KR20230059187A - River safety management system and method using artificial intelligence - Google Patents

Abstract

The present invention relates to a river use safety management system using artificial intelligence, and more specifically, image information is collected using various types of cameras for a measurement target, for example, a river basin, and water level, flow rate, temperature, and humidity are collected through physical sensing. On the other hand, when collecting hyperspectral image information from the sky, an artificial intelligence analysis device analyzes the collected information and accumulates analysis information, while applying continuous analysis and 3D virtual reality to analyze the information. Real-time monitoring and prediction of the occurrence of disasters, including bank collapse or floods, providing intuitive information and risk situations to the relevant river management entity and the public using the river, and ingress and drift of inappropriate people or vehicles in the river It is about a river use safety management system and safety management method using artificial intelligence that can analyze whether or not to respond to an emergency.

Description

River safety management system and method using artificial intelligence {River safety management system and method using artificial intelligence}

The present invention relates to a river use safety management system using artificial intelligence and a method for safely managing river use thereby (safety management method), and more particularly, using various types of cameras for a river or river basin to be monitored. image information is collected, and environmental information including water level, flow rate, temperature and humidity is collected through physical sensing, while hyperspectral image and RGB image information are collected from above, and the collected image information, environmental information and hyperspectral image information are collected. By analyzing image information with an artificial intelligence analysis device, accumulating the analyzed information (analysis information) and incorporating 3D virtual reality, real-time monitoring and prediction of the occurrence of disasters, including levee collapse or flooding, and intuitive information about this and risk situations to the relevant river management entity and the public using the river, as well as to analyze whether inappropriate people or vehicles have entered or drifted in the river to respond to emergency situations. management system and safety management method".

In general, a river (river, stram, river) means that water supplied to the surface through precipitation forms a constant water path and flows down an inclined surface. A certain waterway constituting a river is called a channel, a low valley formed by or flowing through a river is called a river valley, and the bottom part of a river made of rocks or soil is called a river bed. , riverbed). Resistance to flow in natural rivers where the river bed or embankment material is mainly composed of sand and gravel, or in vegetative rivers in which vegetation such as grass, shrubs, and trees are scattered in floodplains and embankments. This embankment and riverside cause changes in flow velocity according to vegetation and river bed data.

On the other hand, embankment means an artificial facility to block the natural (natural) flow of water, and its contents mean the meaning of a dam for storage and a dam for waterproofing. It has two meanings and functions. Among the above two meanings and functions, the embankment refers to the embankment of a general river according to the River Management Act. In Korea, large and small rivers are It refers to a facility built to prevent flood damage that causes farmland to be washed away or buried due to flooding.

Such an embankment is usually built wide enough so that the river flowing at a high water level does not permeate and collapse or be eroded, and as shown in FIG. defend Securing the safety of river basins against unpredictable natural disasters such as earthquakes and continuously guiding the public on the condition of rivers and river basins to the public who use river basins, and promptly evacuating by warning users and managers when disasters are predicted A safety device is required to make this possible. In particular, since earthquake-induced cracks in river banks can cause secondary damage such as bank collapse, monitoring and forecasting are very important.

For reference, in Korea, as strong earthquakes occurred in Gyeongju in September 2016 and Pohang in November 2017, the need for smart ground and structure management is growing. In particular, there is a strong demand for investigation and monitoring using advanced equipment and new ICT convergence technology for the advancement of structure management. Currently, 16,700 reservoirs, or 95% of all reservoirs in the country, have been created for more than 30 years, and among them, reservoirs older than 50 years are at increased risk of leakage. However, due to lack of budget, it was found that there are limitations in practical safety management such as maintenance, repair, reinforcement, and precise safety diagnosis of disaster risk reservoirs. In particular, in the case of Gyeongbuk, there are 479 reservoirs older than 50 years in the province, so the possibility of secondary damage after an earthquake is very high. In addition, after the Pohang earthquake, intensive analysis was conducted on 126 reservoirs vulnerable to disasters and reservoirs weighing 100,000 tons or more within a radius of 50 km from the epicenter, shear cracks in embankments, drainage facilities, and spillways, leaks, and operation of water intake facilities. It is a reality that accurate analysis is difficult because it is carried out.

Rivers including embankments are managed in such a way that after a disaster event, such as an earthquake or a flood, occurs, a manager periodically visits the corresponding embankment to visually check whether or not damage has occurred to the building structure. However, there is a limit to such a manager's visual inspection, and recently, an attempt has been made to analyze the current state by remotely monitoring the embankment through a sensor or/and aerial photograph. Prior art related to monitoring of rivers including embankments includes Korean Patent Registration No. 10-1204426 and Korean Patent Registration No. 10-1881822.

However, in the case of this prior art, since a measuring means (measurement sensor) is buried in the flood control facility and the method of measuring the change in the slope of the flood control facility when the collapse of the flood control facility starts, the danger is recognized at the time of collapse or just before the collapse occurs. There is a problem that it is difficult to secure enough necessary time to cope with this. In addition, since the measurement means is buried in the flood control facility, there is a problem in that maintenance costs are high due to the repeated use of the measurement means being impossible due to the loss or damage of the measurement means during collapse.

On the other hand, in the case of the prior art of collecting and analyzing information on the surface material (low water level) of a reservoir including an embankment through aerial photographs, multi-dimensional and three-dimensional information collection and analysis through this are not performed, and accordingly, the embankment There is a problem of low precision or accuracy as well as limitations of information analysis on the reservoir including the reservoir.

Republic of Korea Patent Registration No. 10-1204426 (Announced on November 26, 2012) Republic of Korea Patent Registration No. 10-1881822 (published on August 24, 2018)

The present invention was developed to solve the problems of the prior art as described above, and collects image information using various types of cameras for monitoring objects, for example, rivers or river basins, and uses physical sensing to perform water level, flow rate, temperature, and humidity. On the other hand, collects environmental information, including hyperspectral image information, and analyzes the terrain-image information and environmental information collected through an artificial intelligence analysis device, while accumulating analysis information to provide continuous analysis and 3D virtual reality. By analyzing information while grafting, monitoring and predicting the occurrence of disasters including bank collapse or flooding in real time, providing intuitive information and risk situations to the relevant river management entity and the public using the river, and providing information within the river. Its purpose is to provide a "safety management system and safety management method for river use using artificial intelligence" that can respond to emergencies by analyzing inappropriate person or vehicle acquisition and drifting.

