KR101784764B1 - The highway toll payment method for electric vehicle - Google Patents

Abstract

The method for collecting tolls on an expressway of an electric vehicle according to the present invention is a method for collecting tolls on an expressway tolls by a tollgate on a highway where a toll charge is settled by a toll charge system or a high- A step of photographing a forward image or a backward image of a vehicle passing through a gantry or a vehicle entering a highway toll gate entrance using a camera using a camera and a license plate recognizer installed in a gantry of a highway; Extracting a license plate area from a forward or backward image of a vehicle passing through a vehicle or a gantry where the vehicle detecting and license plate recognizer enters a highway toll gate entrance; Determining whether an electric vehicle image logo (A) and an EV character logo (B) unique to the electric vehicle are written in the vehicle license plate area from which the vehicle detection and license plate recognizer is extracted; The vehicle detection and license plate recognizer judges that the electric vehicle is an electric vehicle if the logo of the electric car image A and the logo of the EV character (B) are described in the extracted license plate area, and the current highway charge settlement system is a toll charge charge settlement system, When the vehicle is operated by the pass system, it stores the vehicle number detected in the fare settling server, the presence of the electric vehicle, the time of entering the highway, and the name of the highway tollgate entered. If the current highway fare settlement system is operated by the smart tolling system, Storing the vehicle number of the vehicle that has passed through the gantry, whether the vehicle is an electric vehicle, the gantry passage time, the lane position when passing through the gantry, and the identifier (ID) of the vehicle detector and license plate recognizer mounted on the gantry; Tollgate tolls settled at the tollgate fare booth Vehicle detectors and license plate recognizers installed on the tollgate of the highway operated by the fare settlement system or the high pass system are used to detect the forward and backward images of the vehicle leaving the highway toll gate exit The vehicle license plate area is extracted from the forward image or the backward image of the photographed vehicle, and it is confirmed whether the electric vehicle image logo (A) or the EV character logo (B) is written in the extracted license plate area. (A) or the EV character logo (B) is written on the vehicle, it is judged that the vehicle is an electric car, and the vehicle number detected on the fare settling server, whether the electric car is present, the time of departure to the expressway, Storing; And a toll gate fare adjustment system for highway tolls settled in the toll gate fee booth is operated by the vehicle number stored in the fare settling server, the electric car number, the toll gate number of the highway toll gate And the tollgate of the highway to the outgoing highway tolls, and then billing the departing vehicle. The fare adjuster of the highway operated by the smart tolling system compares the toll on the highway toll vehicle with the vehicle number stored in the fare settling server, The gantry passage, the passage through the highway linkage section, and the gantry passage time, and then charging the vehicle to the departing vehicle.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a highway toll payment method for electric vehicles,

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of collecting tolls on an expressway of an electric vehicle, and more particularly, to a method of collecting tolls on an expressway using an expressway, such as a tollgate or a highway gantry, And then charging the electric vehicle leaving the expressway for a discounted highway fee.

At present, the government has decided to reduce the toll of electric cars and hydrogen cars toll roads by 50% from September 2017 in order to reduce fine dust and to support incentives to expand the supply of eco - friendly cars.

However, Hybrid cars that use petroleum fuel during highway driving are excluded.

On the other hand, the design of the electric car license plate was confirmed by the Ministry of Land Transport and Transport as a standard form as shown in FIG. 1. The base plate of the electric car license plate is filled with blue color, .

In addition, an existing vehicle number is described on the front of the electric vehicle license plate, an electric vehicle image logo (A) is written on the upper left side of the electric vehicle license plate, and an EV character logo (B) is written on the right side of the electric vehicle license plate.

On the other hand, the highway toll gatekeepers will be able to visually check the blue license plate number, and then discount the toll charges.

On the other hand, in order to reduce the toll of an electric vehicle using a high pass, a high pass terminal dedicated to an electric car must be installed in the electric vehicle.

Also, if the high-pass terminal is a conventional high-pass terminal, the electric vehicle and the hydrogen vehicle identification code must be input to the existing high-pass terminal by using the Korea Highway Corporation's office or a dedicated homepage.

The electric and hydrogen car expressway toll discount system will be applied for a limited period of time until 2020, which is the target year for eco-friendly cars.

However, there is an inconvenience that the above-mentioned highway toll discount method requires the storage staff in the toll gate fee booth to visually confirm whether or not an electric car is present, and then 50% discount on the expressway toll to the electric car or hydrogen car.

In addition, there is an inconvenience that an electric car or a hydrogen car must be equipped with an electric car or a high-pass terminal dedicated to the hydrogen car when the hydrogen car passes through the high-pass section.

On the other hand, in the prior art of the present invention, "High Pass Fee Reduction System and Method" of the patent registration number "10-1461996" was filed and registered. The high pass toll reduction system and method includes a high pass adjuster installed in a toll gate A high pass terminal operating in any one of a normal mode in which tolls are settled by wireless communication and a toll reduction is not performed and a toll reduction mode in which toll reduction is performed corresponding to boarding of the subject to be abated; The mobile terminal includes a fingerprint recognition function, a GPS function, and a short-range wireless communication function. When the fingerprint is authenticated as a fingerprint of the subject to be reduced by using the fingerprint recognition function, The high-pass terminal transmits a first control signal for instructing the high-pass terminal to switch from the normal mode to the reduced mode, and the high-pass terminal is switched from the normal mode to the reduced mode when the first control signal is received from the portable terminal, When the vehicle passes through the tollgate in the reduction mode, the high-pass toll adjuster is communicatively communicated with the high-pass to settle the reduced toll corresponding to the boarding of the person to be reduced.

