysis based on sculptures within the shooting range of the IP-based camera module and transmits an earthquake evacuation announcement. .

In addition, this video information can be recorded and output to a PC or TV screen, so it can be used as CCTV in normal times, and when an earthquake occurs, CCTV video information is analyzed to provide immediate earthquake information broadcasting.

More specifically, it is as follows.

First, as in the past, this method is based on the assumption that when conducting earthquake monitoring, the control unit of the management information processing device provides response guidance using image analysis techniques.

So, in this state, according to one embodiment, first, a sculpture or object within the shooting range of the IP-based camera module used in the image analysis technique is selected to set a reference point.

Then, the location coordinates of the sculpture or object are confirmed and stored.

Next, image information of the IP-based camera module is captured, and it is checked whether the change value of the position coordinates of the sculpture or object in the image information is greater than the set change value. For example, these changes may be set or confirmed according to changes in the frequency or vibration of earthquakes.

As a result of the confirmation, if the change value of the position coordinates of the sculpture or object in the image information is less than the set change value, the set general monitoring mode is performed, and if it is more than the set change value, the set general monitoring mode is performed.

On the other hand, when the set general monitoring mode is performed, general monitoring information is provided to the registration management information processing device, and when the set emergency monitoring mode is performed, the following operations are started and performed.

Here, the operation of the above-mentioned emergency monitoring mode is as follows.

First, the change and distribution of positional coordinates for the sculpture or object are continuously calculated.

Next, the movement pattern and corresponding type of the sculpture or object are confirmed by the change and distribution of the position coordinates.

Then, the identified movement pattern and type are compared with the specific movement pattern and type of the sculpture or object according to a deep learning or YORO learning format, and it is determined which movement pattern and type it belongs to.

Therefore, the seismic sensitivity corresponding to the identified movement pattern and type is extracted, and information regarding the occurrence of an earthquake is notified differently to the registration management information processing device for each of the multiple different seismic sensitivities. For example, in these cases, earthquake sensitivity is classified using the saved wave file, and notifications are made according to this information.

In addition, when making the above notification, a plurality of registration information processing devices and registration manager terminals belonging to the corresponding areas, groups or channels are sent from the area and group or channel information for a number of different locations according to user settings. Provide text or voice notifications corresponding to different seismic sensitivities.

For example, in situations where earthquake sensitivity appears to be low, notification is sent by text, in situations where earthquake sensitivity appears to be moderate, notification is sent via emergency text, and in more difficult situations, the notification is converted to voice to be notified by a management information processing device, etc. .

Meanwhile, in this case, the format is also as follows.

a) First, from the video signal, the movement pattern and corresponding type are set differently for a number of different earthquakes and the corresponding seismic sensitivity for each type.

In addition, by classifying and learning the location coordinate change and distribution information for each movement pattern and its corresponding type, a specific movement pattern and its corresponding type, and the format from which the seismic sensitivity is derived, are defined.

b) Next, data sets on location coordinate changes and distribution are extracted for each movement pattern and seismic sensitivity for each type.

In addition, by detecting the position coordinate change and distribution information representing the characteristics of each movement pattern and corresponding type in each of the above datasets, and detecting seismic sensitivity in response to each different position coordinate change and distribution information, properties are determined.

c) By normalizing the information for which the above properties have been determined, the position coordinate changes and distributions representing the characteristics of each different movement pattern and corresponding type of seismic sensitivity are obtained.

In addition, independent variables (specific movement patterns and seismic sensitivity representing the characteristics of each movement pattern and type) and dependent variables ( Set location coordinate change and distribution information).

d) So, learn and train this information.

Therefore, through this, one embodiment provides an intelligent earthquake monitoring response guidance method that determines whether an earthquake has occurred through image analysis based on sculptures within the shooting range of the IP-based camera module and transmits an earthquake evacuation announcement. .

That is, specifically, the IP-based camera module collects image information, sets the location coordinates of a sculpture or specific object within the image information shooting range as a standard, calculates the change in the location coordinates through image analysis, and calculates the location coordinates. It provides an intelligent earthquake monitoring response guidance method that determines whether an earthquake has occurred according to the calculated change value and transmits an earthquake evacuation announcement.

In addition, this video information can be recorded and output to a PC or TV screen, so it can be used as CCTV in normal times, and when an earthquake occurs, CCTV video information is analyzed and immediate earthquake information broadcast is made.

