KR102381352B1 - A traffic information assistance system based on license plate recognition - Google Patents

Abstract

The present invention relates to a vehicle number recognition-based traffic information support system that collects, aggregates, and records vehicle traffic information so that it can be utilized in traffic, crime prevention, and control system construction. Vehicle number recognition-based traffic information support system according to the present invention extracts a vehicle object and a license plate area from a photographing means for outputting image data by photographing a vehicle moving by car, and the image data received from the photographing means, and uses the extracted license plate as an image An image extraction module that stores as , a text extraction module that extracts at least a vehicle model identification number as text information from the vehicle license plate image extracted from the image extraction module, and a vehicle model identification number separated from the text extraction module by comparing it with a vehicle model classification table and a traffic information collecting means provided with a vehicle type classification module for classifying, wherein the vehicle classification table includes a first range for classifying passenger vehicles, a second range for classifying vans, and a third range for classifying freight cars A vehicle model classification code range consisting of a fourth range for classifying a range and a special vehicle is provided, and the two or three-digit vehicle model identification number extracted through the text extraction module is matched with the vehicle model classification code range to match the vehicle model are counted separately. According to this, the vehicle type can be easily distinguished by recognizing the license plate given as unique information of the vehicle from the photographed image data and matching the vehicle type classification table with the unique information of the recognized license plate.

Description

A traffic information assistance system based on license plate recognition }

The present invention relates to a vehicle number recognition-based traffic information support system that collects and aggregates vehicle traffic information so that it can be utilized in traffic, crime prevention, and control system construction.

According to the Road Traffic Volume Survey Guidelines of the Ministry of Land, Transport and Maritime Affairs, road traffic means “the number of vehicles passing a point or section of a road in a unit time”. In addition, the road traffic survey means observing the number of passing vehicles using the road for each vehicle type, each direction, and each time period.

The road traffic survey is carried out by dividing the regular and ad hoc surveys for each road in order to accurately grasp statistical data on vehicles passing through high-speed national roads, general national roads, nationally supported local roads, and local roads. Road planning and design, road operation, etc. are in progress.

Currently, road traffic surveys are conducted in parallel with methods using equipment such as AVC, WIM, VDS, and CCTV, and manpower surveys such as dispatching a surveyor directly to the survey site or checking the recorded images.

In detail, the AVC (Automatic Vehicle Classification) equipment is a stationary survey equipment, and it separates and collects the traffic volume, speed, vehicle type, etc. of a running vehicle, and is divided into a buried type and a non-buried type. Currently, the AVC installed in Korea is mainly used in a buried type with a loop-piezo sensor.

The VDS (Vehicle Detection System) equipment provides improved detection performance than the AVC and collects vehicle traffic volume, speed, vehicle type, etc. by embedding a loop sensor.

The WIM (Weight-In-Montion) equipment is a fixed survey equipment that is installed on the road surface to observe traffic parameters such as the weight, speed, and vehicle type of a running vehicle. Collect data such as uniaxial load factor.

The CCTV is a technology that transmits a signal to a limited monitor in a specific place. It is used to collect statistical data.

On the other hand, according to the contents posted on the website of the Ministry of National Highways and Transport (TMS), the traffic volume survey is divided by road type as shown in [Table 1] below, and is performed at a set date and time.

road type form Pause Contents
high-speed national highway
Occasional
Every year on the 3rd Thursday of October Investigation of traffic volume by vehicle type, direction, and time period
Investigation using AVC, VDS, and WIM and manpower survey through CCTV. always 365 days (24 hours a day) Continuous investigation of the use of AVC equipment
general national road
Occasional
Once a year Investigation by point using mobile traffic volume survey equipment
Only at some survey points, the vehicle model survey is conducted once a year for a 6-hour manpower survey. always 365 days (24 hours a day) Continuous investigation of the use of AVC equipment national support map
/locality Occasional Every year on the 3rd Thursday of October Investigate traffic volume by vehicle type, direction, and time by placing a surveyor at the survey point

Referring to [Table 1] above, the domestic road traffic survey is divided into frequent and regular depending on the type of road. Therefore, in the case of an ad hoc survey, a lot of time and money is consumed in determining the location of the survey, the size of the surveyor, and the classification, aggregation, and integrated processing of the surveyed data.