Korean Patent Publication No. 10-2010-0097545 (September 3, 2010) Korea Patent Registration No. 10-1461996 (Aug. 2014)

Accordingly, in order to solve the above problems, the present invention provides a method for automatically collecting tolls according to an electric vehicle, by automatically recognizing an electric vehicle without visually confirming whether or not the storage staff in the toll gate charge booth departs from the expressway SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problems occurring in the prior art.

It is another object of the present invention to provide an electric vehicle which can pass an expressway toll without an existing high-pass terminal or a high-pass terminal, without installing a high- And to provide a method of collecting tolls on an expressway of an electric car so as to be able to do so.

In order to achieve the above object, according to the present invention, there is provided a highway toll collection method for an electric vehicle, comprising: a tollgate system for charging a highway toll at a toll gate fee booth, or a toll gate or a smart tolling system for a highway toll, A step of photographing a forward image or a backward image of a vehicle passing through a gantry or a vehicle entering a highway toll gate entrance using a camera by a vehicle detection and license plate recognizer installed in a gantry of a highway tolls of which the highway toll is settled; Extracting a license plate area from a forward or backward image of a vehicle passing through a vehicle or a gantry where the vehicle detecting and license plate recognizer enters a highway toll gate entrance; Determining whether an electric vehicle image logo (A) and an EV character logo (B) unique to the electric vehicle are written in the vehicle license plate area from which the vehicle detection and license plate recognizer is extracted; The vehicle detection and license plate recognizer judges that the electric vehicle is an electric vehicle if the logo of the electric car image A and the logo of the EV character (B) are described in the extracted license plate area, and the current highway charge settlement system is a toll charge charge settlement system, When the vehicle is operated by the pass system, it stores the vehicle number detected in the fare settling server, the presence of the electric vehicle, the time of entering the highway, and the name of the highway tollgate entered. If the current highway fare settlement system is operated by the smart tolling system, Storing the vehicle number of the vehicle that has passed through the gantry, whether the vehicle is an electric vehicle, the gantry passage time, the lane position when passing through the gantry, and the identifier (ID) of the vehicle detector and license plate recognizer mounted on the gantry; Tollgate tolls settled at the tollgate fare booth Vehicle detectors and license plate recognizers installed on the tollgate of the highway operated by the fare settlement system or the high pass system are used to detect the forward and backward images of the vehicle leaving the highway toll gate exit The vehicle license plate area is extracted from the forward image or the backward image of the photographed vehicle, and it is confirmed whether the electric vehicle image logo (A) or the EV character logo (B) is written in the extracted license plate area. (A) or the EV character logo (B) is written on the vehicle, it is judged that the vehicle is an electric car, and the vehicle number detected on the fare settling server, whether the electric car is present, the time of departure to the expressway, Storing; And a toll gate fare adjustment system for highway tolls settled in the toll gate fee booth is operated by the vehicle number stored in the fare settling server, the electric car number, the toll gate number of the highway toll gate And the tollgate of the highway to the outgoing highway tolls, and then billing the departing vehicle. The fare adjuster of the highway operated by the smart tolling system compares the toll on the highway toll vehicle with the vehicle number stored in the fare settling server, The gantry passage, the passage through the highway linkage section, and the gantry passage time, and then charging the vehicle to the departing vehicle.

The method for collecting tolls on an expressway of an electric vehicle according to the present invention, which is made up of such a procedure, is a method for collecting tolls of an electric car on a freeway tollgate or gantry on a highway operated by a toll charge system, high pass system or smart toll system, And the fare adjuster can charge the discounted highway fare according to the electric car to the electric car owner by understanding the highway route and the highway linkage section by the car number.

Therefore, the highway toll collection method of an electric vehicle according to the present invention can automatically recognize the electric vehicle without confirming visually whether the storage staff in the toll gate booth in the expressway has the electric vehicle of the outgoing vehicle and can automatically calculate the toll of the highway toll driver .

Further, according to the present invention, an electric car passing through a high-pass section and a multi-car smart tolling system can pay a highway toll without a high-pass terminal or a high-pass terminal without installing a high-

크기의 국소 윈도우 영역을 설정한 일실시 예,
도면 5는 전기차 번호판의 흑백 영상에서 숫자 및 문자와 전기차 이미지 로고(A), EV 문자 로고(B)의 후보 영역을 검출한 일실시 예,
도면 6은 차량 번호판 인식용 한글 및 숫자 전용 인공 신경망에서 전기차 이미지 로고(A)와, EV 문자 로고(B)를 인식하기 위해 숫자 인식용 인공 신경망의 구조를 확장한 일실시 예,
도면 7은 본 발명이 스마트톨링 시스템을 통해 전기차에게 고속도로 통행료를 과금하는 절차를 순차적으로 도시한 도면.Figure 1 shows an image of a license plate issued by the Ministry of Land,
2a and 2b are a control flow diagram of the present invention,
FIG. 3 shows an example of conversion of an electric car license plate photographed by a camera into a monochrome image,
4,

Size local window region,
5 shows an example in which candidate regions of numbers and letters, an electric vehicle image logo (A) and an EV character logo (B) are detected in a monochrome image of an electric car license plate,
6 is an embodiment of an extended artificial neural network structure for recognition of an electric vehicle image logo (A) and an EV character logo (B) in Korean and numeric artificial neural networks for license plate recognition,
FIG. 7 is a diagram sequentially illustrating a procedure for charging an electric vehicle with a highway toll through the smart tolling system.