In order to achieve the above object, in the present invention, topographic information including hyperspectral image information and RGB image information is collected in the sky above the monitoring target, that is, the river including the embankment, while multiple points on the ground or above the water including the corresponding embankment. A river information collection device 100 that collects environmental information including water level, flow rate, temperature, and humidity including actual image information in real time; By analyzing the information data collected by the river information collection device 100 and transmitted in real time through the communication network, and continuously analyzing while accumulating the analysis information, whether a disaster including a bank collapse or flood occurs and inappropriate people or vehicles in the river Artificial intelligence analysis device that analyzes the acquisition and drifting of , while implementing image information and environmental information, including the topography to which the analyzed information is grafted, as 3D virtual space images, storing them, and transmitting them to the remote management device 300 at the same time ( 200); And the information transmitted from the river information collection device 100 is transmitted to the artificial intelligence analysis device 200 so that the analysis of the artificial intelligence analysis device 200 is performed, but as a result of the analysis, whether or not there is an abnormality in the river and an abnormal change occurs If this is predicted, the 3D virtual space image and abnormal change alarm are transmitted and guided to the local residents' personal communication terminals 400 subscribed to the forecast service in advance, and temporary stay or 3D virtual space images and abnormal change alerts are transmitted and guided to the personal communication terminal (500) of other people passing through, while announcements are broadcast through an output device adjacent to the location when it is determined that an inappropriate person or vehicle is entering or drifting in the river artificial intelligence characterized by having a configuration comprising; A river use safety management system is provided.

In addition, in the present invention, a method for managing the safe use of a river through the above-described river use safety management system, that is, a river use safety management method is provided.

In the present invention, when the information data collected by the river information collection device 100 is transmitted from the remote management device 300, the artificial intelligence analysis device 200 determines whether a disaster has occurred, the number of inappropriate people or vehicles in the river. An information analysis program that generates predictive information by continuously conducting and informatizing learning analysis while accumulating analyzed data, analyzing in real time whether it is ingested or drifting, and the possibility of disaster occurrence, and terrain-image information, actual image information, environment Analysis and 3D imaging unit 210 including a 3D virtual space program that implements a 3D virtual space image by combining information and analyzed prediction information; and an analysis information management unit 220 that stores information and data predicted and generated by the information analysis program and the 3D virtual space program and transmits them to the remote management device 300 at the same time.

In addition, in the present invention, the remote management device 300 sends a 3D virtual space image and an abnormal change alarm to local residents' personal communication terminals through a communication network when it is determined that there is an abnormality or an abnormal change has occurred as a result of the analysis of the artificial intelligence analysis device 200. (400) and other personal communication terminals (500), while transmitting and guiding, if it is determined that an inappropriate person or vehicle is entering or drifting in the river, an announcement broadcast is sent out through an output device adjacent to the location, and 119 Safety Center An information processing and forecast management unit loaded with an algorithm or program for transmitting an emergency situation along with video information to the emergency response device 600 of the emergency response department including a may be provided.

In the present invention, the remote management device 300 is a remote management device, and includes the terrain information and image information collection unit 110, the environment information collection unit 120, and the artificial intelligence analysis device constituting the river information collection device 100 ( 200) an operation communication unit for communication connection; A monitor unit outputting the information transmitted from the artificial intelligence analysis device 200 and the final analyzed information; An operation storage unit for storing information transmitted from the artificial intelligence analysis device 200 and finally analyzed information; Based on the GPS information, the coordinate values for the shooting area are calculated and stored, while transmitted to the topographic information and image information collection unit 110, or remotely controlling the movement and operation of the topographic information and image information collection unit 110, Image information, environment information, analyzed prediction information, and 3D virtual space images may be output to the monitor unit and stored in the operation storage unit, while further comprising an operation control unit for controlling the operation of the operation communication unit.

In the present invention, the information processing and forecast management unit performs a personal information collection, confirmation, and approval operation in advance for the local residents' personal communication terminal 400 and forecast service use to make a forecast, a local forecast service program; And if the personal communication terminal 500 of another person is captured in the region through the communication network, if the forecast service is used after checking whether or not to use it, the other person who collects personal information, checks, and performs approval operations so that the forecast is made A forecast service program may be further included.

The present invention by means of the above problem solving means collects hyperspectral image information and RGB image information at specific points at intervals along the river including the embankment of the monitoring target area, and the topographic information and image information collection unit and the water level and flow velocity By installing an environmental information collection unit that collects environmental information including temperature and humidity, image information is collected using various types of cameras for the area to be monitored, and environmental information including water level, flow rate, temperature and humidity is collected through physical sensing. On the other hand, hyperspectral image information is collected and analyzed from the sky, and 3D virtual reality is applied to analyze information on river conditions, predict levee collapse and flood possibility, and provide intuitive information and emergency or dangerous situations. The remote management device of the relevant river management entity monitors and analyzes, and provides information to local residents' personal communication terminals or/and other people's personal communication terminals temporarily passing through or staying in the area, in case of disasters on rivers including embankments. It has the effect of allowing civilians to react quickly.

In addition, in the present invention, when a stranger temporarily stays in a river basin including an embankment, for example, camping, accommodation using a car, camping, etc., accesses the remote management device of the subject who manages the river including the embankment in advance. Therefore, after agreeing to the use of the forecast service and the use of personal information, and signing up for the service after identity verification, by receiving environmental information about the area in real time, other people temporarily staying or passing through the area Get the effect of being able to respond quickly to disaster situations.

In addition, the present invention collects image information using various types of cameras for the monitoring target area, collects environmental information including water level, flow rate, temperature and humidity through physical sensing, and collects hyperspectral image information from the sky. Analysis and 3D virtual reality are applied, and when it is determined that an inappropriate person or vehicle is ingested or drifted in the river, an announcement is sent through an output device adjacent to the location, and emergency response departments including 119 Safety Center By transmitting an emergency situation along with video information to a response device, an effect of emergency response and rescue is obtained.

1 is a photograph in place of a drawing showing an example of an embankment.
2 is a schematic system configuration diagram briefly showing the configuration of the present invention.
3 is a system block diagram showing the configuration of the present invention in detail.
4 is a block diagram showing an example of collecting terrain information, spectroscopic image and RGB image information, and environment information around a river by the topographic information and image information collection unit and the environment information collection unit according to the present invention.
5 is a schematic block diagram of the configuration of a hyperspectral image analysis program of a geographic information and image information collection unit.
6 is a schematic configuration diagram showing an example of capturing an actual image and extracting an image and terrain data from it in the present invention.
7 is a schematic block diagram of the configuration of an artificial intelligence analysis device.
Figure 8 is a schematic block diagram showing the configuration of the information analysis program provided in the analysis and 3D imaging unit of the artificial intelligence analysis device.
9 is a schematic block diagram showing the configuration of a 3D virtual space program provided in the analysis and 3D imaging unit of the artificial intelligence analysis device.
10 is a schematic block diagram showing the configuration of a remote management device provided in the river use safety management system according to the present invention.
11 is a schematic illustration showing an example of setting an area to implement a 3D virtual space.
12 is a virtual view showing an example of an image implementing a 3D virtual space.

Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings. The present invention has been described with reference to the embodiments shown in the drawings, but this is described as one embodiment, and thereby the technical idea of the present invention and its core configuration and operation are not limited. For reference, in the disclosure of this specification, including the claims, like reference numerals designate like elements. In describing the embodiments of the present invention, detailed descriptions of well-known functions or configurations may be omitted. Combinations of each block of the accompanying block diagram and each step of the flowchart may be performed by computer program instructions (execution engine), and these computer program instructions are executed by a processor of a general-purpose computer, special-purpose computer, or other programmable data processing equipment. Since it can be mounted, the instructions executed through the processor of a computer or other programmable data processing equipment create means for performing the functions described in each block of the block diagram or each step of the flowchart. These computer program instructions may also be stored in a computer usable or computer readable memory that can be directed to a computer or other programmable data processing equipment to implement functionality in a particular way, such that the computer usable or computer readable memory The instructions stored in are also capable of producing an article of manufacture containing instruction means for performing the functions described in each block of the block diagram or each step of the flow chart. And since the computer program instructions can also be loaded on a computer or other programmable data processing equipment, a series of operating steps are performed on the computer or other programmable data processing equipment to create a computer-executed process to create a computer or other programmable data processing equipment. It is also possible that the instructions performing the data processing equipment provide steps for executing the functions described in each block of the block diagram and each step of the flow chart. Additionally, each block or each step may represent a module, segment or portion of code that contains one or more executable instructions for executing specified logical functions, and in some alternative embodiments referred to in blocks or steps. It is also possible that the functions described may occur out of sequence. For example, two blocks or steps shown in succession may in fact be performed substantially concurrently, or the blocks or steps may be performed in the reverse order of their corresponding functions, if necessary. In particular, as used herein, '... module', '… unit', '… wealth', '… energy', '… A term such as 'part' refers to a unit that processes at least one function or operation, and may be implemented as hardware, software, or a combination of hardware and software.

In addition, the river use safety management system and river use safety management method using artificial intelligence according to the present invention are abbreviated as "river use safety management system" and "river use safety management method", respectively, and The streams and river basins included are abbreviated as “monitored stream zones”.

2 is a schematic system configuration diagram briefly showing the configuration of the present invention, and FIG. 3 is a system block diagram showing the configuration of the present invention in detail. As shown in FIGS. 2 and 3, the river use safety management system according to the present invention has a configuration including a river information collection device 100, an artificial intelligence analysis device 200 and a remote management device 300. .

The river information collection device 100 collects "topographic-image information" including hyperspectral image information and RGB image information from the sky above the river area to be monitored, and "actual image information" and "real image information" at multiple points in the river area to be monitored. It collects "environmental information" including water level, flow rate, temperature and humidity of the river in real time.

The information and data collected by the river information collection device 100 are transmitted to the artificial intelligence analysis device 200 in real time through a communication network, and the artificial intelligence analysis device 200 analyzes the transmitted information and data and analyzes the information. Accumulation and continuous analysis are performed to analyze whether a disaster has occurred, including bank collapse or flooding, and whether inappropriate people or vehicles have entered or drifted in the river, while topographic-image information, actual image information, environmental information, and analyzed predictions The information is integrated, implemented as a 3D virtual space image, stored, and transmitted to the remote management device 300 at the same time.

The remote management device 300 transmits information and data transmitted from the river information collection device 100 to the artificial intelligence analysis device 200. In addition, in the remote management device 300, according to the analysis result of the artificial intelligence analysis device 200, if an abnormality occurs in the river area to be monitored and an abnormal change is predicted, local residents' personal communication terminals subscribed to the forecast service in advance (400) and other personal communication terminals (500), 3D virtual space images and abnormal change alerts are transmitted and guided, and when it is determined that an inappropriate person or vehicle is entering or drifting in the river, announcements are broadcast through an output device adjacent to the location is transmitted, and the emergency situation is transmitted along with the image information to the emergency response device 600 of the emergency response department.

In the river use safety management method according to the present invention, topography-image information including hyperspectral image information and RGB image information is collected for the river area to be monitored through the river information collection device 100, and the Real image information and environmental information including water level, flow rate, temperature and humidity are collected in real time from multiple points on the ground or on the water, and are collected and transmitted by the river information collection device 100 through the artificial intelligence analysis device 200. Based on the information and data to be monitored, analyze whether there is a disaster including bank collapse or flooding in the river area to be monitored, whether inappropriate people or vehicles are entering or drifting in the river, and analyze topographic-image information and environmental information. It is implemented, stored, and output as a 3D virtual space image by grafting it with predicted information (prediction information). According to the analysis results, if an abnormality occurs or an abnormal change is predicted in the river area to be monitored, or if a situation such as a person entering or drifting in the river occurs, an announcement is made through an output device adjacent to the location. and transmits the emergency situation along with the image information to the emergency response device 600 of the emergency response department.

Specifically, the river information collection device 100 is a topographic information and image information collection unit ( 110), and environmental information that is installed at multiple points along the river in the river area to be monitored to collect actual image information about the river area to be monitored and collect environmental information including water level, flow rate, temperature and humidity through physical sensing It has a configuration including the collection unit 120.

4, in the present invention, the topographic information and image information collection unit 110 and the environment information collection unit 120 provide topography-image information including hyperspectral image information and RGB image information of the river area to be measured, and actual image information. , And a schematic configuration diagram showing an example of collecting environmental information is shown.

In the present invention, the terrain information and image information collection unit 110 collects "terrain-image information" including hyperspectral image information and RGB image information for a monitoring target. The topographic information and image information collection unit 110 is provided with a photographing device for capturing hyperspectral image information and RGB image information. It may be configured to include a hyperspectral sensor (hyperspectral camera) having spectral information characteristics in the infrared region and a digital camera such as a DSLR for capturing RGB images.

In the topographic information and image information collection unit 110, the photographing device is fixed at a certain height of the monitoring target or installed on a flying vehicle such as a drone to acquire information while moving along the monitoring target. Here, the flight vehicle may be one for which a manager directly adjusts the flight status, but may also be an unmanned aerial vehicle without a manager or an autonomous unmanned aerial vehicle that autonomously flies within a pre-set area. In the case of an unmanned air vehicle, a battery for power supply, a flight drive unit, a GPS unit that provides current location information, a photographing means for photographing spectral and RGB image information, a storage means for storing the photographed image information, and communication connection with other devices Controls the driving of the flight drive unit to move according to the coordinates transmitted from the satellite or other devices through a communication means capable of exchanging signals and information, and an automatic navigation program, and collects the hyperspectral image information or RGB image information It may have a configuration including a control means for controlling the operation of the photographing means, storing the hyperspectral image information or RGB image information and analysis information in the storage unit, and controlling the operation of the communication unit to transmit them to other devices.