In addition, even in the case of regular irradiation, the fixed irradiation equipment such as the AVC, VDS, and WIM has a problem in that it is required to bury the sensor before construction and installation as it is installed on the road surface.

On the other hand, in order to solve the above problems, a system for collecting traffic volume using an image capturing device is being developed.

For example, in Republic of Korea Patent No. 10-1973725 “Unburied type classification device and method”, a technology to analyze road traffic by analyzing road images using two cameras without burying sensors in the road such as AVC, VDS, and WIM this is posted

In detail, in the prior art, a camera that divides two roads is configured to analyze the two road images, determine the number of axles and the length of the vehicle being driven, and compare them with the vehicle model table to distinguish the types of vehicles do.

To this end, the camera is configured as a thermal imaging camera, and configured to determine the number of axles and the vehicle length through thermal image analysis.

However, when a thermal imaging camera is used as described above, the recognition rate for a low-speed vehicle in summer with a high temperature may decrease, and in a city with a lot of traffic, a two-wheeled vehicle may be located between the vehicles while moving a lot, so the recognition accuracy is poor. There is a problem that can be lowered further.

In addition, a camera for recognizing the upper part of the vehicle and the license plate may be further provided to compensate for such a decrease in the recognition rate, but in this case, a processing process for vehicle object recognition and license plate recognition is further added, thereby complicating the overall system. .

As a result, the prior art as described above is more suitable for roads such as highways or general national roads in which vehicles can be driven at a constant speed or more, and accuracy may be lowered in urban areas where traffic is high and vehicle density is high.

KR 10-1973725 B1

It is an object of the present invention to provide a vehicle number recognition-based traffic information support system that can divide and aggregate vehicle types and form a DB while grasping the amount of traffic based on license plate recognition given as unique information of a vehicle regardless of road type.

Another object of the present invention is to provide a vehicle number recognition-based traffic information support system capable of providing traffic volume information according to a specific point as well as a set area range.

Another object of the present invention is to provide a vehicle number recognition-based traffic information support system capable of providing movement route data of a vehicle when it is necessary to track a vehicle for criminal tracking or crime prevention purposes.

Another object of the present invention is to provide a vehicle number recognition-based traffic information support system capable of classifying a plurality of movement routes existing between two selected points and providing traffic volume data for each movement route for comparison of each movement route. will be.

Another object of the present invention is to provide a vehicle number recognition-based traffic information support system in which image information captured through a photographing means is stored separately as images and texts, and vehicle types are distinguished based on low-capacity text.

Another object of the present invention is based on vehicle number recognition that can prevent exposure of personal information by identifying only some elements of the license plate without providing the entire data of the vehicle number, which is personal information, and assigning a safety number to prevent duplication of vehicle types. To provide a traffic information support system.

The vehicle number recognition-based traffic information support system according to the present invention recognizes a vehicle number from a photographing means for outputting image data by photographing a vehicle moving in a lane, and image information photographed from the photographing means, and sets the recognized vehicle number on the basis of the vehicle model classification table and a traffic information collection means for classifying and counting passenger vehicles, vans, freight vehicles and special vehicles, wherein the traffic information collection means sets a license plate area to the image data captured by the photographing means, and from the set area An image extraction module for extracting a license plate as an image, a text extraction module for recognizing the vehicle number and license plate color from the vehicle license plate image extracted from the image extraction module and extracting it as text information, and text information extracted from the text extraction module a vehicle model classification module for classifying the vehicle model by comparing it with the vehicle model classification table, wherein the vehicle model classification table includes a vehicle model classification code range that defines the vehicle model information range assigned at the time of vehicle registration from the vehicle number text information extracted from the text extraction module; , a color classification standard that defines the vehicle license plate color information given at the time of vehicle registration in the vehicle license plate color text information from the text extraction module is provided, matching at least one of the vehicle type classification symbol range and the color classification standard to classify and count the vehicle types characterized.