On the other hand, when confirming the type with this movement information, this method determines whether there is a direct type of earthquake using the method below, and in the case of a direct type earthquake, the operation of the facility is directly controlled.

In other words, in this case, the facility is directly controlled, and in other cases, the facility is controlled according to external commands.

For example, these facilities include gas locks and lighting devices. So, through this, the lights can be flashed directly or indirectly, or the gas can be directly turned off. In addition, for example, the facilities may differ depending on whether the building is built on bedrock or general soil.

Specifically, as described above, when confirming the type of earthquake, whether or not there is direct earthquake sensitivity is determined depending on whether the size of the movement pattern is greater than the set size.

Therefore, as a result of the above judgment, if the size of the movement pattern is greater than the set size, it is judged as a direct earthquake sensitivity, that is, it is judged as a direct earthquake and the operation of the facility is directly controlled, and if it is less than the set size, the operation of the facility is directly controlled. Try to control it indirectly.

Figure 4 is a diagram showing images for CCTV and earthquake judgment functions in an earthquake monitoring response guidance method using a video analysis technique according to another embodiment.

Referring to Figure 4, specific objects are attached to the four corners within the camera radius and their position coordinates are matched. With image capture, the change in position coordinates is calculated. At this time, the position coordinate change calculation is performed only in the area marked in yellow and containing the specific object, thereby speeding up image analysis.

If an earthquake occurs during CCTV filming, the image shakes and the location coordinates of the specific object within the recording radius change. At this time, if the change value of the location coordinates increases beyond a certain value and continues for more than a certain time, it is determined that an earthquake has occurred, and the stored earthquake evacuation broadcast is immediately transmitted.

On the other hand, this configuration allows for the simple use of road facilities that are frequently seen in the above-mentioned sculpture. This is because video surveillance devices (i.e., IP-based camera modules) can be found in many places, making installation and operation convenient. It can be easily used for.

Additionally, these road facilities are installed on the road for the convenience of drivers and pedestrians to help them find their way. For example, they include street lights, traffic lights, and guidance signs.

For example, guide signs provide directions to drivers while they are driving, or traffic signals are displayed so that drivers can recognize road conditions in advance, allowing drivers to drive defensively in advance.

However, when trying to use road facilities in this way, there is an existing management device for road facilities, that is, a so-called safety monitoring information processing device for road facility management, which repeatedly checks whether the road facilities are in normal condition. Confirm.

Therefore, in this specification, when using the above-mentioned sculpture, for example, when using a safety sign as a sculpture, abnormalities in the sculpture are repeatedly checked in conjunction with this information processing device to first check whether the sculpture is in a normal state. Easily determine whether or not

For this purpose, first, the video surveillance device repeatedly receives road facility information from the above-mentioned safety monitoring information processing device, and detects whether there is an abnormality in the road facility from this road facility information, that is, in this case, detects whether there is an abnormality in the sculpture and returns to normal state Determine whether you are aware of it or not.

Additionally, in this case, by receiving road facility information from each safety monitoring information processing device for multiple different devices, that is, receiving information appropriate for each location, the status of road facilities, i.e., sculptures, is collectively determined for each location. Check to ensure smooth installation and maintenance.

For example, the shape of the sculpture in the road facility information is compared with the shape of the setting sculpture inside the device, and according to the comparison result, that is, if the shape of the sculpture in the road facility information and the shape of the setting sculpture are the same, the shape is returned to normal. Determine.

On the other hand, if, as a result of the comparison, the appearance of the sculpture in the road facility information is different from the appearance of the setting sculpture, it is determined to be in an abnormal state, and a notification is made to the administrator, that is, to the registration management information processing device.

Another example is a method that links to the method described above, for example, as follows.

First, as described above, when a sculpture or object within the shooting range of the IP-based camera module is selected and a reference point is set, the reference point for the sculpture in the road facility information is compared with the reference point for the set sculpture, and the reference point is normalized according to the comparison result. It can also determine the status.

In other words, as a result of the above comparison, if the reference point confirmed by the position coordinates of the sculpture in the road facility information and the reference point of the set sculpture are the same or within the set error range, it is determined to be in a normal state, and if they are different or outside the set error range, It is determined to be in an abnormal state.

That is, when an abnormality in the sculpture is detected, the abnormality information is transmitted to the registration management information processing device and the manager is notified, so that the sculpture can be managed and maintained in a normal state.

