KR102306854B1 - System and method for managing traffic event - Google Patents

Abstract

The present invention provides a traffic situation management system and method. Provided is a method for learning various event situations on the road of the vehicle with artificial intelligence, monitoring the traffic situation in real time through CCTV, and providing a zoom-in image to the manager by judging the situation with artificial intelligence when a specific event occurs.

Description

Traffic situation management system and method {System and method for managing traffic event}

The present invention is a traffic situation management system and method, and more specifically, it learns various event situations on the vehicle road with artificial intelligence, monitors the traffic situation in real time through CCTV, and determines the corresponding situation with artificial intelligence when a specific event occurs to zoom in. -In) It is about the method of providing one video to the manager.

With the recent development of information and communication technology, the spread of telematics devices and the establishment of a ubiquitous environment, an Intelligent Transport System (ITS) is being applied.

In particular, on highways and automobile-only roads, CCTVs are installed in each road section, and various situations occurring on the road are monitored and dealt with through an image monitoring system.

Through this video monitoring system, when an unexpected situation occurs, the manager at a remote location detects the occurrence of an event based on the video screen for the relevant road section and takes follow-up measures accordingly. There are limits.

Furthermore, there is a problem that it takes time for the manager to determine what kind of situation has occurred through the video screen after the event occurs. is occurring frequently.

Therefore, it is necessary to find a way to automatically recognize various unexpected situations occurring on the vehicle road and actively perform appropriate follow-up measures in real time.

Korean Patent Registration No. 10-2061264 Korean Patent Registration No. 10-1931298

The present invention has been devised in order to solve the problems of the prior art as described above, so that it is possible to recognize various events occurring on the vehicle road, automatically grasp the occurrence situation, and accurately grasp the corresponding road section together with the event information. We would like to suggest a way to provide zoom-in images to administrators.

It is intended to solve the problem that the administrator cannot monitor the numerous video screens transmitted from the vehicle road in real time by focusing.

In particular, as it takes time for the manager to determine the situation through the video screen after the event occurs, appropriate measures cannot be taken for vehicles moving at high speed in the section where the situation occurs, thereby solving the problem that leads to a major traffic accident.

The object of the present invention is not limited to the above, and other objects and advantages of the present invention not mentioned can be understood by the following description.

An embodiment of the method for managing traffic conditions according to the present invention includes: a road condition monitoring step of monitoring the vehicle road condition by photographing it in real time through a camera; an event detection step of detecting an event occurrence by checking an object or movement pattern on the vehicle road; a traffic condition determination step of analyzing the corresponding event to determine the traffic condition; and providing traffic condition information providing an image in which the event occurrence area is zoomed-in.

Preferably, the method further comprises an artificial intelligence learning step of retaining object definition information about an object, object movement pattern information, and event situation information corresponding to the object definition information and the object movement pattern information through machine learning, wherein the event detection The step may include: an object determination step of detecting an object existing on a real-time image screen of a vehicle road and determining the detected object based on the object definition information; a movement pattern determination step of determining a movement pattern of the object based on the object movement pattern information; and an event generation detection step of detecting occurrence of an event based on the object determination result and the object movement pattern determination result, wherein the traffic condition determination step includes a determination result of the object and the object based on the event condition information By analyzing the event situation corresponding to the movement pattern determination result of , the zoom-in image and the traffic condition information may be provided to a manager terminal or a driver terminal.

Further, the object determination step may include extracting a feature point and a feature line from the image screen, detecting an object based on the feature point and the feature line, and comparing the detected feature point and feature line with the object definition information to determine the object and, the moving pattern determination step sets a mesh point based on a lane on the image screen, warps the mesh point to generate a warping image, and extracts the movement pattern of the object sensed from the warping image can do.

As an example, the object determination step classifies the object as an unidentified object when determination of the detected object is impossible based on the object definition information, and the traffic condition determination step includes an event for the unidentified object and the movement pattern of the unidentified object Generating the occurrence information, the step of providing the traffic situation information, setting the event occurrence area on the video screen to take a zoom-in, and provide the zoom-in image and the event occurrence information to the manager terminal, through the manager terminal The artificial intelligence learning step may be additionally performed by receiving new event information corresponding to the event occurrence information.