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

As shown in FIGS. 2A to 2B, the method of collecting tolls on an expressway of an electric vehicle according to the present invention is a method of collecting tolls on a tollgate or a tollgate on a highway tolls of highway tolls settled by a toll- , A vehicle detection and license plate recognizer (1) installed in a gantry of a highway where a toll charge is settled by a smart tolling system, and a forward image of a vehicle passing through a gantry or a vehicle entering a highway toll gate entrance using a camera A step S1 of photographing a backward image or a backward image; (S2) extracting a license plate area from a forward or backward image of a vehicle passing through a vehicle or a gantry where the vehicle detecting and license plate recognizer (1) enters a highway toll gate entrance; (S3) judging whether an electric vehicle image logo (A) and an EV character logo (B) unique to the electric vehicle are written in the vehicle license plate area from which the vehicle detection and license plate recognizer (1) is extracted; The vehicle identification and license plate recognizer 1 judges that the electric vehicle is an electric car if the logo of the electric car image A and the logo of the EV character B are described in the extracted license plate area and the current highway charge settlement method is a toll charge settlement System, or a high-pass system, it stores the vehicle number detected in the fare settling server 7, the presence of an electric car, the time of the highway entrance, and the name of the toll gate of the highway. The vehicle number of the vehicle that has passed through the gantry, the presence or absence of the electric car, the gantry passage time, the position of the lane when passing through the gantry, and the vehicle detection and license plate recognizer mounted on the gantry 1) (S4); A vehicle detection and license plate recognizer (1) installed on a toll gate of a highway operated by a toll charge settlement system or a high pass system that clears a toll charge toll gate at a toll gate fee booth is a forward image of a vehicle After the rear image is photographed, the vehicle license plate area is extracted from the forward image or the backward image of the photographed vehicle, and it is checked whether the electric vehicle image logo (A) or the EV character logo (B) If the logo of the electric car image (A) or the logo of the EV character (B) is written in the license plate area, it is determined that the vehicle is an electric car and the vehicle number detected by the charge settlement server (7) Storing the name of the toll gate on the highway (S5); And a toll charge system of a highway operated by a toll charge system for setting up toll charge on a toll gate fee booth and a highway system operated by a high pass system compares the toll on the highway toll vehicle with the vehicle number stored in the toll charge server 7, The toll account of the expressway tollgate, the tollgate name of the tollgate on the highway, and the tollgate name of the highway to which the vehicle has departed, and then bill the vehicle to the departing vehicle. The fare adjuster of the highway operated by the smart tolling system, (Step S6) of charging the vehicle with reference to the number, the electric vehicle, the gantry passage, the passage of the highway linkage, and the gantry passage time.

The EV character logo B is an electric vehicle (EV).

The design of the electric car license plate has been established in a standard form as shown in FIG. 1, which will be described in more detail. The base plate of the electric car license plate is filled with blue color, It is engraved.

In addition, an existing vehicle number is written on the front face of the electric car license plate, an electric vehicle image logo A is written on the upper left side of the electric car license plate, and an EV character logo B is written on the right side of the electric car license plate.

The image of the electric car license plate is an image announced by the Ministry of Land, Transport and Maritime Affairs. As of the end of May 2017, there are 14,861 electric cars registered in Korea.

Since most of the cameras installed in the existing license plate recognition system do not utilize the color information of the license plates, it is only necessary to refer to the logos added to the license plate of the electric car without using the blue color information of the license plate, It should be possible to judge whether

In order to judge whether the vehicle is an electric car only with the logos added to the electric vehicle license plate, development of an algorithm that can display the electric vehicle image logo (A) or the EV character logo (B) clearly when the electric car license plate is converted into a monochrome image This is very important.

The electric vehicle image logo (A) or the EV character logo (B) engraved on the electric car license plate has an intermediate brightness level between the car number color engraved on the electric car license plate and the blue background color.

Accordingly, when converting an electric car license plate into a black and white image, it is possible to eliminate the blue-colored flag patterns displayed on the base of the license plate, to display the electric vehicle image logo (A), the EV character logo (B) Algorithm is needed.

The step S2 of extracting the license plate area from the forward or backward image of the vehicle passing through the vehicle or the gantry of the vehicle detection and license plate recognizer 1 enters the highway toll gate entrance, A step of performing Gaussian smoothing and median filtering on the front image or the rear image of the electric car in order to remove the front image or the back image noise of the electric car; Extracting an edge component in a horizontal direction after the vehicle detection and license plate recognizer (1) performs second differential filtering on the entire forward or backward image of the electric vehicle; Generating a character segment image by assigning the same label value to successive segments generated by vertically stacking edge segments connected to extracted edge components; And detecting a candidate region of the electric license plate by finding four consecutive number blobs in the generated edge-based character segment image.