In another aspect, the present invention includes a photographing means for outputting image data by photographing a vehicle moving by car, and an image extraction module for extracting a vehicle object and a license plate area from the image data received from the photographing means, and storing the extracted license plate as an image; , a text extraction module for extracting at least a vehicle model identification number as text information from the license plate image extracted from the image extraction module, and a vehicle model classification module for comparing the vehicle model identification number separated from the text extraction module with a vehicle model classification table The vehicle classification table includes a first range for classifying passenger vehicles, a second range for classifying vans, a third range for classifying freight cars, and special vehicles. A vehicle model classification code range consisting of a fourth range for .

In another aspect, the present invention recognizes a vehicle number from a photographing means for outputting image data by photographing a vehicle moving by car, and image information photographed from the photographing means, and uses the recognized vehicle number on the basis of a vehicle type classification table for a passenger car and a van , a traffic information collection means for classifying and counting freight cars and special vehicles, wherein the traffic information collection means sets a license plate area in the image data captured from the photographing means, and uses the license plate as an image from the set area An image extraction module for extracting, a text extraction module for classifying at least a vehicle model identification number from the vehicle license plate image extracted from the image extraction module and extracting it as text information, and a vehicle model identification number and vehicle license plate image separated from the text extraction module A vehicle type classification module is included for comparing and classifying the vehicle license plate color with the vehicle type classification table, wherein the vehicle type classification table includes a first range for classifying passenger vehicles, a second range for classifying vans, and cargo A vehicle type classification symbol range divided into a third range for classifying automobiles and a fourth range for classifying special vehicles, and a sub-item of the vehicle type divided into the first to fourth ranges, according to the color of the license plate , a color classification standard for distinguishing diplomatic vehicles, transportation vehicles, heavy equipment vehicles and eco-friendly vehicles is included, and the 2- or 3-digit vehicle model identification number and vehicle license plate color extracted through the text extraction module are included in the vehicle model classification It is characterized in that the car model and use are classified and counted by matching with the symbol range and color classification criteria.

In another aspect, the present invention includes an image extraction module for receiving image data of a vehicle moving by a vehicle photographed from one or more photographing means, and extracting a license plate as an image by setting a license plate area from the received image, and the image extraction module A text extraction module for classifying at least a vehicle model identification number from the vehicle license plate image extracted from and extracting it as text information, a vehicle model identification module for comparing the vehicle model identification number separated from the text extraction module with a vehicle model classification table, and at least the photographing and an information unit forming module for forming an information unit by classifying the ID information of the means and analysis information items including time, location, vehicle number, and vehicle type classification, wherein the vehicle type classification table includes a first for classifying a passenger vehicle A vehicle type classification code range consisting of a range, a second range for classifying a van, a third range for classifying a freight car, and a fourth range for classifying a special vehicle is provided, and through the tact extraction module It is characterized in that the extracted two-digit or three-digit vehicle model identification number is matched with the vehicle model classification code range, and the vehicle model is classified and counted as the vehicle model classification item of the information unit.

According to the present invention, the vehicle type can be easily distinguished by recognizing the vehicle license plate given as unique information of the vehicle from the image data photographed through the photographing means and matching the vehicle type classification table with the unique information of the recognized license plate.

In addition, metadata is added to images taken from a plurality of photographing means, and analysis information items for each photographing means are formed as an information unit for analysis. By providing traffic analysis information in various forms by selecting it, it can be used in various ways in connection with traffic volume identification as well as crime prevention and control systems.

1 is a view showing an embodiment of a vehicle number recognition-based traffic information support system according to the present invention.
2 is a flowchart illustrating a process of processing the image data received by the traffic information collecting means, which is the main component of the present invention, into data for aggregation.
3 is a view for showing a text classification standard for classification of vehicle types in an extracted license plate image and an insertion form of metadata;
4 is a flowchart showing a text conversion process for car type classification from a car number image by a text extraction module, which is a main component of the present invention.
5 is a view for showing the range of the vehicle type classification symbol in the vehicle type classification table, which is the main component of the present invention.
6 is a view for showing a color classification standard in the vehicle model classification table, which is a main part of the present invention;
7 is a view showing an embodiment of a vehicle license plate for each purpose according to the color classification standard among the vehicle type classification table, which is the main component of the present invention.
8 to 13 are flowcharts illustrating a process in which a vehicle is classified by a vehicle type classification module, which is a major component of the present invention.
14 is a view showing an embodiment of an information unit, which is a main part of the present invention.
15 is a view for explaining a case of collecting data for traffic analysis by using the information unit according to the present invention.