In addition, an embodiment of the traffic situation management system according to the present invention, a plurality of photographing devices for providing a zoom-out or zoom-in image by photographing a vehicle road in real time; A real-time video screen of the vehicle road is provided from the plurality of photographing devices to detect an object existing on the video screen, and an event is determined by analyzing the object and movement pattern based on machine learning information, and an event according to the occurrence of an event a traffic condition management device that provides traffic condition information and a zoom-in image of a corresponding event occurrence area; a manager terminal that receives traffic condition information and a zoom-in image from the traffic condition management device, and receives new event information about an unidentified object and its movement pattern from a manager and provides it to the traffic condition management device; and a driver's terminal mounted on a driving vehicle on a vehicle road or possessed by a driver, and receiving and displaying traffic situation information and a zoom-in image of a corresponding event occurrence area from the traffic situation management device.

According to the present invention as described above, the traffic situation management device detects and analyzes the event situation occurring on the vehicle road in real time and provides the traffic situation information and zoom-in image thereto to the manager, so that the manager can more quickly and accurately control the situation of the vehicle road can recognize

In particular, through the traffic condition management device, the manager can receive real-time traffic condition information and zoom-in images according to the occurrence of events on the vehicle road. Recognize it so you can take follow-up action.

In addition, by providing traffic situation information and a zoom-in image of the event to the driver's terminal of the vehicle driving on the vehicle road in which the event occurred, the vehicle driver can recognize the traffic situation and prevent accidents in advance. If necessary, it is possible to avoid traffic jams by diverting to another road.

Effects of the present invention are not limited to those mentioned above, and other effects not mentioned will be clearly understood by those of ordinary skill in the art from the description below.

1 shows an embodiment of a traffic situation management system according to the present invention.
2 is a block diagram of an embodiment of a traffic condition management apparatus according to the present invention.
3 is a flowchart illustrating an embodiment of a method for managing traffic conditions according to the present invention.
4 is a flowchart illustrating an embodiment of detecting the occurrence of an event and providing traffic situation information thereto in the traffic situation management method according to the present invention.
5 shows an embodiment of determining an event by detecting an object on a vehicle road in the present invention.
6 shows an embodiment of determining the movement pattern of an object by generating a warping image in the present invention.
FIG. 7 shows an embodiment of detecting a region in which an event occurs by tracking a movement pattern of an object in the present invention.
8 shows an embodiment of a zoom-in image for an event occurrence area in the present invention.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings, but the present invention is not limited or limited by the embodiments.

In order to explain the present invention, the operational advantages of the present invention, and the objects achieved by the practice of the present invention, preferred embodiments of the present invention are exemplified below and will be described with reference to them.

First, the terms used in the present application are only used to describe specific embodiments, and are not intended to limit the present invention, and the singular expression may include a plural expression unless the context clearly indicates otherwise. Also in the present application, terms such as “comprise” or “have” are intended to designate that a feature, number, step, operation, component, part, or combination thereof described in the specification exists, but one or more other It is to be understood that this does not preclude the possibility of addition or presence of features or numbers, steps, operations, components, parts, or combinations thereof.

In describing the present invention, if it is determined that a detailed description of a related known configuration or function may obscure the gist of the present invention, the detailed description thereof will be omitted.

The present invention automatically analyzes the traffic situation by monitoring the vehicle road, and provides traffic situation information according to the analysis result to a manager or a driver. Here, the vehicle road may include various vehicle roads to which the present invention can be applied according to necessity, such as a national road or an urban road, in addition to including a highway or an exclusive road for automobiles.

In particular, the present invention learns various event conditions on the vehicle road with artificial intelligence, monitors the traffic condition of the vehicle road through CCTV in real time, and determines the situation with artificial intelligence when a specific event occurs to zoom-in with traffic situation information. A traffic situation management system that provides an image to a manager and a traffic situation management method using it are presented.

1 shows an embodiment of a traffic situation management system according to the present invention.

The traffic situation management system according to the present invention may include a photographing device 100 , a traffic situation management device 200 , and a manager terminal 300 , and may further include a driver terminal 340 according to a communication network construction situation. have.

The photographing apparatus 100 includes a plurality of cameras 110a, 110b, and 110c installed along the vehicle road, photographing the vehicle road, and providing a real-time photographed image to the traffic condition management apparatus 200 .