If distortion occurs in the license plate area, many unique geometric features of the license plate may be lost, but regularity remains in the size and position of the four consecutive large numbers.

Therefore, in order to realize the license plate invariant license plate recognition and recognition system, four consecutive large digits are firstly found and then the license plate area is estimated.

In general, when an adaptive binarization algorithm is used to convert an electric license plate into a black and white image, the binarization parameters, that is, the local window size used for obtaining the threshold value for binarization and the reflection ratio of the brightness level distribution of the pixels are fixed The image quality of the converted black and white image may be changed, so that the logo type of the electric car can not be accurately extracted.

For this purpose, it is necessary to set variable binarization parameters in order to convert the electric car image logo (A) or the EV character logo (B) into black and white images accurately when converting the license plate into a black and white image.

The step S3 of judging whether or not the electric vehicle image logo A and the EV character logo B unique to the electric vehicle are described in the vehicle license plate area from which the vehicle detection and license plate recognizer 1 is extracted is performed by the vehicle detection and license plate recognizer (A) and an EV character logo (B) are described in a vehicle license plate converted into a monochrome image by the vehicle identification and license plate recognizer (1) (A) and an EV character logo (B) are extracted from the vehicle license plate and the candidate region extracted and classified from the license plate is extracted from the vehicle license plate, (Artificial Neural Network), which recognizes and classifies the candidates, and confirms whether the logo of the electric vehicle image (A) and the logo of the EV character (B) are included in the candidate region. And classifying whether the region is an electric vehicle image logo (A) or an EV character logo (B).

을 구한 다음, 평균 명도값

이 상기 차량 검지 및 번호판 인식기(1)에 설정된 명도 임계값

이하인지 확인하는 단계와; 상기 차량 검지 및 번호판 인식기(1)는 평균 명도값

이 상기 차량 검지 및 번호판 인식기(1)에 설정된 명도 임계값

이하라면 촬영된 전기차 번호판 영상이 흐린 영상이라고 판단하여 선명한 이진화 영상을 얻기 위해 각 화소들의 명도 레벨 분포에 대한 반영 비율

값을 상기 차량 검지 및 번호판 인식기(1)에 저장된 하이 값(High value)으로 설정하는 단계; 상기 차량 검지 및 번호판 인식기(1)는 평균 명도값

이 차량 검지 및 번호판 인식기(1)에 설정된 명도 임계값

을 초과하였다면 불필요한 영상 일부가 이진화됨을 막기 위해 각 화소들의 명도 레벨 분포에 대한 반영 비율

값을 차량 검지 및 번호판 인식기(1)에 저장된 로우 값(Low value)으로 설정하는 단계; 상기 차량 검지 및 번호판 인식기(1)는 전기차 번호판 영상 내 숫자들의 평균 높이

을 구한 다음, 상기

이 차량 검지 및 번호판 인식기(1)에 설정된 숫자 높이의 임계값

이하일 경우 전기차 번호판 영상이 작고 흐린 영상이라고 판단하여 국소 윈도우(Window) 크기 값

을 차량 검지 및 번호판 인식기(1)에 저장된 하이 값(High value)으로 설정하는 단계; 상기 차량 검지 및 번호판 인식기(1)는 전기차 번호판 영상 내 숫자들의 평균 높이

을 구한 다음, 상기

이 차량 검지 및 번호판 인식기(1)에 설정된 숫자 높이의 임계값

을 초과하였을 경우 국소 윈도우 크기 값

을 차량 검지 및 번호판 인식기(1)에 저장된 로우 값(Low value)으로 설정하는 단계; 상기 차량 검지 및 번호판 인식기(1)는

식을 이용하여 국소 임계치(Local threshold value) T를 구하는 단계; 및 상기 차량 검지 및 번호판 인식기(1)는 국소 윈도우 영역 내 포함된 각 화소들의 명도 레벨과 국소 임계치 T를 순차적으로 비교하여 특정 화소의 명도 레벨이 국소 임계치 T보다 크면 특정 화소를 백색 화소('1')로 변환하고 특정 화소의 명도 레벨이 국소 임계치 T보다 낮으면 특정 화소를 흑색 화소로 변환함으로써 전기차 번호판의 명도 영상을 흑백 영상으로 변환 및 이진화하는 단계로 이루어질 수 있다.As shown in FIGS. 3 and 4, the step of converting the vehicle license plate into a black and white image is carried out by the vehicle detection and license plate recognizer 1, ) ≪ / RTI > for all the pixels of the license plate of the &

And the average brightness value

The vehicle identification and license plate recognizer (1)

Or less; The vehicle detection and license plate recognizer (1)

The vehicle identification and license plate recognizer (1)

Or less, it is determined that the photographed license plate image is a blurry image, and in order to obtain a bright binary image, a reflection ratio