Hereinafter, some embodiments of the present invention will be described in detail with reference to exemplary drawings. In adding reference numerals to components in each drawing, the same components are denoted by the same reference numerals as much as possible even though they are indicated in different drawings. is described In addition, in the description of the embodiment, when it is determined that a detailed description of a related known configuration or function interferes with the understanding of the embodiment of the present invention, the description is simplified or omitted, and the terms described below refer to the function in the present invention. As terms defined in consideration of these terms, they may vary depending on the intention or custom of the user or operator, and therefore, the definition will be preferably understood based on the content throughout this specification describing the present invention.

The present invention is configured to easily supply information generated by collecting, aggregating, classifying, and processing vehicle traffic information so that it can be utilized in traffic, crime prevention, and construction of a control system to a linkage system.

In detail, FIG. 1 is a view showing an embodiment of a vehicle number recognition-based traffic information support system according to the present invention.

Referring to the drawings, the vehicle number recognition-based traffic information support system 100 (hereinafter referred to as 'traffic information support system 100') according to the present invention is an image obtained by a photographing means 200 for photographing a vehicle moving by car. It is configured to include a traffic information collection means 400 that recognizes a vehicle number based on data and classifies it according to vehicle type. And, as described above, based on the information collected and generated in connection with the external system 600 including the monitoring unit 620 and the client control unit 640, traffic information according to the purpose of traffic, crime prevention, control, etc. is provided. do.

Here, traffic information refers to information that can be grasped by limiting one or more conditions such as time, location, range, route, type, and direction of movement with respect to the movement of the vehicle. have

Meanwhile, for convenience of explanation, referring first to FIG. 12 , the photographing means 200 is installed at a distance spaced apart from each other in a specific area in which the main road and the intersection road are extended as shown. do. Here, the installation location is preferably installed at a uniform distance, but in reality, it can be determined through an analysis of the installation location according to road conditions such as geographic location, cost, traffic flow, and traffic volume.

The photographing means 200 installed as described above is configured to photograph a vehicle moving in a detection target lane, respectively, and to store and transmit the photographed image data.

In addition, the photographing means 200 may be applied to any device capable of generating image data by photographing a moving vehicle, such as a vehicle number recognition camera (LPR Camera), a digital CCTV for crime prevention, an analog CCTV, and an IOT device.

In addition, the image data captured by the photographing means 200 may have various storage paths. That is, it may be recorded on a separate storage medium directly connected to the photographing means 200 , and may be stored in an integrated server of a management entity for managing the photographing means 200 through a network.

Therefore, the traffic information collecting means 400 according to the present invention is connected to the photographing means 200 through a network network to collect and process image data transmitted in real time, or to collect image data through a path in which previously photographed image data is stored. It is configured to receive and process.

In detail, when various types of photographing means 200 are installed as described above, each photographing means 200 has a different message format (API, Application Programming Interface), source and tool package (SDK, Software Development Kit). .

Therefore, in the traffic information collecting means 400, which is a major component of the present invention, a data processing unit corresponding to the type of the photographing means 200 to collect and aggregate image data in response to various APIs/SDKs for each photographing means 200 410 is provided.

That is, the data processing unit 410 includes the first to N ^th SDK/APIs 432 , 434 , and 436 to transmit/receive control signals and data to/from each of the photographing means 200 .

In addition, the video provided to the monitoring unit 620 is provided with a dedicated API/SDK for transmitting data to the above-described external system 600 and based on Real-Time Streaming Protocol (RTSP) and Simple Object Access Protocol (SOAP). In addition to data transmission, the control signal can be transmitted by linking the client control unit 640 and each photographing means 200 through the Internet.

Meanwhile, the traffic information collecting means 400 according to the present invention recognizes a vehicle number through the image data received from the photographing means 200, classifies the vehicle type based on the recognized vehicle number, and stores it in a database.

2 is a flowchart showing a process of processing the image data received from the traffic information collecting means, which is the main component of the present invention, into data for aggregation. A diagram is shown to show the classification criteria and the insertion form of metadata.