The number and spacing of the cameras 110a, 110b, and 110c installed on the vehicle road may be variously changed as needed. For example, multiple cameras can be installed at relatively close distances in sections where traffic jams are severe or where traffic accidents occur frequently. A plurality of cameras may be installed to take pictures.

Furthermore, the camera of the photographing apparatus 100 may include an infrared camera to enable photographing of the vehicle road even at night time.

The real-time captured images of the cameras 110a, 110b, and 110c installed for a plurality of vehicle roads are transmitted to the traffic condition management device 200 located at a remote location. of communication means may be provided.

The traffic condition management apparatus 200 receives images of real-time vehicle roads from the photographing apparatus 100 installed on a plurality of vehicle roads.

The traffic situation management device 200 receives a real-time image screen of the vehicle road from a plurality of photographing devices 100 and detects an object existing in the image screen, and the detected object and the object based on machine learning information of artificial intelligence The event is determined by analyzing the movement pattern. In addition, when an event occurs, the traffic condition management apparatus 200 provides event traffic condition information and a zoom-in image of the corresponding event occurrence area to the manager terminal 300 .

In particular, the traffic condition management device 200 controls the photographing device 100 corresponding to the occurrence of an event on the vehicle road to obtain an image captured by zoom-in in the event occurrence area, and displays the zoomed-in image to the manager terminal 300 . By providing this, it supports the manager to judge the vehicle road condition more accurately.

In addition, the traffic condition management device 200 provides the corresponding event occurrence information to the manager terminal 300 for an event situation that cannot be determined by artificial intelligence, and receives new event information analyzed by the manager from the manager terminal 300 to receive new It may be configured so that machine learning of artificial intelligence for event information can be additionally performed.

The manager terminal 300 receives traffic condition information for an event occurring on a specific vehicle road and a zoom-in image corresponding thereto from the traffic condition management device 200 . Based on the traffic situation information and the zoom-in image provided from the traffic situation management device 200, the manager can more quickly and accurately recognize the situation of the corresponding vehicle road, thereby enabling quick follow-up actions.

Furthermore, the driver terminal 450 of the vehicle 400 traveling on the vehicle road may receive various traffic condition information generated on the vehicle road being driven from the traffic condition management device 200 . Here, the driver terminal 450 may be a vehicle terminal such as a navigation device fixedly disposed on the vehicle 400 or a mobile terminal such as a smartphone owned by the driver, and the traffic condition management device 200 may be configured to operate a driver terminal through various wireless communication methods. Data may be sent to 450 .

The driver can recognize the situation of the road ahead or the road of a nearby vehicle by receiving the traffic situation information and zoom-in image on the road condition ahead while the vehicle 400 is running, so that the driver can prevent accidents or delays due to traffic jams in advance. can be prevented.

2 is a block diagram of an embodiment of a traffic condition management apparatus according to the present invention.

The traffic condition management device 200 may include a photographing device control unit 210 , an image analysis unit 220 , an event determination unit 230 , an artificial intelligence learning unit 240 , a traffic information providing unit 250 , and the like. have.

The photographing device control unit 210 receives a real-time photographed image of the vehicle road from the photographing device 100 while controlling the operation of the photographing apparatuses 100 distributed on a plurality of vehicle roads.

In particular, the photographing device control unit 210 may selectively receive a zoom-out image or a zoom-in image depending on the situation while controlling the photographing direction, the photographing angle, etc. with respect to the plurality of photographing apparatuses 100 . can be controlled so that

The image analysis unit 220 processes and analyzes the real-time captured image obtained from the photographing device control unit 210 to identify an object existing on a vehicle road and determine a movement pattern of the object.

As an example, the image analyzer 220 may extract a feature point and a feature line from the image screen and detect an object existing on the vehicle road based on the feature point and the feature line. In addition, the image analysis unit 220 may determine the movement pattern of the detected object based on the real-time continuous image.

As an example, the image analyzer 220 generates a warping image by setting a mesh point based on a lane on the image screen and warping the mesh point. And it is possible to extract the movement pattern of the detected object from the generated warping image.

The event determination unit 230 determines whether the object detected by the image analysis unit 220 is suitable to exist on the vehicle road or is suitable for the moving direction of the vehicle road with respect to the movement pattern of the object identified by the image analysis unit 220 . Determine whether it is a movement pattern.

The event determination unit 230 may determine an object and a movement pattern through artificial intelligence. To this end, the event determination unit 230 may have an artificial intelligence configuration. If necessary, the event determination unit 230 may use an external artificial intelligence system.

The event determination unit 230 may determine what kind of object the corresponding object is by comparing the object definition information and the detected object through machine learning of artificial intelligence, and also object movement pattern information and the corresponding object through machine learning of artificial intelligence. It can also be determined whether the object is properly moving on the vehicle road in preparation for the movement pattern of In addition, the event determination unit 230 may determine what kind of situation the corresponding event is based on the object definition information and the event situation information corresponding to the object movement pattern information through artificial intelligence.

The event determination unit 230 may generate traffic condition information according to the determination result of the artificial intelligence, and may also obtain a zoom-in image of a vehicle road area corresponding to the corresponding traffic condition information by interworking with the photographing device control unit 210 .

The artificial intelligence learning unit 240 provides object definition information for an object on a vehicle road through machine learning, object movement pattern information for a movement pattern of the object, and event situation information corresponding to the object definition information and the object movement pattern information. hold the

To this end, the artificial intelligence learning unit 240 is provided with learning data for premise learning, and based on the learning data, it learns the object definition for the object on the video screen through supervised learning or the movement pattern of the object. An object movement pattern may be learned, and further, an object definition may be established or an object movement pattern may be established through clustering through unsupervised learning. Furthermore, the artificial intelligence learning unit 240 may learn event situation information for a corresponding traffic situation corresponding to the object definition information and the object movement pattern information.

In addition, when there is an unidentified object that cannot be determined through the event determination unit 230 , event occurrence information is provided to the manager terminal 300 , and new event information corresponding thereto is provided from the manager terminal 300 . Continuous machine learning can be made based on new event information.

The traffic information providing unit 250 provides traffic situation information and a zoom-in image corresponding thereto to the manager terminal 300 . To this end, the traffic information providing unit 250 supports data communication with the manager terminal 300 in various wired or wireless communication methods.

In addition, the traffic information providing unit 250 may provide the traffic situation information and a zoom-in image corresponding thereto to the driver terminal 450 of the vehicle driving on the vehicle road on which the corresponding traffic situation information is generated. To this end, the traffic information providing unit 250 supports data communication with the driver terminal 450 through various wireless communication methods.

The present invention proposes a traffic situation management method using the traffic situation management system as described above. Since the traffic situation management method according to the present invention uses the traffic situation management system according to the present invention as described above, hereinafter, a traffic situation according to the present invention is provided. Let's take a look at the traffic situation management method by referring to the management system together.

3 is a flowchart illustrating an embodiment of a method for managing traffic conditions according to the present invention.

According to the present invention, the traffic situation management device 200 acquires an image of the vehicle road in real time through the photographing device 100 installed on the vehicle road, and based on this, determines various event situations occurring on the vehicle road through artificial intelligence, As to provide situational information and images, for this purpose, machine learning of artificial intelligence (S100) is preceded.

Machine learning of artificial intelligence can learn object definition for an object on a video screen or learn an object movement pattern for an object movement pattern through supervised learning based on pre-prepared learning data, and furthermore Object definitions or object movement patterns may be established through clustering through unsupervised learning. In addition, it is possible to learn event situation information about the corresponding traffic situation corresponding to the object definition information and the object movement pattern information.

Through this machine learning of artificial intelligence, object definition information about an object on a vehicle road, object movement pattern information, and event situation information corresponding to the object definition information and the object movement pattern information are retained.

In a state in which artificial intelligence is learned, the traffic condition management apparatus 200 receives a real-time photographed image of the vehicle road from the photographing device 100 and monitors the vehicle road in real time (S200).

The traffic condition management device 200 detects an object on the vehicle road during real-time monitoring, recognizes a movement pattern of the detected object, and detects whether an event occurs through artificial intelligence (S300).

If it is sensed that a specific event has occurred, the traffic condition management apparatus 200 analyzes the event based on the captured image and determines the traffic condition accordingly ( S400 ).

In addition, the traffic condition management apparatus 200 acquires a zoomed-in image by controlling the photographing apparatus 100 to zoom-in photographing (S500) with respect to an event occurrence area corresponding to the determined traffic condition.