Setting a value to a high value stored in the vehicle identification and license plate recognizer (1); The vehicle detection and license plate recognizer (1)

The vehicle detection and license plate recognizer (1)

It is possible to prevent the unnecessary part of the image from being binarized. In order to prevent the unnecessary part of the image from being binarized,

Setting a value to a low value stored in the vehicle identification and license plate recognizer (1); The vehicle detection and license plate recognizer (1) recognizes the average height of the numbers in the license plate image

And then,

The numerical height threshold value set in the vehicle identification and license plate recognizer (1)

, It is determined that the image of the electric license plate is a small and blurry image, and the local window size value

To a high value stored in the vehicle identification and license plate recognizer (1); The vehicle detection and license plate recognizer (1) recognizes the average height of the numbers in the license plate image

And then,

The numerical height threshold value set in the vehicle identification and license plate recognizer (1)

The local window size value

To a low value stored in the vehicle identification and license plate recognizer (1); The vehicle identification and license plate recognizer (1)

Obtaining a local threshold value T using an equation; And the vehicle detection and license plate recognizer 1 sequentially compares the brightness level of each of the pixels included in the local window region with the local threshold T, and if the brightness level of the specific pixel is greater than the local threshold T, Converting the brightness image of the electric vehicle license plate into a black and white image by converting a specific pixel into a black pixel if the brightness level of the specific pixel is lower than the local threshold value T, and binarizing the brightness image.

는 전기차 번호판의 명도 영상을 흑백 영상으로 변환할 때 명도 영상의 국소 윈도우 영역에 포함된 각 화소들의 명도 레벨과 비교되는 국소 임계치(Local threshold value)이고, 상기

은 도면 4에 도시한 바와 같이, 전기차 번호판의 명도 영상 내 국소 윈도우(Window)의 가로 세로 크기이며, 상기

은

크기의 국소 윈도우 영역 내에 포함된 화소들의 명도 레벨을 평균 낸 값이다.Here,

Is a local threshold value which is compared with the brightness level of each pixel included in the local window region of the brightness image when converting the brightness image of the license plate into the black and white image,

As shown in Fig. 4, the horizontal and vertical sizes of the local window in the brightness image of the license plate,

silver

The average brightness level of the pixels included in the local window region of the size is averaged.

는 국소 임계치

를 계산하기 위해 국소 윈도우 내 각 화소들의 명도 레벨 분포에 대한 반영 비율이고, 상기

는

크기의 국소 윈도우 영역 내에 포함된 화소들의 명도 레벨에 대한 표준 편차이며, 상기

은 상수로서 128로 고정한다.In addition,

Lt; RTI ID = 0.0 >

Is a reflection ratio for the brightness level distribution of each pixel in the local window in order to calculate

The

The standard deviation of the brightness level of the pixels included in the local window region of the size

Is fixed to 128 as a constant.

The step of dividing the candidate region on which the EV logo (A) and the EV character logo (B) are described and extracting the candidate region from the license plate is performed on the license plate of the vehicle detection and license plate recognizer As shown in FIG. 5, the vehicle identification and license plate recognizer 1 recognizes a vehicle number engraved on an electric license plate; The position of the EV logo A and the EV character logo B engraved on the license plate are predicted based on the position information of the left end and the right end of the vehicle number recognized by the vehicle detection and license plate recognizer 1, (A) and an EV character logo (B) of the EV image; And normalizing the size of the EV image logo A and the EV character logo B extracted from the vehicle detection and license plate recognizer 1 to the same size as the number engraved on the EV license plate.

The step of dividing the candidate region on which the EV logo (A) and the EV character logo (B) are described and extracting the candidate region from the license plate is performed on the license plate of the vehicle detection and license plate recognizer And connecting broken pixels using a morphology technique when pixels that are cut off from the license plate are converted into monochrome images.

That is, as shown in FIG. 5, the vehicle detection and license plate recognizer 1 extracts the electric vehicle image logo A located on the left side of the vehicle number and the EV character logo B located on the right side of the vehicle number, It can be determined whether the vehicle is a fuel car or an electric car.

In order to recognize the electric vehicle image logo A and the EV character logo B by using the artificial neural network, the vehicle detection and license plate recognizer 1 first recognizes the electric vehicle image logo A and the EV character logo B And normalizes the size of the EV image logo (A) and the EV character logo (B).

Next, the size-normalized electric vehicle image logo (A) and the EV character logo (B) are trained by adding to the numerical recognition neural network as shown in FIG.

That is, the number of output neurons is 0 to 9, the electric vehicle image logo (A), the EV character (A), and the EV character image The logo B is added to 12, and the electric vehicle image logo A and the EV character logo B are recognized as two increased neurons.

That is, the conventional artificial neural network for numerical recognition is modified to learn 12 input patterns as shown in FIG.

After learning 12 input patterns for numerical recognition artificial neural network, we confirmed and analyzed the reliability of artificial neural network for arbitrary input samples. As a result, we have reliability of 95% or more for 12 input patterns, , It is possible to accurately recognize not only the numbers engraved on the license plate but also the EV image logo (A) and the EV character logo (B).

By expanding the artificial neural network for numerical recognition, artificial neural network for numerical recognition is learned so as to distinguish 12 types of numbers and logos including an electric vehicle image logo (A) and an EV character logo (B), thereby using a high performance artificial neural network So that the electric vehicle image logo (A) and the EV character logo (B) can be recognized.

Therefore, instead of introducing a separate independent decision classifier for recognizing the electric vehicle image logo (A) or the EV character logo (B), the output neuron is increased to the conventional numerical recognition artificial neural network, Whether it is a fuel car or an electric car.

The vehicle detection and license plate recognizer 1 extracts candidate regions extracted from the license plate and inputs the classified candidate regions into an artificial neural network for recognizing and classifying the electric vehicle image logo A and the EV character logo B, (A) and the EV character logo (B) are included in the candidate region, and the artificial neural network classifies whether the candidate region is the electric vehicle image logo (A) or the EV character logo (B) (A) or EV (B) extracted from an electric vehicle license plate converted into a black and white image is used as a feature extraction method for an electric vehicle image logo (A) and an EV character logo (B) in the artificial neural network. A mesh feature extraction method of dividing a character logo (B) into a grid shape and calculating a distribution ratio of black pixels; Distance to calculate the horizontal and vertical distances from the rectangle circumscribing the EV image logo (B) or the EV image logo (A) extracted from the plate to the stroke of the EV logo (B) Extraction method, a projection extraction method for calculating the distribution of black pixels when black pixels are accumulated in a horizontal direction or a vertical direction with respect to an electric vehicle image logo (A) or an EV character logo (B) extracted from an electric vehicle license plate converted into a monochrome image And a line crossing the electric motor vehicle image logo A or the EV character logo B extracted from the electric vehicle license plate converted into the monochrome image in the horizontal direction or the vertical direction is the electric motor image logo A or the EV character logo B A crossing feature extraction method may be used to calculate the number of times a stroke is encountered.

The method for recognizing and identifying the vehicle number and the electric vehicle of the vehicle that the vehicle detecting and license plate recognizer 1 exits from the highway recognizes and identifies the vehicle number or the electric vehicle of the vehicle passing through the highway or the gantry The detailed description is omitted.

The vehicle detection and license plate recognizer (1) mounted on the smart tolling system comprises a vehicle detection control device (3) for detecting a vehicle passing through a gantry, a camera using the vehicle when the vehicle is detected from the vehicle detection control device And an image processing control device 5 for capturing a forward or backward image of the detected vehicle, extracting a license plate area from the photographed image, and recognizing and identifying the vehicle number and the electric vehicle from the extracted license plate area.

The fare adjuster of the highway operated by the toll charge system or the high pass system that clears the highway toll charge at the toll gate booth can calculate the toll of the outbound vehicle on the highway to the vehicle number stored in the fare settling server 7, The highway toll gate name of the highway toll gate, and the highway toll gate name of the highway to the departing vehicle, and the fare adjuster of the highway operated by the smart tolling system transmits the toll to the fare settling server 7 (S6), which is calculated by referring to the stored vehicle number, the presence or absence of the electric vehicle, the gantry passing path, the passage of the highway linking section, and the gantry passing time, and then charging the departing vehicle, Smart talling to detect vehicles entering the highway gantry as a procedure The system in-vehicle detection device 3 (VDU: Vehicle Detecting Unit) requests the charge settlement server 7 to make a socket connection, the charge settlement server 7 receives the socket connection, 3) and a fare settlement server (7) socket connection (S6-1); The charge settlement server 7 confirms the current time information and then transmits the current time information to the vehicle detection control device 3 and the vehicle detection control device 3 stores the received current time information S6- 2); The vehicle detection control device 3 that detects the vehicle entering the highway gantry by the highway toll billing procedure transmits a trigger signal that is a vehicle detection signal to the image processing control device 5 and a unique number of the trigger signal Next, a step (S6-3) of transmitting the position of the detected vehicle, the identifier (ID) of the vehicle detection control device 3, and the time when the detected vehicle passes the gantry to the image processing control device 5; The image processing control device 5 receives the trigger signal transmitted from the vehicle detection control device 3, the unique number of the trigger signal, the vehicle position of the detected vehicle, the identifier (ID) of the vehicle detection control device 3, (S6-4) of transmitting the time that the detected vehicle passes through the gantry to the fare adjustment server 7; The vehicle control unit 3 receives the trigger signal, the unique number of the trigger signal, the vehicle position of the detected vehicle, the identifier (ID) of the vehicle detection control device, and the time when the detected vehicle passes the gantry The processing control device 5 extracts the license plate area from the forward image or the backward image of the gantry passing vehicle photographed through the camera, identifies and identifies the vehicle number and the electric vehicle in the extracted license plate area, (S6-5) whether the electric car is transferred to the fare adjustment server (7); And the fare settling server 7 calculates the fare of the expressway by referring to the vehicle number transmitted from the image processing control device 5, whether the vehicle is an electric car, whether the gantry passage is passing through the highway linkage section, And a step S6-6 of charging the borrower of the car number.