Referring to these drawings, in the traffic information collecting means 400, first, an image extraction module for setting a license plate area in the image data photographed from the photographing means 200, and extracting the license plate as an image from the set area ( 442) is provided.

The image extraction module 442 acquires a basic background image after storing the entire received image data.

That is, after separating the image data extraction frame 520 from the entire image data, a region of interest 524 for detecting the vehicle object 522 is set on the separated image data extraction frame 520, and the interest set in this way The area 524 is applied as a whole to the received image data. Here, a plurality of regions of interest 524 may be set according to a detection target lane.

When the image in which the region of interest 524 is set as described above is received, the shadow is removed through the background difference operation, and the object is detected based on the morphology operation.

That is, when a target passing through the region of interest 524 is identified, it is detected as an object, and the detected object image is received to determine whether the vehicle object 522 is present.

Here, the determination of the vehicle object 522 is made by comparing the shape of the object detected through the morphology operation with the trained data. By increasing the count, the total traffic volume can be counted regardless of the vehicle type.

On the other hand, license plates are formed of rectangular plates attached to the front and rear of all officially registered vehicles. The countries of the United States use a mixture of roman characters and numbers on license plates.

Therefore, the license plate used as described above contains vehicle-specific information and, in particular, classifies the vehicle type, so the present invention recognizes the vehicle number based on such a classification criterion, extracts it as text information, and extracts the extracted text information. The vehicle model classification task is performed by matching the unique information about the vehicle model in the

Hereinafter, an embodiment of the present invention will be described with a focus on the Republic of Korea in order to more clearly explain the present invention.

As described above, when the vehicle object 522 is determined through the image extraction module 442 , the vehicle license plate image is extracted based on the license plate area 524 .

The license plate region 524 is formed inside the region of interest 524 on the image data extraction frame 520 in the same manner as the region of interest 524 .

In the setting of the license plate area 524, since there is a difference in the size, shape and position of the license plate depending on the vehicle model, the license plate area 524 is formed on the upper side and the lower side along the inner center of the area of interest 524, respectively. Detection omissions can be reduced.

Even if the license plate area 524 is set as described above, there may be cases in which license plate recognition is impossible depending on the speed of the vehicle or the degree of contamination of the license plate. Accordingly, when the vehicle object 522 is recognized, but the license plate is not recognized, the image of the corresponding object is divided into error information and stored separately.

On the other hand, when the license plate is correctly recognized, a license plate image 540 is generated, and a license plate recognition area is set in the generated license plate image 540 .

The vehicle number recognition area is divided in detail into a vehicle model identification number area 542, a use identification character area 544 and a registration serial number area 546 as shown in FIG. 3, and when each of the detailed areas is recognized, the vehicle number The image is extracted.

After the vehicle number image is extracted as described above, metadata 541 for adding information on the photographing means 200 for photographing the vehicle number is further added, and metadata 541 is linked to the extracted vehicle number image A car number image is created with

In addition, the vehicle number image may further include color information of the vehicle license plate in addition to the vehicle number, which may be used to classify vehicle types based on reference information of a vehicle classification table to be described below.

On the other hand, the vehicle number image generated as described above is converted and extracted into text format through a text extraction module, which will be described below, in which the vehicle number and metadata 541 record information included therewith are converted.

4 is a flowchart showing a text conversion process for classifying a vehicle type from a vehicle number image by the text extraction module, which is the main component of the present invention.

Referring to the drawings, the text extraction module 444 loads the vehicle number image extracted through the image extraction module 442, and sets an area to be converted into text and extracted.

That is, a feature of the present invention is to set an area for identifying the minimum information for car model classification in the extracted car number image, and classify and aggregate car models based on the recognized information.

In addition, in the present invention, when the entire information of the vehicle number image is disclosed, exposure of personal information may be concerned, and even when the storage deadline for image data according to the regulations is exceeded, the vehicle's travel information can be checked without exposing personal information. Data reprocessing is performed to

To this end, the text extraction module 444 sets the part corresponding to the vehicle model identification number area 542 and the part corresponding to the use identification character 544 in the vehicle number image, and identifies the vehicle model from the loaded vehicle number image. Converts numbers and usage identification characters to text and extracts them.