In addition, the traffic condition management apparatus 200 provides traffic condition information according to the occurrence of an event and a zoom-in image corresponding thereto (S600). At this time, the traffic condition management device 200 may provide traffic condition information and a zoom-in image to the set manager terminal 300 , and depending on the situation, the traffic condition information and A zoomed-in image may also be provided.

A detailed process of the traffic condition management method according to the present invention will be described in more detail through the embodiments of FIGS. 4 to 8 .

4 is a flowchart for an embodiment of detecting the occurrence of an event and providing traffic situation information thereto in the traffic situation management method according to the present invention, and FIG. 5 is an event by detecting an object on the vehicle road in the present invention. Fig. 6 shows an embodiment of determining the movement pattern of an object by generating a warping image in the present invention, and Fig. 7 shows an event occurring by tracking the movement pattern of the object in the present invention It shows an embodiment of detecting a region that has been changed, and FIG. 8 shows an embodiment of a zoomed-in image for an event occurrence region in the present invention.

The traffic condition management apparatus 200 receives a real-time photographed image of the vehicle road from the photographing apparatus 100, determines whether an object exists on the image screen, and detects the object (S310).

For example, by extracting feature points and feature lines from the image screens 510 and 520 as shown in FIG. 5 , lanes of the vehicle road 10 are distinguished based on the feature lines, and objects 410a, 410b, 410c, and 410d based on the feature points. , 420a, 420b, 420c, 430) can be distinguished.

Also, as an example, in the case of the vehicle road 20 having a turning section as in the video screen 530a shown in FIG. 6A , the mesh point 531a based on the lane 20 in the video screen 530a ) and warping the mesh point 531a, as in the image screen 530b shown in FIG. 6(b), a warping image for the vehicle road 20 based on the mesh point 531b. After generating , the objects 440a, 440b, and 440c may be distinguished based on the feature point and the feature line. When a warping image is applied, the corresponding objects 440a, 440b, and 440c are in a distorted state depending on the degree of movement of the mesh point 531a. show the normal appearance.

When an object is detected on the video screen, the traffic condition management apparatus 200 determines what kind of object the object detected on the video screen is based on the object definition information held through artificial intelligence learning (S321).

If the determined object is an abnormal object that should not exist on the vehicle road ( S323 ), the traffic condition management apparatus 200 determines that an event has occurred. In this case, the object determination may be determined differently by the artificial intelligence according to the characteristics of the vehicle road. As an example, if an object determined to be a person is present on the image screen for a highway or automobile-only road, it may be determined as an abnormal object, but if an object determined to be a person exists on the image screen for an urban vehicle road This may be determined as a normal object.

As a specific embodiment related to object determination, the objects 410a, 410b, 410c, and 410d detected on the vehicle road 10 of the image screen 510 as shown in FIG. When all are identified as vehicles, they are determined as normal objects, but among the objects 420a, 420b, 420c, and 430 detected on the vehicle road 10 of the video screen 520 as shown in FIG. When an object 430 other than (420a, 420b, 420c, 430) is identified, it is determined that an abnormal object exists and this is detected as an event occurrence.

In addition, the traffic condition management device 200 determines whether the detected object moves in an abnormal movement pattern by determining the movement pattern of the object detected on the video screen based on the object movement pattern information retained through artificial intelligence learning (S325). .

For example, the movement pattern of the objects 410a, 410b, 410c, and 410d determined on the vehicle road 10 of the video screen 510 as shown in FIG. It is judged on the basis of whether an abnormal movement pattern exists, and since the vehicles 410c and 410d due to a contact accident on the vehicle road 10 have a movement pattern different from the normal speed and direction of the surrounding vehicles, it is determined as an abnormal movement pattern. Accordingly, the occurrence of an event can be detected.

And, based on the object movement pattern information among the objects 420a, 420b, 420c, 430 determined on the vehicle road 10 of the image screen 520 as shown in FIG. 5(b), the vehicle on the vehicle road 10 Since the object 430 continuously having a movement pattern different from the moving direction appears as an abnormal movement pattern, it can be detected as an event occurrence.

Furthermore, when the vehicle road 20 of the image screen 530a has a rotation section as shown in FIG. By generating the warped image screen 530b as shown in b) and identifying the movement patterns of objects, it is possible to more easily and clearly distinguish a normal movement pattern from an abnormal movement pattern.