The fare adjuster of the highway operated by the toll charge system or the high pass system that clears the highway toll charge at the toll gate booth can calculate the toll of the outbound vehicle on the highway to the vehicle number stored in the fare settling server 7, The highway toll gate name of the highway toll gate, and the highway toll gate name of the highway to the departing vehicle, and the fare adjuster of the highway operated by the smart tolling system transmits the toll to the fare settling server 7 (Step S6), in which the image processing control device 5 determines whether or not the image processing control device 5 has received the vehicle number, the electric vehicle number, the gantry passage, the highway linkage period, and the gantry passage time, Passing vehicle captured through the gantry-passing vehicle is transmitted to the video server 9 (S6-7). ≪ / RTI >

The method for collecting tolls on an expressway of an electric vehicle according to the present invention, which is made up of such a procedure, is a method for collecting tolls of an electric car on a freeway tollgate or gantry on a highway operated by a toll charge system, high pass system or smart toll system, And the fare adjuster can charge the discounted highway fare to the electric car owner by identifying the highway route and highway linkage zone by car number.

Accordingly, the highway toll collection method of an electric vehicle according to the present invention can automatically recognize the electric vehicle without confirming visually whether the storage staff in the toll gate charge booth is the electric vehicle of the vehicle coming out of the toll gate, and charge the automatically calculated highway toll to the electric vehicle driver can do.

Further, according to the present invention, an electric car passing through a high-pass section and a multi-car smart tolling system can pay a highway toll without a high-pass terminal or a high-pass terminal without installing a high-

1. Vehicle detection and license plate recognition device 3. Vehicle detection control device
5. Image processing control device 7. Fee settlement server
9. Video server

Claims (4)

Tollgate tolls are settled on the toll gate tollgate in the tollgate booth. Vehicle detection and installation on the gantry of the expressway tolls are settled by highway toll or highway toll system where highway tolls are settled by highway system. (S1) photographing a forward image or a backward image of a vehicle passing through a gantry or a vehicle entering a highway toll gate entrance using a camera using the license plate recognizer (1);
(S2) extracting a license plate area from a forward or backward image of a vehicle passing through a vehicle or a gantry where the vehicle detecting and license plate recognizer (1) enters a highway toll gate entrance;
(S3) judging whether an electric vehicle image logo (A) and an EV character logo (B) unique to the electric vehicle are written in the vehicle license plate area from which the vehicle detection and license plate recognizer (1) is extracted;
The vehicle identification and license plate recognizer 1 judges that the electric vehicle is an electric car if the logo of the electric car image A and the logo of the EV character B are described in the extracted license plate area and the current highway charge settlement method is a toll charge settlement System, or a high-pass system, it stores the vehicle number detected in the fare settling server 7, the presence of an electric car, the time of the highway entrance, and the name of the toll gate of the highway. The vehicle number of the vehicle that has passed through the gantry, the presence or absence of the electric vehicle, the gantry passage time, the lane of the vehicle when passing through the gantry, and the vehicle detection and license plate recognizer mounted on the gantry, (S4) of storing an identifier (ID) of the terminal (1);
A vehicle detection and license plate recognizer (1) installed on a toll gate of a highway operated by a toll charge settlement system or a high pass system that clears a toll charge toll gate at a toll gate fee booth is a forward image of a vehicle After the rear image is photographed, the vehicle license plate area is extracted from the forward image or the backward image of the photographed vehicle, and it is checked whether the electric vehicle image logo (A) or the EV character logo (B) If the logo of the electric car image (A) or the logo of the EV character (B) is written in the license plate area, it is determined that the vehicle is an electric car and the vehicle number detected by the charge settlement server (7) Storing the name of the toll gate on the highway (S5);
And a toll charge system of a highway operated by a toll charge system for setting up toll charge on a toll gate fee booth and a highway system operated by a high pass system compares the toll on the highway toll vehicle with the vehicle number stored in the toll charge server 7, The toll account of the expressway tollgate, the tollgate name of the tollgate on the highway, and the tollgate name of the highway to which the vehicle has departed, and then bill the vehicle to the departing vehicle. The fare adjuster of the highway operated by the smart tolling system, (S6) of charging the vehicle with reference to the number of the vehicle, whether or not the electric vehicle is present, the gantry passage, the passage of the highway linkage, and the gantry passage time.
The method according to claim 1,
Step S3 of determining whether the electric vehicle image logo A and the EV character logo B specific to the electric vehicle are described in the vehicle license plate area from which the vehicle detection and license plate recognizer 1 is extracted
Converting the license plate into a monochrome image by the vehicle identification and license plate recognizer (1)
A step of extracting a candidate region in which a candidate region on which an EV logo (A) and an EV character logo (B) are described is divided and divided into candidate regions from a license plate on a license plate on which the vehicle identification and license plate recognizer (1)
And the vehicle detecting and license plate recognizing device (1) inputs the classified candidate regions extracted from the license plate into an artificial neural network for recognizing and classifying the electric vehicle image logo (A) and the EV character logo (B) (A) and an EV character logo (B) are included in the region, and classifying whether the candidate region is an electric vehicle image logo (A) or an EV character logo (B) Highway toll collection method.
3. The method of claim 2,
The step of converting the vehicle license plate into a monochrome image by the vehicle detection and license plate recognizer (1)
The average value of brightness of all the pixels of the electric car number plate in the state of the grayscale image in which the vehicle detection and license plate recognizer 1 is extracted

And the average brightness value

The vehicle identification and license plate recognizer (1)

Or less;
The vehicle detection and license plate recognizer (1)

The vehicle identification and license plate recognizer (1)

Or less, it is determined that the photographed license plate image is a blurry image, and in order to obtain a bright binary image, a reflection ratio

Setting a value to a high value stored in the vehicle identification and license plate recognizer (1);
The vehicle detection and license plate recognizer (1)

The vehicle detection and license plate recognizer (1)

It is possible to prevent the unnecessary part of the image from being binarized. In order to prevent the unnecessary part of the image from being binarized,

Setting a value to a low value stored in the vehicle identification and license plate recognizer (1);
The vehicle detection and license plate recognizer (1) recognizes the average height of the numbers in the license plate image

And then,

The numerical height threshold value set in the vehicle identification and license plate recognizer (1)

, It is determined that the image of the electric license plate is a small and blurry image, and the local window size value

To a high value stored in the vehicle identification and license plate recognizer (1);
The vehicle detection and license plate recognizer (1) recognizes the average height of the numbers in the license plate image

And then,

The numerical height threshold value set in the vehicle identification and license plate recognizer (1)

The local window size value

To a low value stored in the vehicle identification and license plate recognizer (1);
The vehicle identification and license plate recognizer (1)

Obtaining a local threshold value T using an equation;
And the vehicle detection and license plate recognizer 1 sequentially compares the brightness level of each of the pixels included in the local window region with the local threshold T, and if the brightness level of the specific pixel is greater than the local threshold T, Converting the brightness image of the license plate into a black and white image and converting the brightness image into a black and white image by converting a specific pixel into a black pixel if the brightness level of the specific pixel is lower than the local threshold value T. [

Here,

Is a local threshold value that is compared with the brightness level of each pixel included in the local window region of the brightness image when converting the brightness image of the license plate into the black and white image,
remind

The size of the local window in the brightness image of the electric car license plate,
remind

silver

A value obtained by averaging the brightness level of the pixels included in the local window region of the size,
remind

Lt; RTI ID = 0.0 >

The reflectance ratio for the brightness level distribution of each pixel in the local window,
remind

The

The standard deviation of the brightness level of the pixels included in the local window region of the size,
And

128 as a constant
The method according to claim 1,
The vehicle detection and license plate recognizer (1) mounted on the smart tolling system includes a vehicle detection control device (3) for detecting a vehicle passing the gantry,
When a vehicle is detected from the vehicle detection control device 3, a front image or a rear image of the detected vehicle is photographed using the camera, and then a license plate area is extracted from the photographed image. And an image processing control device (5)
The toll charge system for the highway toll charge system in the toll gate charge booth or the highway toll system operated by the high pass system stores the toll on the highway toll vehicle in the toll gate system, The highway tollgate name and the highway toll gate name of the highway to the departing vehicle, and the fare adjuster of the highway operated by the smart tolling system stores the toll of the highway toll vehicle in the fare settling server 7 (S6) of charging the vehicle to the departing vehicle based on the reference to the vehicle number, whether or not the electric vehicle is present, the gantry passage, the passage of the highway linkage, and the gantry passage time
The vehicle detection control unit 3 (VDU: Vehicle Detecting Unit) in the smart tolling system that detects the vehicle entering the highway gantry in the initialization procedure requests a socket connection to the fare settling server 7 The charge settlement server 7 accepts socket connection and socket connection of the vehicle detection control device 3 and the charge settlement server 7 (S6-1);
The charge settlement server 7 confirms the current time information and then transmits the current time information to the vehicle detection control device 3 and the vehicle detection control device 3 stores the received current time information S6- 2);
The vehicle detection control device 3 that detects the vehicle entering the highway gantry by the highway toll billing procedure transmits a trigger signal that is a vehicle detection signal to the image processing control device 5 and a unique number of the trigger signal Next, a step (S6-3) of transmitting the position of the detected vehicle, the identifier (ID) of the vehicle detection control device 3, and the time when the detected vehicle passes the gantry to the image processing control device 5, ;
The image processing control device 5 receives the trigger signal transmitted from the vehicle detection control device 3, the unique number of the trigger signal, the vehicle position of the detected vehicle, the identifier (ID) of the vehicle detection control device 3, (S6-4) of transmitting the time that the detected vehicle passes through the gantry to the fare adjustment server 7;
The trigger signal, the unique number of the trigger signal, the position of the detected vehicle, the identifier (ID) of the vehicle detection control device 3, and the time at which the detected vehicle passes the gantry are calculated from the vehicle detection control device 3 The transmitted image processing control device 5 extracts the license plate area from the forward image or the backward image of the gantry passing vehicle photographed through the camera, identifies and identifies the vehicle number and the electric vehicle in the extracted license plate area, (S6-5) of transmitting the vehicle number and the electric vehicle to the fare adjustment server 7;
And the fare settling server 7 calculates the fare of the expressway by referring to the vehicle number transmitted from the image processing control device 5, whether the vehicle is an electric car, whether the gantry passage is passing through the highway linkage section, And a step (S6-6) of charging the owner of the car number to the owner of the car number.

Images