And, the vehicle model identification number and the use identification character extracted as described above are used as a target for determining the vehicle type by the vehicle model classification module 446, which is a major component of the present invention.

In detail, FIG. 5 is a diagram showing the range of the vehicle type classification symbol in the vehicle type classification table, which is a major component of the present invention, and FIG. 7 is a diagram showing an embodiment of a vehicle license plate for each purpose according to the color classification standard.

Referring to these drawings, the vehicle type classification table is configured to include a criterion for classifying the type of vehicle and a criterion for classifying the type of the vehicle.

In detail, the vehicle type classification code range, which is included in the vehicle type classification table and is a standard for classifying vehicle types, is divided into four vehicle types based on the vehicle type identification number in the text information extracted from the text extraction module 444. .

The four vehicle types classified by the vehicle type classification code range are divided into passenger vehicles, vans, freight vehicles, and special vehicles, and a range serving as a criterion for judgment is allocated to each.

The first range is a range applicable to passenger vehicles, and is defined as 01 to 69 and 3 digits 100 to 699 of the two-digit vehicle model identification number.

And, the second range corresponding to the van is defined as 70 to 79 of the two-digit vehicle model identification number, and the third range corresponding to the freight vehicle is defined as the vehicle model identification number 80 to 97. Finally, the fourth range corresponding to special vehicles is defined by vehicle model identification numbers 98 and 99.

On the other hand, in relation to the classification of vehicles by use, the number of eco-friendly vehicles such as electric vehicles is increasing, and a color classification standard for dividing license plate images into five colors may be further included in order to classify heavy equipment vehicles.

In detail, the color classification standard divides the color filling in between the number and letter and the border of the license plate into first to fifth colors, and the first color is defined as a general vehicle, white.

The second color is defined as royal blue for diplomatic vehicles, the third color is defined as yellow for transportation, the fourth color is defined as yellow for heavy equipment, and the fifth color is defined as blue for eco-friendly vehicles.

Hereinafter, a process of classifying a vehicle by the vehicle type classification module 446 will be described.

8 to 13 are flowcharts illustrating a process of classifying a vehicle by a vehicle type classification module, which is a major component of the present invention.

8 is a process for classifying the type of vehicle using the vehicle identification number. When the vehicle number text is loaded, the process of matching the vehicle model identification number with the first to fourth ranges in the loaded vehicle number text is performed. .

In the matching process, if the vehicle model identification number falls within the first range, it is classified as a passenger vehicle; if it falls within the second range, it is a van; if it falls within the third range, it is a freight car; are separated and stored in a DB separated by vehicle type.

However, if the vehicle identification number is not properly recognized due to the speed of the detected vehicle object or the degree of contamination of the license plate, and for this reason, it is not included in the first to fourth ranges. They are stored together as a temporary DB.

On the other hand, in the case of the vehicle number text matched for each range and stored in the DB for each vehicle type, a process of matching with the color classification standard for classification by use is performed.

In addition, the data stored in the temporary DB as described above is classified for the purpose of the vehicle through matching with the color classification criterion.

Matching of the text stored in the temporary DB is made for the color of the license plate extracted from the license plate image, and as shown in FIG. 9, the color of the license plate corresponds to any one of white, yellow, yellow, blue, and royal blue. In each case, it is stored in the DB for each use of the vehicle.

However, as described above, when the color classification of the license plate is not clearly made when the object is detected, it is separately stored in the error DB as the final unrecognized vehicle.

In addition, the classification by use of the vehicle performs matching with the color classification criteria first, and if not classified, the vehicle number text and the vehicle license plate color are stored together in a temporary DB to reclassify.

Meanwhile, the information classified and stored through the matching process with the vehicle type classification table as described above is reprocessed into an information unit to be utilized as the aforementioned traffic information.

To this end, the traffic information collection module 400 is further provided with an information unit formation module 448 .

In addition to the type and use of the vehicle described above, the information unit forming module 448 includes the ID of the photographing means included in the metadata 541, the time at which the object was detected, the detection location information, as well as the original image in which the object was photographed. The analysis information item is formed to include the stored route, the related DB information, so that it can be utilized for various traffic analysis.