In the present invention, the geographic information and image information collection unit 110 can be operated by being equipped with a unique <hyperspectral image analysis program (algorithm)>. 5 is a schematic block diagram of the configuration of the <hyperspectral image analysis program> of the topographic information and image information collection unit. As illustrated in FIG. 5, the <hyperspectral image analysis program> of the topographic information and image information collection unit 110 includes a ground control point measurement module for measuring ground control points set for accuracy of observation coordinates in the shooting area, and a captured second. A data processing module that performs image matching processing on a spectral image or RGB image according to a ground control point for coordinate accuracy of the shooting area, a data analysis and mapping module that analyzes the data processed by the data processing module and maps it to a map, and the photographing A mapping module that maps the topography, water level, flow rate, temperature and humidity of the shooting area so that condition profiling by depth as well as longitudinal and lateral directions of the area is possible, and the data analysis and mapping of the area mapped by the mapping module It may have a configuration including an event output module that outputs an event generation signal according to the occurrence of an abnormality or change so that an RGB image is captured.

The hyperspectral sensor used as the topographic information and image information collection unit 110 is preferably a near-infrared (NIR) image sensor, an optical multispectral sensor, or a hyperspectral image sensor. In the sky, near-infrared (NIR) image sensors have spectral reflectivity in the 600nm-900nm wavelength range or red wavelength range and NIR wavelength range, multispectral image sensors in the 400nm-500nm wavelength range and 600nm-700nm wavelength range or blue and red wavelength ranges, and The hyperspectral image sensor preferably takes a spectroscopic image to measure spectral reflectance in a wavelength range of 400 nm to 600 nm and 650 nm to 750 nm. For reference, the hyperspectral image sensor is a digital camera installed in an unmanned aerial vehicle flying at an altitude of 200 m above the shooting area, for example, captures RGB images, and a near-infrared (NIR) image sensor captures a wavelength range of 600 nm to 900 nm or Red Spectral reflectance of the wavelength band and the NIR wavelength band, the multispectral image sensor has a 400nm to 500nm wavelength band and a 600nm to 700nm wavelength band or a blue and red wavelength band, and the hyperspectral image sensor has a 400nm to 600nm and 650nm to 750nm wavelength band While taking a spectroscopic image to measure the spectral reflectance of the wavelength band, a general RGB image is taken and 3D virtual space information is collected through topographical survey. In hyperspectral imaging, since up to 1,000 narrow spectral bands can be captured, the reflection characteristics of an object can be recorded as it is to distinguish terrain and features. detect radiant energy.

Every imaging system has a measured wavelength range depending on the filter and sensor configuration. When only one type of wavelength band is used, it is called panchromatic, and in the case of a commonly used RGB image, it can be said to be a kind of multispectral imaging in which the sensor responds to three wavelength bands, R, G, and B. Since RGB is recognized by the human eye, RGB images are used a lot, but in reality, various types of configurations are possible. At this time, hyperspectral imaging is obtained by adding spatial imaging to a number of wavelengths as high as those used in general spectrometers. Each pixel has a spectrum as if taken with a spectrometer. When the wavelengths to be photographed are tens to hundreds and the relative spectral resolution is about 0.01, it is called hyperspectral, and the relative spectral resolution is called absolute spectral resolution/length of the spectrum. It is the value obtained by dividing the spectral resolution by the total wavelength range being photographed. Just as a general RGB image is in the form of a cube in which three images of red, blue, and green are stacked, in the case of hyperspectral image, a data cube in the form of stacking images of each wavelength to be captured is formed. There are several methods for capturing spectroscopic images, and measurement devices are classified accordingly. In the case of multi-spectral devices that take pictures by attaching filters according to wavelengths to the sensor like a general RGB image, the image is taken at once as a snapshot, and the data obtained varies depending on how the filter of the sensor is configured. Compared to the Pushbroom method that performs line scan or the Whisk-broom method that is point scan, it is much simpler as it is a snapshot method and has the advantage that all data are captured at the same time. less than the method.

On the other hand, in the case of the scanning method, as the name suggests, scanning is performed, so simultaneous shooting is not possible. Therefore, the device should not shake or the subject should not move. Pushbroom is the so-called line scanning method and is the most commonly used hyperspectral imaging method. At the time of one shot, light in a linear direction enters through the slit, and the light passes through a spectrometer and spreads for each wavelength band, becoming a plane and detected by a 2D detector. 2d images in the spatial and spectral directions are taken. Then, while the camera or object moves in the y direction, line scanning is performed to construct the entire 3D cube.

6 is a schematic configuration diagram showing an example of capturing an actual image and extracting an image and terrain data from it in the present invention. The present invention includes an environmental information collection unit 120, which collects "actual image information" using various types of cameras installed at multiple points along the river in the monitoring target area, By physical sensing, "environmental information" including the water level and velocity of the river, and the temperature and humidity of the river basin is collected. To this end, the environmental information collection unit 120 is installed on the ground along the river, including the embankment, and captures video and still images to acquire actual images, a plurality of CCTV modules, a water level sensor installed on the water surface of the river to measure the water level, A temperature sensor installed on the ground and on the water along the river including the embankment to measure land and water temperature, a humidity sensor installed on the ground along the river including the embankment to measure the humidity around the river, and the measured or measured information over time It may have a configuration including a data logger that stores and manages each data logger, and a network sensor that provides a communication path for collecting and transmitting the measured or measured information by wire or wirelessly.

In the river use safety management system according to the present invention, the information and data collected by the river information collection device 100 are transmitted to the artificial intelligence analysis device 200 in real time through a communication network, and the artificial intelligence analysis device 200 Analyzing transmitted information and data, accumulating analysis information, and performing continuous analysis to analyze whether a disaster including levee collapse or flood occurred, whether an inappropriate person or vehicle entered or drifted in a river, etc., while analyzed forecast information The grafted “topography-image information and environment information” is implemented as a 3D virtual space image, stored and transmitted to the remote management device 300 at the same time.

7 is a schematic block diagram of the configuration of the artificial intelligence analysis device 200 is shown. As illustrated in FIG. 7 , the artificial intelligence analysis device 200 in the present invention may include an analysis and 3D imaging unit 210 and an analysis information management unit 220 . The analysis and 3D imaging unit 210, when the information data collected by the river information collection device 100 is transmitted through the remote management device 300, whether a disaster including an embankment collapse or a flood has occurred, and an inappropriate person in the river It is equipped with an <information analysis program (algorithm)> that analyzes in real time whether a vehicle has been acquired or drifted, and the possibility of a disaster, while accumulating the analyzed data, continuously performing learning analysis, and generating predictive information by informatizing it. . In addition, the analysis and 3D imaging unit 210 is equipped with a <3D virtual space program (algorithm)> that implements a 3D virtual space image by combining terrain-image information, actual image information, environment information, and analyzed prediction information. .

The analysis information management unit 220 of the artificial intelligence analysis device 200 stores data predicted and generated by <information analysis program> and <3D virtual space program> of the analysis and 3D imaging unit 210, and at the same time, the remote management device It performs a function of transmitting to (300).