Another example is a method that links to the method described above, for example, as follows.

First, as described above, when a sculpture or object within the shooting range of the IP-based camera module is selected and a reference point is set, the reference point for the sculpture in the road facility information is compared with the reference point for the set sculpture, and the reference point is normalized according to the comparison result. It can also determine the status.

To this end, the reference point for the sculpture suitable for the corresponding location, that is, the road facility, is set and registered for each different sculpture for each of the multiple different devices. Additionally, these reference points may be set slightly in the center or on the sides depending on the shape of each sculpture, for example, a circle, triangle, or square.

In other words, as a result of the above comparison, if the reference point confirmed by the position coordinates of the sculpture in the road facility information and the reference point of the set sculpture are the same or within the set error range, it is determined to be in a normal state, and if they are different or outside the set error range, It is determined to be in an abnormal state.

In addition, when abnormalities in road facilities are detected in this way, that is, when abnormalities are detected under normal circumstances, a notification is sent to the relevant location through the RSU (Road Side Unit) to notify drivers traveling in places where abnormal facilities are located. You can also let them know. In other words, notifications are sent to each place where abnormalities are detected to facilitate smooth passage.

On the other hand, if the seismic sensitivity is recognized to a certain degree as described above through these sculptures, that is, in an emergency situation, notification is sent to the relevant place and surrounding places within a set radius through RSU, so that the relevant place and surrounding places are notified. Be sure to inform the driver, etc.

For example, these notifications set locations in response to different seismic sensitivities.

Specifically, these methods are as follows.

First, each sculpture is set as a road facility for a number of different IP-based camera modules, and partial information representing reference point information and sculpture characteristics is set and registered, and information to collectively estimate the overall size and location of each sculpture is provided for each partial information. Register settings.

For example, in the case of a guide sign, the overall size and location are estimated from a specific location in the 'ㄷ' shape.

Additionally, RSU (Road Side Unit) information corresponding to the location where the IP-based camera module belongs is set and registered.

Next, in this state, each of the multiple different IP-based camera modules receives road facility information repeatedly from each registered road facility safety monitoring information processing device.

Therefore, this road facility information is converted into the corresponding sculpture information format for each IP-based camera module.

Then, partial information representing characteristics is extracted from this sculpture information.

Additionally, from this extracted partial information, the overall size and location of each sculpture are estimated to obtain sculpture information.

Then, the reference point of the sculpture information set above is compared with the reference point of the corresponding sculpture information within the obtained road facility information.

As a result of the comparison, if the reference point of the set sculpture information and the reference point of the corresponding sculpture information in the received road facility information are the same, it is determined to be in a normal state, and if they are different, it is determined to be in an abnormal state.

As a result of the determination, if an abnormal state is determined, notification is made to the registration management information processing device, and notification is made to the registration RSU corresponding to the location where the sculpture is abnormal for each location.

Additionally, when notifying by text or voice according to the seismic sensitivity according to the above-mentioned emergency monitoring mode settings, notification is made differently by text or voice for each registered RSU corresponding to the place where the sculpture belongs and the surrounding places. do.

In addition, in this configuration, when performing notification in this way, it is determined whether the seismic sensitivity corresponds to a direct earthquake depending on whether the size of the movement pattern for this sculpture is greater than the set size.

Therefore, as a result of the above judgment, if the size of the movement pattern is larger than the set size, it is determined by the seismic sensitivity directly below and a voice notification is made for each registered RSU corresponding to the place where the sculpture belongs and the surrounding places.

10: IP-based camera module
11: Camera
12: encoder
20: Voice broadcasting department
30: Network
40: display unit

Claims (3)