14 is a diagram showing an embodiment of an information unit, which is a main part of the present invention.

Referring to this, the analysis information items included in the information unit include photographing means ID, photographing time, photographing location, vehicle number, safety number, vehicle model classification, use classification, and related DB information. After creation of the number item, it may be masked.

First, the photographing means ID item is for identifying the photographing means 200 installed to photograph the vehicle, and is given based on the installation location.

For example, the photographing means ID may indicate the name of the installed street as an abbreviation, and a serial number may be assigned in the order of installation in the corresponding street.

The time at which the vehicle object is detected by the corresponding photographing means 200 is recorded in the photographing time item.

The photographing location item is information corresponding to the installation location of the corresponding photographing means 200, and may be given by applying an Open Location Code (OLC) provided by Google or a geographic coordinate system indicated by latitude and longitude. .

The vehicle number item is given a vehicle number according to vehicle number recognition as described above, and when it is transmitted to the external system 600 for utilizing traffic information, only the safety number item is exposed, and the vehicle number is masked. The leakage of personal information is prevented.

The safety number item is given so that the vehicle number can be tracked in order to provide various traffic information such as movement route information of a specific vehicle when an actual vehicle number is not required based on the collection of traffic information.

In detail, as for the relief number given based on the vehicle number, the vehicle model identification number and the use identification letter are the same in the recognized vehicle number, and the first recognized vehicle among the vehicles having the vehicle identification number and the use identification letter is English letters A serial number is assigned as a combination of and a number.

In the information unit transmitted to the outside after the serial number is assigned as described above, at least the registration serial number of the corresponding vehicle number is masked, and when the real vehicle number of the given security number is required in the future, such as a police investigation, a code for removing the masking by providing the actual vehicle number can be confirmed.

On the other hand, as described above, the vehicle type classification item and the use classification item are given by matching the vehicle number text and vehicle number image with the vehicle type classification table, and related DB information is set before and after based on the object detection time of the vehicle It is given as storage location information of the edited original video by the time range.

For example, the vehicle number and safety number of each vehicle detected at the same location, the vehicle type, use, and related DB information are checked, and related DB information of the vehicle photographed at 09:30, 09:32, and 09:35 In the item, the storage path of the video edited with VDB20200312-0930 is identified, and in the case of a vehicle filmed at 09:40 and 09:45, the storage path is written as VDB20200312-0940.

That is, in this embodiment, an image edited in units of 10 minutes is stored in the image data DB 462, and information about the storage path is recorded in the related DB information item of the information unit. Therefore, when the user wants to check the driving image of a specific vehicle, he or she can easily check it through the file stored in the corresponding route.

On the other hand, referring back to FIG. 1, the data generated by the image extraction module 442, the text extraction module 444, the vehicle model classification module 446, and the information unit calculation module 448 are classified according to the type of data. It is stored in the data DB.

In detail, information such as the license plate image 540 and the detected vehicle object generated by the image extraction module 442 is stored in the image data DB 464, and vehicle number text information generated by the text extraction module 444 is stored in the text data DB (466).

In addition, the data processed through the information unit forming module 448 is stored separately in the information unit DB 468, and the data stored in the separate DB are linked as family information.

Hereinafter, a situation of grasping traffic information using the present invention having the above configuration will be described with reference to the accompanying drawings.

15 is a diagram for explaining a case of collecting data for traffic analysis by using the information unit according to the present invention.

Referring to, on the monitor screen 610 that can be checked from the external system 600, the main road and the intersection road are linked to each other and a road network appears, and a plurality of photographing means 200 are identified at major locations on the road network.

The information unit according to the present invention can be utilized when the system operator wants to understand the movement path of a specific vehicle on the road network screen as described above.

First, when the system operator recognizes the vehicle number, the masking area may be removed by using the masking removal code to confirm the corresponding vehicle number, and the vehicle number may be searched.

And, when the vehicle number is searched, the information unit can determine the ID of the photographing means from which the vehicle number is detected, the photographing time, the photographing location, etc. can

That is, as shown in FIG. 15 , when a vehicle “123 is 4567” is searched for, the vehicle of the corresponding vehicle number is detected by the photographing means 1 → photographing means 4 → photographing means 8 → photographing means 10, and the vehicle is It can be determined that the vehicle has moved along the movement path 1 (612) shown by a yellow line.