8 is a schematic block diagram showing the configuration of <information analysis program> provided in the analysis and 3D imaging unit 210 of the artificial intelligence analysis device 200 is shown. As illustrated in FIG. 8, the <information analysis program> provided in the analysis and 3D imaging unit 210 collects hyperspectral image information or RGB image information and analysis information transmitted from the topographic information and image information collection unit 110. A data integration module that converts images and text into images and text, outputs and stores them to the monitor unit, and finally analyzes the integrated analysis data created by the data integration module, the topography, water level, flow rate, temperature, and humidity of the shooting area. Determines whether a disaster including a bank collapse or flood has occurred and whether an inappropriate person or vehicle is entering or drifting in the river, predicts whether a disaster is likely to occur, outputs it to the monitor unit and stores it It may have a configuration including a change occurrence analysis module and an environment evaluation module that inputs final analysis information according to a format set in advance, processes it into environmental evaluation data for the shooting area, and provides it.

Meanwhile, <3D virtual space program> of the analysis and 3D imaging unit 210 implements terrain-image information, actual image information and environment information, and analyzed prediction information as 3D virtual space images. 9 is a schematic block diagram showing the configuration of <3D virtual space program> provided in the analysis and 3D imaging unit 210 . As illustrated in FIG. 9, the <3D virtual space program> includes a virtual space information management part and an image control part. The virtual space information management part of <3D virtual space program> is composed of a collection information DB module, an image collection module, an image matching processing module, and a monitoring module. In the collection information DB module, terrain-image information from the terrain information and image information collection unit 110, actual image information from the environment information collection unit 120, reference images and video images, and SIFT descriptor data is stored That is, the collection information DB module stores topographic-image information of the topographic information and image information collection unit 110, reference image and video image of the environment information collection unit 120, and SIFT descriptor data. The image collection module collects reference images of the space according to the shooting angles for each camera position of the topographic information and image information collection unit 110 and the environment information collection unit 120, The shooting point is set according to the reference image and stored in the collection information DB module. The image matching processing module analyzes the video image, sets SIFT descriptor data for each section in the video image according to the SIFT algorithm, stores it in the collection information DB module, and cross section of adjacent video images based on the SIFT descriptor data By matching with , a virtual space image is created and stored in the collection information DB module. The monitoring module performs a function of monitoring the processing status of the image matching processing module.

The image control part of the <3D virtual space program> includes a search module for searching the collection information DB module for necessary information, a data processing module for controlling the execution of the image collection module, image matching processing module, and monitoring module; It has a configuration including a spatial image editing module that executes a dashboard for outputting the control status of the data processing module and inputting control information.

In the present invention, in the <3D virtual space program> of the analysis and 3D imaging unit 210, in addition to the virtual space information management part and the image control part, GPS location information and INS information included in the captured hyperspectral image or RGB image information An orthoimage construction part may be further provided to construct an orthoimage by generating a masking file by separately extracting only the corresponding region by combining the images by the luminous flux adjustment method after calculating the shooting point inversely from . Furthermore, in the <3D virtual space program>, the captured image information is stored for each band, B1 (Red), B2 (Green), B3 (Blue), B4 (Nir, near infrared), B5 (multispectral, optical Multispectral), B6 (Hyperspectral, Hyperspectral) bands are calculated by the band combination formula, selecting representative pixels for each cluster, including people and vehicles, and different reflectance values and sizes (reflection area) in the near-infrared and hyperspectral bands The pixel (s) included in a window of a predetermined size is selected as a representative pixel centering on the pixel representing A <river object recognition program (algorithm)> is further installed to generate one learning model by combining both and sizes, and to calculate and determine for each pixel the presence or absence of a person or vehicle at a target point based on the learning model. is also desirable

10 is a schematic block diagram showing the configuration of the remote management device 300 provided in the river use safety management system according to the present invention. As illustrated in FIG. 10 , the remote management device 300 includes an “information processing and forecast management unit”. In the "information processing and forecast management unit", if it is determined that there is an abnormality or an abnormal change occurs according to the analysis result of the artificial intelligence analysis device 200, a 3D virtual space image and an abnormal change alarm are sent to the personal communication terminal of the local resident through a communication network (400) and other people's personal communication terminal (500), while transmitting and guiding, in the case of an emergency such as inappropriate person or vehicle entering or drifting in the river, through an output device adjacent to the location. An algorithm or program that transmits an announcement and transmits an emergency along with video information showing an emergency situation to the emergency response device 600 of the emergency response department including the 119 safety center is loaded.

Specifically, in the "information processing and forecast management unit", in relation to the personal communication terminal 400 of local residents, personal information collection, confirmation, and approval are performed in advance to make a forecast for the use of forecast service <local forecast service program (algorithm)> may be installed, and furthermore, when a personal communication terminal 500 of another person is captured in the region through a communication network, personal information collection, confirmation, and approval operation when application for use is made after checking whether or not the forecast service is used. A <foreigner forecast service program (algorithm)> that performs forecasting by performing can be additionally installed in the "information processing and forecast management unit".

On the other hand, the remote management device 300 is a remote device that remotely manages related devices, specifically, the terrain information and image information collection unit 110 and the environment information collection unit constituting the river information collection device 100 ( 120) and an operation communication unit that communicates and connects the artificial intelligence analysis device 200. In addition, the remote management device 300 includes a monitoring unit that outputs information transmitted from the artificial intelligence analysis device 200 and finally analyzed information, and information transmitted from the artificial intelligence analysis device 200 and finally analyzed information. The topographical information and image information collection unit 110, which calculates and stores coordinate values for the shooting area based on an operation storage unit that stores information, and GPS information, and transmits them to the topography and image information collection unit 110. ) Remotely controls the movement and operation, outputs image information, environmental information, analyzed prediction information, and 3D virtual space images to the monitor unit and stores them in the operation storage unit, while further comprising an operation control unit that controls the operation of the operation communication unit. can have

The local residents' personal communication terminal 400 may be a smart phone or PC loaded with a program including an application for activating forecast information transmitted from the remote management device 300 and information on embankments and rivers including 3D virtual space. In addition, the personal communication terminal 500 of a stranger may be a smartphone loaded with a program including an application for activating forecast information transmitted from the remote management device 300 and information on embankments and rivers including 3D virtual space. The emergency response device 600 is installed in competent government offices such as each emergency safety situation center, 119 safety report center, maritime emergency report center, police report center, and local government disaster response center of the Ministry of Public Administration and Security, and the remote management device 300 It may be an emergency response computer device loaded with an algorithm or program that instructs an emergency response team to mobilize to the scene when an emergency state is transmitted along with video information about an emergency situation from the .

In the present invention, the communication network includes various online data including LAN (Local Area Network), MAN (Metropolitan Area Network), WAN (Wide Area Network), Internet communication network, WiFi (Wireless Fidelityv: Wi-Fi) It is connected to a communication network and GPS (Global Positioning System) to enable bi-directional communication of collected information and control signals between each device 100 (300) and the terminal 400 (500).

The remote management device 300 controls the operation of the topographic information and image information collection unit 110. For example, shooting and angle change or rotation of a photographing unit fixed at a corresponding location, or an unmanned aerial vehicle operating in a corresponding area. Set the flight route or remotely control it in real time to make the movement. At this time, if the device for collecting the hyperspectral image and RGB image of the topographic information and image information collection unit 110 is an unmanned aerial vehicle, since an automatic navigation device is installed to enable autonomous driving, GPS coordinate points for the shooting area are determined in advance. When is input, hyperspectral images and RGB images are captured and acquired while moving according to the set coordinates of the embankment along a preset navigation route. In addition, the topographic information and image information collecting unit 110 moves along a pre-set navigation route, but detects a location through a GPS signal and automatically navigates to correct data for rolling and pitching of the topographic information collecting device 110. simultaneously record the attitude value and save their flight path. In this state, the topographic information and image information collection unit 110 captures a hyperspectral image with a hyperspectral sensor such as a near-infrared image sensor, a multispectral image sensor, or a hyperspectral image sensor, and images of previously set spectral or RGB image information. Depending on the degree of (color) gradient, RGB image information is additionally captured and analyzed in close proximity to a place where an event has occurred, that is, a flood has occurred or a flood is suspected.

As described above, in the present invention, as a result of analysis of the spectroscopic image taken by the hyperspectral image sensor of the topographical information and image information collection unit 110, a rapid increase in the water level or flow velocity or a suspected flood occurs in a river including an embankment. When the flight object temporarily departs from the preset flight path, moves to the corresponding coordinates, takes an RGB image, transmits it to the remote management device 300, and analyzes it by the artificial intelligence analysis device 200, so that the analysis algorithm Rather than relying solely on software or programs, administrators can further improve the accuracy of analysis by adding intuitive evaluation analysis through photo verification.

In the present invention, the spectroscopy image and the RGB image that are transmitted by the hyperspectral sensor of the photographing means including the unmanned aerial vehicle flying along a pre-set flight route or the various types of cameras fixed in the sky are captured by the hyperspectral image analysis program. analyze this According to the analysis results, if the water level or flow rate of the river increases rapidly or flooding is suspected in the river area to be monitored, temporarily deviate from the set navigation route and move to the corresponding coordinates to take RGB images and remotely manage the spectroscopic and RGB images. When transmitted to the device 300, the artificial intelligence analysis device 200 collects the transmitted information and approaches or includes the water level and flow speed in the flood range when flooding is expected in the embankment, which is the shooting area, through the self-loaded information analysis program. data, and analyze the cross section of the river including the embankment. Therefore, in the present invention, it is possible to improve the precision or accuracy of information analysis on the embankment, which is a shooting area, and it is possible to provide the analyzed data as evaluation data for the possibility of disasters such as river environmental evaluation and flood occurrence.

In addition, in the present invention, multi-layered information collection and analysis are made based on hyperspectral image information by utilizing ultraviolet and infrared characteristics, which cannot be confirmed by general RGB images, that is, characteristics that cannot be seen by the human eye. By making it possible to collect multi-dimensional and three-dimensional information and analyze the condition of rivers, including embankments, through this, a wide range of information analysis on monitoring targets including embankments, rivers, and river basins is possible. In particular, in the present invention, information is collected and analyzed based on hyperspectral image information that utilizes ultraviolet and infrared characteristics, not the naked eye, so that the manager directly checks the water level, flow rate, temperature, humidity, etc. of the river without visually checking the embankment. it is possible to analyze

Furthermore, according to the present invention, through the precision or accuracy of the analysis of the state information of the monitoring target, including the shooting area, the embankment and the river, and the additional analysis scalability, the environmental investigation of the monitoring target can be quantitatively performed, and accordingly The advantage is that it can be additionally evaluated for the possibility of disasters such as earthquakes and floods as well as the assessment of the embankment environment in the future.

In particular, the present invention collects image information using various types of cameras for monitoring objects, collects environmental information including water level, flow rate, temperature, and humidity through physical sensing, and collects and analyzes hyperspectral image information from the sky. and 3D virtual reality is applied to determine whether an inappropriate person or vehicle is entering or drifting in the river. To this end, in the present invention, the 3D virtual space algorithm or program inversely calculates the shooting point from the GPS location information and INS information included in the photographed hyperspectral image or RGB image information, and then joins the images by the luminous flux adjustment method. A masking file is created by extracting only the corresponding region separately to build an ortho image, and the captured image information is stored for each band, B1 (Red), B2 (Green), B3 (Blue), B4 (Nir, near infrared) , B5 (optical multispectral), and B6 (hyperspectral, hyperspectral) bands are calculated by the band combination formula, and representative pixels for each cluster including people and vehicles are selected, and different Centering on the pixel representing the reflectance value and size, pixels included in a window of a predetermined size are selected as representative pixels, and the reflectance value and size of each material for each of the two bands are combined in each band, and the reflectivity of each of the two bands combined A single learning model is created by combining all values and sizes, and based on the learning model, whether a person or a vehicle exists at a target point is calculated and determined for each pixel. Through this process, if it is judged that a person or vehicle is inappropriately entering the river, an announcement broadcast is sent out through an output device adjacent to the location to block the entry. If it is determined to do so, the emergency situation is transmitted along with the video information to the emergency response device 600 of the emergency response department including the 119 safety center so that emergency response and rescue are performed.

As described above, in the present invention, moving images and still images for the monitoring target are taken through the CCTV module constituting the environmental information collection unit 120, the water level of the river is measured through the water level sensor, and the monitoring target is monitored through the temperature sensor. , Remote management device (300 ) to be analyzed by the artificial intelligence analysis device 200.

The artificial intelligence analysis device 200 collects the data captured or measured by the geographic information and image information collection unit 110 and the environmental information collection unit 120, corrects the deviation, and outputs it as integrated analysis data, or sets the deviation If the range is exceeded, it has a control function to manage re-shooting and measurement for the location. When the artificial intelligence analysis device 200 collects the data captured or measured by the topographic information and image information collection unit 110 and the environmental information collection unit 120 and corrects the deviation, various types of analysis are performed on the collected photographed or measured data. It is desirable to calibrate the total sum of the values by the average value divided by the number of measurements, and if the set deviation range is exceeded, it means that a change has occurred in the corresponding location of the river, including at least the embankment, for example, a flood (or drought) has occurred. Therefore, by operating the unmanned aerial vehicle of the topographic information and image information collection unit 110 to retake and measure the entire embankment, or move to the corresponding coordinates of the embankment and retake and measure to perform reanalysis, the state of the embankment allows for more accurate analysis.

In the present invention, the collected and analyzed information about whether or not there is an abnormality or an abnormal change in the monitoring target is implemented as a 3D virtual space image (image). To this end, the operation control unit of the remote management device 300 sets a 3D virtual space implementation area. It is to select the scope of EDC virtual space based on the basic information of EDC (Eco Delta City) obtained by collecting data to review the scope of virtual space. 11 is a schematic illustration showing an example of setting an area to implement a 3D virtual space.

In the present invention, by the analysis of the artificial intelligence analysis device 200 and the 3D virtual space program provided in the 3D imaging unit 210, the terrain and image information of the topographic information and image information collection unit 110, and the environmental information collection unit (120) stores the reference image and video image of the environment information and SIFT descriptor data, and stores the geographic information and image information collection unit 110 and the environment information collection unit 120. Space according to the shooting angle for each camera position The reference image of is collected, and the shooting point of the video image collected by re-shooting of the indoor space is set according to the reference image and stored in the collection information DB module. In addition, the video image is analyzed by a 3D virtual space program, SIFT descriptor data is set for each section in the video image according to the SIFT algorithm, and stored in the collection information DB module. Based on matching, a virtual space image is created and stored in the collection information DB module. In addition, the 3D virtual space program monitors the processing status of the image matching processing module, searches the collection information DB module to retrieve necessary information, and executes the dashboard for outputting the control status of the data processing module and inputting control information.

To explain this in more detail, when the topographic information and image information collection unit 110 and the environment information collection unit 120 collect and transmit video images taken according to the corresponding shooting angle, <3D virtual space program> In the image collection part, the transmitted video image is stored in the collection information DB module. The image matching processing module compares the collected image with the standard image. If the image is found to be inconsistent with the reference image, the image matching processing module adjusts the shooting angle information of the inconsistent image to the shooting angle of the reference image. correct Specifically, the collected image and the video image must be taken and collected at the same shooting angle, and furthermore, the collected image and the video image must be taken in the same section. The collected image of the shooting angle and the video image are compared with each other to determine whether there is an error. If the error is confirmed, the shooting angle information of the corresponding video image is adjusted with the shooting angle information of the collected image, and the corrected video image is stored in the collection information DB module. is to save

In addition, the image matching processing module analyzes the video image, sets SIFT descriptor data for each section in the video image according to the SIFT algorithm, stores it in the collection information DB module, and determines the intersection of adjacent video images based on the SIFT descriptor data. By matching, an image of an indoor space is created. Since the cameras of the topographic information and image information collection unit 110 and the environmental information collection unit 120 capture the monitoring target, that is, the river including the embankment, at various shooting angles, the image matching processing module has a number of It is necessary to select only a specific section from a video image and perform processing to connect neighboring video images. To this end, the image matching processing module analyzes the video images linked to the corresponding shooting angles, finds sections serving as references from neighboring shooting angles, connects them, and performs matching. For this matching, a specific target that can be a matching criterion is required for video images, and the image matching processing module checks the image of the specific target, matches the video images with each other, and edits them.

This image matching processing module analyzes video images based on the SIFT algorithm to classify and set specific targets. SIFT of SIFT algorithm stands for Scale-Invariant Feature Transform, and is a basic solution for matching multiple video images to each other. The image matching processing module determines based on the location pixel value and distance measurement information of feature points through 3D interpolation, which is an extreme localization process, in a section that can be targeted according to the SIFT algorithm.

The image matching processing module finds a locally swollen part in the image of the object (Scale-space extrema detection). That is, it is to find a characteristic part (a part that appears prominently) in the video image, and the found part is classified as a feature point of the object. In addition, the image matching processing module selects a key point, which is a reliable final feature point, from among previously identified feature points. To this end, the image matching processing module first selects the final feature point based on the pixel value (intensity) of the corresponding feature point and the position or size corresponding to the corner of the object among the feature points, and sets it as a key point of the corresponding video image. Then, the image matching processing module obtains the gradient of the area around the key point, and obtains the directions indicated by the pixel values of the area around the key point as a whole. That is, since the feature point set as the key point corresponds to a corner of the corresponding object image, one or more side images are included around the feature point, and pixel values of the side image are formed toward pixels of the corresponding key point.

Thereafter, the image matching processing module rotates the direction of the constituent pixels of the peripheral area to be 0 degrees, sets the portion corresponding to the peripheral area as SIFT descriptor data, and stores it in the collection information DB module. Through this, one or more SIFT descriptor data is configured in the video image. Therefore, even if there is a change in the video image according to the shooting angle, the video matching processing module recognizes the surrounding area set as the SIFT descriptor data as the same section. For reference, the SIFT descriptor data includes pixel values of a region surrounding a key point, and histogram information based on direction information of pixel values of a region surrounding the key point. Therefore, the SIFT descriptor data enables the image matching processing module to identify targets representing the same section in the image changed according to the photographing angle and accurately match adjacent image images.

In general, the recommended histogram dimension in the SIFT algorithm is 128-dimension, so regardless of the shooting angle of the camera, the video image containing the same section is accurately extracted by the image matching processing module and shared in neighboring video images. A video image is matched based on the SIFT descriptor data, and other necessary sections are combined with each other to generate a river image including an embankment. An example of the generated image is illustrated in FIG. 12 .

In the present invention, existing template data of a 3D virtual space image to be completed may be utilized in the process of performing video image matching by the image matching processing module. To this end, the virtual space information management part included in the <3D virtual space program> of the analysis and 3D imaging unit 210 may further include an existing information DB for storing existing video images and template data for the same indoor space. The image matching processing module searches the existing information DB for the existing image of the river including the embankment under the control of the search part, and performs matching processing of the new image image according to the SIFT algorithm based on the template data do. That is, if there is an image of a river including the same embankment in the past, the image matching processing module searches the existing template data of the spatial image in the existing information DB, and based on the template data, the combination of the image images is more easily performed. can do.

The image matching processing module formats the spatial image generated according to the above process and stores it in the collection information DB module. The formatting converts the spatial images completed by matching all the video images to each other and converts them into a designated format, through which The operator can search and output the spatial image to be output with the video control part.

In the present invention, the monitoring target, that is, the topographic information and image information collection unit 110 that collects hyperspectral image information and RGB image information at specific points at intervals along the river including the embankment, and the water level, flow rate, temperature, and humidity By installing an environmental information collection unit 120 that collects environmental information including water level, flow rate, temperature and humidity, collecting image information using various types of cameras for monitoring targets, and collecting environmental information including water level, flow rate, temperature and humidity through physical sensing On the other hand, hyperspectral image information is collected and analyzed from the sky, and 3D virtual reality is applied to analyze information on river conditions, predict levee collapse and flood possibility, and provide intuitive information on this and whether or not there is an emergency or danger situation in the corresponding river. The remote management device 300 of the management body monitors and analyzes, and provides information to the personal communication terminal 400 of local residents or / and the personal communication terminal 500 of other people temporarily passing through or staying in the area, including the embankment. It is possible to enable nearby civilians to quickly respond to a disaster situation on a river.

In addition, in the present invention, when a stranger temporarily stays in a river basin including an embankment, for example, camping, accommodation using a car, camping, etc., the remote management device 300 of the subject managing the river including the embankment By accessing, agreeing to the use of advance forecast service and utilization of personal information, verifying identity and subscribing to the service, and receiving environmental information about the area in real time, other people temporarily staying or passing through the area It is possible to quickly respond to disaster situations of

In addition, in the present invention, image information is collected using various types of cameras for monitoring objects, and environmental information including water level, flow rate, temperature, and humidity is collected through physical sensing, while hyperspectral image information is collected and analyzed from above. and incorporating 3D virtual reality, if it is determined that an inappropriate person or vehicle is entering or drifting in the river, an announcement broadcast is sent through an output device adjacent to the location, and the emergency response device of the emergency response department including the 119 safety center It is possible to transmit an emergency along with video information to 600 so that emergency response and rescue can be performed.

Although the present invention has been described and illustrated in relation to preferred embodiments for illustrating the principles of the present invention, the present invention is not limited to the configuration and operation as shown and described. It will be appreciated by those skilled in the art that many other changes and modifications may be made to the present invention without departing from the spirit and scope of the appended claims. Accordingly, all such appropriate changes and modifications and equivalents should be regarded as falling within the scope of the present invention.

100: river information providing device
110: topographic information and image information collection unit
120: environmental information collection unit
200: artificial intelligence analysis device
300: remote management device
400: personal communication terminal for local residents
500: personal communication terminal of stranger
600: emergency response device

Claims (6)

For the river area to be monitored, topography-image information including hyperspectral image information and RGB image information is collected, and actual image information at multiple points on the ground or above water in the river area to be monitored, and the environment including water level, flow rate, temperature and humidity River information collection device for collecting information in real time;
Continuous analysis is performed while accumulating information and data collected and transmitted by the river information collection device to determine whether a disaster including bank collapse or flood occurs in the river area to be monitored, and whether inappropriate people or vehicles in the river are entering or entering the river. Artificial intelligence that analyzes and determines whether drifting occurs, predicts the possibility of a disaster, implements and stores terrain-image information, actual image information and environmental information, and analyzed prediction information as 3D virtual space images and transmits them to a remote management device intelligence analysis device; and
The information and data transmitted from the river information collection device are transmitted to the artificial intelligence analysis device. According to the analysis results of the artificial intelligence analysis device, if a disaster is judged to occur or a disaster is predicted, local residents who have subscribed to the forecast service in advance 3D virtual space images and warnings are transmitted and guided to personal communication terminals and other personal communication terminals, and when it is determined that an inappropriate person or vehicle is entering or drifting in the river, an announcement is sent through an output device adjacent to the location, A river use safety management system using artificial intelligence, characterized in that it has a configuration including a remote management device that transmits an emergency situation along with video information as an emergency response device of the emergency response department. According to claim 1,
The artificial intelligence analysis device is composed of an analysis and 3D imaging unit and an analysis information management unit;
The analysis and 3D imaging department analyzes in real time whether a disaster has occurred, whether an inappropriate person or vehicle has entered or drifted in the river, and the possibility of a disaster when the information data collected by the river information collection device is transmitted, and analyzes the analyzed data in real time. 3D virtual space image that implements 3D virtual space image by combining information analysis program that continuously performs learning analysis and informatization to generate prediction information while accumulating information, terrain-image information, actual image information, environment information, and analyzed prediction information contains a spatial program;
The analysis information management unit stores the information and data predicted and generated by the information analysis program and the 3D virtual space program and transmits them to the remote management device. According to claim 2,
The 3D virtual space program calculates the shooting point inversely from the GPS location information and INS information included in the photographed hyperspectral image or RGB image information, and then joins the images by the luminous flux adjustment method, extracts only the corresponding area separately, and creates a masking file. an orthoimage building module for generating and constructing an orthoimage;
In the 3D virtual space program, the captured image information is stored for each band, but representative pixels for each cluster including people and vehicles are calculated by the band combination formula for the B1, B2, B3, B4, B5, and B6 bands. Select pixels included in a window of a preset size centered on pixels showing different reflectance values and sizes in the near-infrared and hyperspectral bands as representative pixels, and select the reflectance value of each material for each band in each band Combining the sizes, combining the reflectivity values and sizes of the two combined bands to create one learning model, and calculating and determining the presence of people or vehicles at the target point by pixel based on the learning model. River use safety management system using artificial intelligence, characterized in that the "object recognition program in the river" is mounted. According to any one of claims 1 to 3,
remote management device,
As a result of the analysis of the artificial intelligence analysis device, if it is determined that there is an abnormality or an abnormal change has occurred, the 3D virtual space image and abnormal change alert are transmitted to the personal communication terminals of local residents and other people through the communication network, and guides the inappropriate person in the river Or, if it is determined that the vehicle has been acquired or drifted, an algorithm or program is installed that transmits an announcement broadcast through an output device adjacent to the location and transmits the emergency state along with emergency video information to the emergency response device of the emergency response department. River use safety management system using artificial intelligence, characterized in that the information processing and forecast management unit is provided. According to claim 4,
In the information processing and forecast management unit,
A local forecast service program that collects, confirms, and approves personal information in advance for the use of local residents' personal communication terminals and forecast services so that forecasts are made; and
When a personal communication terminal of another person is captured in the region through a communication network, a forecast service program of a third party that collects, confirms, and approves personal information is performed to make a forecast in case of application for use after checking whether or not the forecast service is used. A river use safety management system using artificial intelligence, characterized in that it is mounted. Through the river use safety management system equipped with a river information collection device, artificial intelligence analysis device, and remote management device, terrain-image information, environmental information, and actual image information for the river area to be monitored are collected, analyzed, and provided to use the river. As a method for safely managing
Through the river information collection device, topographical image information including hyperspectral image information and RGB image information is collected for the river including the embankment in the measurement target area, and actual image information is obtained from multiple points on the ground or on the water surface of the river including the corresponding embankment. , and collecting environmental information including water level, flow rate, temperature and humidity in real time;
Through the artificial intelligence analysis device, based on the information and data collected and transmitted by the river information collection device, whether a disaster including bank collapse or flood occurs in the river area to be monitored, and whether inappropriate people or vehicles in the river It analyzes and determines whether it is drifting or drifting, predicts the possibility of disaster, implements and stores topographic-image information, actual image information and environmental information, and analyzed prediction information as 3D virtual space images and transmits them to remote management devices. step; and
According to the analysis result, when a disaster is judged to occur or a disaster is predicted, 3D virtual space images and alerts are transmitted through a remote management device to local residents' personal communication terminals subscribed to the forecast service in advance and to other people's personal communication terminals. When it is judged that an inappropriate person or vehicle is entering or drifting in the river, an announcement is sent through an output device adjacent to the location, and an emergency situation is transmitted along with video information to the emergency response device of the emergency response department. River use safety management method comprising the step.

Images