If the object detected on the image screen through this process has a normal movement pattern as a normal object, the traffic situation management device 200 continues real-time monitoring (S200) of the vehicle road as a situation in which an event does not occur. will perform

However, when an abnormal object that should not exist on the vehicle road is detected or the movement pattern of the detected object is abnormal, the traffic situation management device 200 determines that an event occurs and analyzes the event based on the object and the movement pattern and accordingly The traffic condition is determined (S411) and traffic condition information is generated (S413).

As an example, based on the object definition information and the event situation information corresponding to the object movement pattern information through artificial intelligence, the event situation for the corresponding object and the corresponding movement pattern is analyzed to determine the traffic situation, and the corresponding traffic situation information is generated. do.

For example, in the image screen 510 as shown in FIG. 5A , the objects 410c and 410d determined to be vehicles based on the object definition information are selected from the surrounding vehicles 410a and 410b based on the object movement pattern information. When a movement different from the driving direction and driving speed appears or exists in a non-ideal contact state, an access accident with respect to the corresponding vehicles 410c and 410d may be determined based on the corresponding event situation information.

In addition, the object 430 detected on the basis of the object definition information on the image screen 520 as shown in FIG. In the case of continuously representing a movement different from the driving direction and driving speed of , the risk of the obstacle 430 may be determined based on the corresponding event situation information.

In addition to the event situation according to the above-described embodiment, the traffic condition management apparatus 200 may analyze various event conditions and determine the traffic conditions accordingly to generate traffic condition information on the corresponding vehicle road.

In addition, the traffic condition management apparatus 200 sets an event generation area based on the movement pattern of the object on the image screen of the vehicle road (S511), and scans the object on the vehicle road along the vehicle driving direction on the image screen to generate the event area can be found, and the event occurrence area is set based on this.

For example, while determining the movement patterns of the vehicles 450a, 450b, and 450c along the vehicle movement direction from the bottom to the top on the image screen 540 as shown in FIG. By scanning, a connection accident point of the vehicles 460a and 460b may be found.

As described above, by scanning the image screen based on the movement pattern of the object on the image screen, an event occurrence region may be found and an event occurrence region may be set.

Then, the traffic condition management device 200 sets the event occurrence area on the video screen in which the event situation occurs (S511), and then sets the event occurrence area in which the photographing device 100 is shooting the vehicle road in which the event situation occurs. A zoom-in image may be acquired by controlling the zoom-in to take a picture ( S515 ).

Furthermore, the traffic condition management apparatus 200 provides the traffic condition information analyzed and determined for the event and a zoomed-in image thereto to the manager terminal 300 ( S610 ).

Since the manager can receive traffic condition information according to the occurrence of an event from the traffic condition management device 200 and a zoomed-in image thereof through the manager terminal 300, the manager can immediately zoom-in image without going through a separate judgment or analysis process. of traffic conditions can be recognized.

Going one step further, the traffic condition management apparatus 200 may provide traffic condition information according to the occurrence of the event and a zoom-in image thereof to the driver terminal 450 of the vehicle driving on the vehicle road in which the corresponding event has occurred. By receiving such traffic situation information and zoom-in image, the vehicle driver can recognize the traffic situation and prevent an accident in advance, and, if necessary, can avoid traffic jam by detouring to another road.

In addition, in the present invention, the traffic condition management device 200 can continuously perform machine learning of artificial intelligence based on new event information. In the determination (S321), if it is impossible to determine the object based on the object definition information, it is classified as an unidentified object (S451) and event occurrence information is generated (S455).

In addition, the traffic condition management apparatus 200 sets an event occurrence area corresponding to an unidentified object on the image screen (S551), controls the corresponding photographing apparatus 100, and receives a zoomed-in image (S555).

And the traffic condition management device 200 provides event occurrence information on the unidentified object and a zoom-in image corresponding thereto to the manager terminal 300 (S650), and the manager selects the unidentified object based on the event occurrence information and the zoom-in image. It is determined and input new event information to the manager terminal 300 (S700). In this case, the new event information may include object definition information and object movement pattern information, and may optionally further include event context information corresponding to the corresponding object definition information and object movement pattern information.

The new event information input to the manager terminal 300 is transmitted to the traffic condition management device 200, and the traffic condition management device 200 additionally performs artificial intelligence machine learning by using the received new event information as learning data. do.

Through this process, the object definition information, object movement pattern information, and event situation information of artificial intelligence can be continuously expanded.

According to the present invention as described above, the traffic situation management device detects and analyzes the event situation occurring on the vehicle road in real time and provides the traffic situation information and zoom-in image for this to the manager, so that the manager can more quickly and accurately control the situation of the vehicle road can recognize

The above description is merely illustrative of the technical spirit of the present invention, and various modifications and variations will be possible without departing from the essential characteristics of the present invention by those skilled in the art to which the present invention pertains. Therefore, the embodiments described in the present invention are not intended to limit the technical spirit of the present invention, but to explain, and the technical spirit of the present invention is not limited by these embodiments. The protection scope of the present invention should be construed by the following claims, and all technical ideas within the scope equivalent thereto should be construed as being included in the scope of the present invention.

100, 110a, 110b, 110c: photographing device;
200: traffic condition management device,
210: photographing device control unit;
220: image analysis unit,
230: event judgment unit,
240: artificial intelligence learning department,
250: traffic information provider,
300: administrator terminal;
400: vehicle,
450: driver terminal.

Claims (5)

a road condition monitoring step of monitoring the vehicle road condition by photographing it in real time through a camera;
an event detection step of detecting an event occurrence by checking an object or movement pattern on the vehicle road;
a traffic situation determination step of analyzing the corresponding event to determine the traffic condition; and
A traffic condition information providing step of providing a zoom-in image of the event occurrence area,
The method further comprises an artificial intelligence learning step of retaining object definition information, object movement pattern information, and event situation information corresponding to the object definition information and the object movement pattern information through machine learning,
The event detection step is
an object determination step of detecting an object existing on a real-time image screen of a vehicle road and determining the detected object based on the object definition information;
a movement pattern determination step of determining a movement pattern of the object based on the object movement pattern information; and
An event occurrence detection step of detecting the occurrence of an event based on the object determination result and the object movement pattern determination result,
The traffic condition determination step is,
Determine the traffic situation by analyzing the event situation corresponding to the determination result of the object and the movement pattern determination result of the object based on the event condition information, and generate traffic condition information corresponding thereto,
The step of providing the traffic condition information,
Setting an event occurrence area on the video screen to zoom in, and providing the zoomed-in image and the traffic situation information to a manager terminal or a driver terminal,
The object determination step is
extracting feature points and feature lines from the image screen, detecting an object based on the feature points and feature lines, and comparing the detected feature points and feature lines with the object definition information to determine an object,
The movement pattern determination step is,
A method for managing a traffic situation, characterized in that a mesh point is set based on a lane on the image screen, a warping image is generated by warping the mesh point, and a movement pattern of the object detected from the warping image is extracted. . delete delete The method of claim 1,
The object determination step is
Classifying the object as an unidentified object when it is impossible to determine the detected object based on the object definition information,
The traffic condition determination step is,
generating event occurrence information for the unidentified object and the movement pattern of the unidentified object,
The step of providing the traffic condition information,
Set an event occurrence area on the video screen to zoom in, and provide the zoom-in image and the event occurrence information to an administrator terminal,
Traffic situation management method, characterized in that the artificial intelligence learning step is additionally performed by receiving new event information corresponding to the event occurrence information through the manager terminal. A plurality of photographing devices for providing a zoom-out or zoom-in image by photographing a vehicle road in real time;
A real-time video screen of the vehicle road is provided from the plurality of photographing devices to detect an object existing on the video screen, and an event is determined by analyzing the object and movement pattern based on machine learning information, and an event according to the occurrence of an event a traffic condition management device that provides traffic condition information and a zoom-in image of a corresponding event occurrence area;
a manager terminal that receives traffic condition information and a zoom-in image from the traffic condition management device, and receives new event information about an unidentified object and its movement pattern from a manager and provides it to the traffic condition management device; and
A traffic situation management system, comprising a driver terminal mounted on a driving vehicle on a vehicle road or held by a driver, and receiving and displaying traffic situation information and a zoom-in image of a corresponding event occurrence area from the traffic situation management device.

Images