Meanwhile, the information unit according to the present invention can be utilized even when the system operator wants to compare two or more movement paths formed between two points.

In detail, the path moving between the photographing means 1 and the photographing means 10 includes, in addition to the movement path 1 (612), the movement path of the green line leading to the photographing means 1 → photographing means 6 → photographing means 11 → photographing means 10 (hereinafter referred to as “movement path 2”). ', 614) can be compared.

That is, in the case of comparing the amount of traffic in the same time period for the movement route 1 (612) and the movement route 2 (614), the traffic amount of the photographing means 1, 4, 8, and 10 included in the movement route 1 (612) and the movement route Comparison between routes can be made by determining the amount of traffic of the photographing means 1, 6, 11, and 10 included in 2 (614) and comparing them.

In addition, the information unit according to the present invention can be utilized even when the system operator wants to determine the density by checking the number of vehicles that exist in a specific area for a certain period of time.

That is, when checking the vehicle density in the area "B" shown in FIG. 15 for a specific time period, the security number recognized by the photographing means 1, 3, 5, 6, 7, 9, 12 included in the area is identified. Recognized vehicles for each photographing means are detected, and overlapping vehicles are removed from the detected vehicles to determine the vehicle density in the corresponding time period.

In addition, even in the case of counting eco-friendly vehicles or heavy equipment vehicles that have moved for a certain period of time, it can be identified by using the information unit according to the present invention, and more diverse applications can be made in addition to that.

200........ Means of filming 400........ Means of collecting traffic information
442......... Image extraction module 444........ Text extraction module
446........ Vehicle type classification module 448........ Information unit formation module
600............. External system

Claims (9)

a photographing means for photographing a vehicle moving by car and outputting image data; and
Traffic information collecting means for recognizing the vehicle number from the image information photographed by the photographing means, and for classifying the recognized vehicle number into passenger vehicles, vans, freight vehicles and special vehicles based on the vehicle type classification table; and
In the traffic information collection means,
An image extraction module for setting a license plate area in the image data taken from the photographing means, and extracting a license plate from the set area as an image;
A text extraction module for recognizing the license plate number and the license plate color from the license plate image extracted from the image extraction module and extracting it as text information; and
A vehicle type classification module that compares the text information extracted from the text extraction module with the vehicle type classification table to classify the vehicle model, extracts the vehicle registration serial number from the vehicle number converted into text, and converts it into a safety number; and
an information unit forming module that forms an information unit by classifying at least the ID information of the photographing means, and analysis information items including time, location, vehicle number, and vehicle type classification;
a traffic analysis information calculation module for calculating traffic analysis information through classification and calculation processing for the information unit based on one or more items selected from among the analysis information items; and
An information unit DB for storing the information unit is included,
In the vehicle model classification table,
In the vehicle number text information extracted from the text extraction module, the vehicle model classification code range that defines the vehicle model information range given at the time of vehicle registration, and the vehicle license plate color information given at the time of vehicle registration from the vehicle license plate color text information from the text extraction module. A color classification standard is provided so that at least one of the vehicle model classification symbol range and the color classification standard is matched to classify and count vehicle types,
In the traffic information collection means,
In the information unit transmitted to the outside, at least the registration serial number of the corresponding vehicle number is masked. based traffic information support system.
delete delete delete delete According to claim 1, wherein the traffic information collection means,
an image DB for storing image data received from one or more photographing means;
an image DB in which the image data extracted from the image extraction module is stored;
A text DB in which the text data extracted from the text extraction module is stored is further included,
Vehicle number recognition-based traffic information support system, characterized in that the analysis information item further includes at least a related DB information item in which the file storage location of the image DB in which the original image captured by the corresponding photographing means is stored. 7. The method of claim 6,
The related DB information includes a file storage location of the image DB in which the original image captured by the corresponding photographing means is stored or the file storage location of the image DB in which the vehicle object image and license plate image extracted from the original image are stored. Number recognition-based traffic information support system. delete delete

Images