In the case of earthquake monitoring, a method of providing response guidance using image analysis techniques in the control unit of the management information processing device,
A first step of setting a reference point by selecting a sculpture or object within the shooting range of the IP-based camera module used in the video analysis technique;
A second step of confirming and storing the location coordinates of the sculpture or object;
A third step of capturing image information of the IP-based camera module and checking whether a change value in the positional coordinates of the sculpture or object in the image information is greater than a set change value;
As a result of the confirmation, a fourth step of performing a set general monitoring mode if the change value of the position coordinates of the sculpture or object in the image information is less than a set change value, and performing a set emergency monitoring mode if it is more than the set change value; and
A fifth step of providing general monitoring information to the registration management information processing device when the set general monitoring mode is performed, and starting and performing the following operations when the set emergency monitoring mode is performed; Including,
Here, the operation of the emergency monitoring mode is as follows:
A first process of continuously calculating changes and distribution of positional coordinates for the sculpture or object; and
A second process of confirming the movement pattern and corresponding type of the sculpture or object by the change and distribution of the position coordinates;
Compare the confirmed movement patterns and types with the specific movement patterns and types of the sculpture or object according to the format, identify which movement patterns and types it belongs to, and determine the seismic sensitivity corresponding to the identified movement patterns and types. Third process of extracting; and
Different notifications are sent to the registration management information processing device according to the extracted seismic sensitivity,
When making the above notification, from the area and group or channel information for a number of different places according to user settings, to the registration management information processing device and the registration manager terminal belonging to the corresponding area and group or channel in the surrounding area, a number of different seismic sensitivities are provided. A fourth process of providing different notifications with corresponding text or voice for each type; Including,

The format is:
a) From the video signal, the movement pattern and corresponding type are set differently for a number of different earthquakes and the corresponding seismic sensitivity for each type,
By classifying and learning the location coordinate change and distribution information for each movement pattern and its corresponding type, a specific movement pattern and its corresponding type, and a format in which seismic sensitivity according to the type is derived are defined,

b) Data sets on location coordinate changes and distribution are extracted for each movement pattern and seismic sensitivity for each type,
In each of the above datasets, the location coordinate change and distribution information representing the characteristics of each movement pattern and corresponding type are detected, and the seismic sensitivity is detected for each different location coordinate change and distribution information, so that the properties of each information are determined. become,

c) By normalizing the information for which the above properties have been determined, the change and distribution of location coordinates representing the characteristics of each different movement pattern and corresponding type of seismic sensitivity are obtained,
Independent variables (specific movement patterns and seismic sensitivity representing the characteristics of each movement pattern and type) and dependent variables (position coordinates) for calculating movement patterns representing each characteristic of each movement pattern and corresponding seismic sensitivity for each type. By setting change and distribution information,

d) Constructing learning and training data with the set position coordinate change and distribution information; Earthquake monitoring response guidance method using video analysis techniques characterized by . In claim 1,
After the first step,
For each different IP-based camera module, each sculpture is set as a road facility to set and register reference point information and partial information representing the sculpture characteristics, and for each partial information, information that collectively estimates the overall size and location of each sculpture is set and registered.
Step 1-1 of setting and registering RSU (Road Side Unit) information corresponding to the location to which the IP-based camera module belongs;
Steps 1-2 of repeatedly receiving road facility information from each registered road facility safety monitoring information processing device for each of the multiple different IP-based camera modules;
Steps 1-3 of converting the received road facility information into a sculpture information format set by an IP-based camera module;
Steps 1-4 of extracting the corresponding sculpture information for each IP-based camera module from the converted road facility information into partial information representing the characteristics of the sculpture;
Steps 1-5 of collectively estimating the overall size and position of each sculpture using the extracted partial information;
Steps 1-6 of comparing the reference point of the set sculpture information with the reference point of the corresponding sculpture information in the estimated road facility information;
As a result of the comparison, if the reference point of the set sculpture information and the reference point of the corresponding sculpture information in the received road facility information are the same, it is determined to be in a normal state, and if they are different, it is determined to be in an abnormal state; and
Steps 1-8 of notifying the registration management information processing device if an abnormal state is determined as a result of the determination, and notifying the registration RSU corresponding to the location where the sculpture is abnormal for each location; It further includes,

After the fifth step above,
When notifying by text or voice according to the seismic sensitivity according to the emergency monitoring mode set above, step 5-1 of performing different notifications by text or voice for each registered RSU corresponding to the place where the sculpture belongs and surrounding places. ; An earthquake monitoring response guidance method using an image analysis technique, further comprising: In claim 2,
The third process is,
When checking the type of the above earthquake, it is determined whether the seismic sensitivity corresponds to a direct earthquake depending on whether the size of the movement pattern is greater than the set size,
As a result of the above determination, if the size of the movement pattern is greater than the set size, a 3-1 process of performing a voice notification for each registered RSU corresponding to the place where the sculpture belongs and surrounding places by determining the seismic sensitivity directly below; An earthquake monitoring response guidance method using an image analysis technique, further comprising: