KR102465167B1 - Vehicle number recognition system for increasing recognition rate of vehicle number using rear license plate - Google Patents

Abstract

The present invention relates to a vehicle license plate recognition system using a rear license plate, which is configured to recognize a vehicle license plate by utilizing a rear license plate of a vehicle to be enforced when detecting the vehicle to be enforced, thereby accurately recognizing the vehicle license plates even for vehicles such as a vehicle with a damaged front license plate or a two-wheeled vehicle with a rear license plate attached to improve service reliability and an enforcement rate.

Description

Vehicle number recognition system for increasing recognition rate of vehicle number using rear license plate

The present invention relates to a vehicle number recognition system using a rear license plate with an increased number recognition rate, and more specifically, when detecting an intercepted vehicle, it is configured to recognize the vehicle number using the rear license plate of the intercepted vehicle, so that the front license plate is damaged or the rear It relates to a vehicle number recognition system using a rear license plate that can recognize a number for a two-wheeled vehicle having only a license plate, thereby not only increasing service reliability and accuracy, but also maximizing the accuracy and precision of vehicle detection.

Recently, with the development of information and communication technology, the spread of telematics devices and the establishment of a ubiquitous environment, an intelligent transportation system (ITS) has been studied and widely used.

In general, an intelligent transportation system (ITS) primarily detects a vehicle and generates vehicle information, then analyzes and processes the generated vehicle information to generate traffic information and at the same time generate whether or not an unexpected situation occurs. The performance of the Intelligent Transportation System (ITS) is determined according to how accurately the vehicle number is recognized because it is configured to recognize the vehicle number of

However, in the prior art, it was mainly configured to recognize the vehicle number based on the front license plate by photographing the front license plate of the vehicle. It has limitations.

Accordingly, in order to improve the license plate recognition rate of the vehicle, various studies are being conducted on a technique for using the image of the rear license plate together with the image of the front license plate of the vehicle.

1 is a block diagram illustrating a multi-lane-based vehicle photographing apparatus disclosed in Korean Patent Registration No. 10-1584105 (Title of the Invention: Multi-lane-based vehicle photographing apparatus and control method thereof).

The multi-lane-based vehicle imaging device (hereinafter referred to as the prior art) 100 of FIG. 1 detects the entry state and location of a vehicle entering by a plurality of laser sensors installed for each lane and lane on the entry side of the first gantry G1. The passing state of the vehicle passing through the first vehicle detection unit 110 that detects and determines whether the vehicle enters the center of the lane or along the lane, and a plurality of laser sensors installed for each lane on the exit side of the first gantry G1 A plurality of lasers installed for each lane on the second vehicle detection unit 120 for detecting the position of the vehicle and outputting a front photographing signal, and the second gantry G2 located at the rear of the first gantry G1 in the traveling direction of the vehicle. The third vehicle detection unit 130 for detecting the passing state and position of the vehicle passing through the sensor and outputting a rear photographing signal, the front photographing signal output from the second vehicle detection unit 120 and the third vehicle detection unit ( 130), the front camera 152 and the rear camera 154 are operated according to the outputted rear camera signal to control matching data for matching the photographed front photographed image 152 and the rear photographed image 154 to a lane to be described later. The entry state of the vehicle inputted from the photographing unit 150 and the first to third vehicle detection units, which transmits the front photographed image and the rear photographed image to the image server 160 according to the matching data requested by 140 and provided , output the front shooting signal and the rear shooting signal based on the passing state and location, predict whether the vehicle has changed lanes and the driving speed of the vehicle, and output the front shooting signal according to the passing state of the second vehicle detection unit 120, and According to the passing state of the third vehicle detection unit 130, a rear photographing signal is output, and when the vehicle is traveling on a lane, the photographing signal is simultaneously output so that all cameras installed in adjacent lanes can be operated, and the front photographed image and the rear It consists of a vehicle control unit 140 that generates matching data for matching the captured image and provides it to the photographing unit 150 .

The prior art 100 configured as described above accurately matches the front and rear photographed images when photographing the vehicle through the front camera 152 and the rear camera 154, thereby recognizing the vehicle number and the influence of the lighting state or the surrounding environment. It has the advantage of not only increasing the recognition rate by minimizing the

However, the prior art 100 has disadvantages in that the installation work is complicated and the operation is complicated because a plurality of laser sensors and front and rear cameras must be installed in a gantry spaced apart in the front and rear directions in order to implement the system.

In addition, the prior art 100 is configured to generate matching data using the business office number, lane number, processing serial number, shooting time, and recognized vehicle number as a medium when generating matching data of the front photographed image and the rear photographed image. While this is dense, if an obstacle occurs in the recognition of the front license plate or the rear license plate, the image matching rate is significantly lowered.

That is, when vehicles are adjacent to each other in the same lane, they must be accurately identified, and a valid number plate must be selected from among the front and rear license plates for each of the adjacent vehicles to perform number recognition. In order to accurately identify these adjacent vehicles, detection The resolution of the means should be improved, but the prior art 100 does not describe a technology for increasing the resolution of the detection means and a technology for selecting a valid license plate when detecting an adjacent vehicle. It has the disadvantage of lowering the efficiency.

The present invention is to solve this problem, and the present invention is configured to recognize the vehicle number by using the rear license plate of the intercepted vehicle when detecting the intercepted vehicle, so that the front license plate is damaged or the rear license plate is attached to a two-wheeled vehicle, etc. The purpose of this is to provide a vehicle number recognition system using the rear license plate that can accurately recognize the vehicle number for the vehicle of

In addition, another solution of the present invention is to match the front and rear images using the vehicle location as a medium in normal times when the license plate recognition of the intercept vehicle is normally performed, but in an unexpected situation where the license plate of the intercept vehicle is not recognized, the width of the intercept vehicle object Based on the first judgment whether the enforcement vehicle is a two-wheeled vehicle with only the rear license plate attached, the front license plate is damaged or only the rear license plate is attached by matching the image of the unrecognized front license plate with the image of the object of the enforcement vehicle. This is to provide a vehicle number recognition system using a rear license plate that can accurately recognize vehicle numbers for two-wheeled vehicles as well as increase image matching rate.

The solution means of the present invention for solving the above problems is a sensing means for detecting a vehicle; a controller that analyzes the detection signal detected by the detection means and determines whether the detection vehicle is an intermittent vehicle; a rear camera for photographing the rear of the intermittent vehicle to obtain a rear image according to the control of the controller; and a front camera for capturing a front image by photographing the front of the intercepted vehicle under the control of the controller, wherein the controller analyzes the detection signal detected by the detection means to determine the lane, position and speed ( After calculating V), a vehicle information generating unit that generates vehicle information including a lane, a position, and a speed (V) of the detected vehicle; By comparing the speed (V) information of the vehicle information generated by the vehicle information generating unit with the speed limit (V'), a detection vehicle having a speed (V) exceeding the speed limit (V') is determined as the intermittent vehicle an enforcement vehicle determination unit; a trigger signal generating unit for outputting a trigger signal including lane information to the rear camera and the front camera when the intercept vehicle is detected by the interception vehicle determining unit; a vehicle number recognition unit for recognizing a vehicle number by analyzing the rear image and the front image obtained by the photographing of the rear camera and the front camera; It is executed when at least one of the front/rear vehicle numbers is recognized by the vehicle number recognition unit, and if the recognized front vehicle number and the rear vehicle number are the same, it is determined that the intermittent vehicle is a four-wheeled vehicle and the same front/rear vehicle number The vehicle number is determined as the vehicle number of the intercepted vehicle, and if the front vehicle number is omitted from the front/rear vehicle number, the vehicle width (D) of the intercepted vehicle is calculated by analyzing the front image or rear image, and then the calculated vehicle width (D) ) is less than the threshold, a number recognition verification unit for determining the enforcement vehicle as a two-wheeled vehicle and at the same time determining the rear vehicle number as the vehicle number of the enforcement vehicle; A vehicle number input module that receives the vehicle number recognized by the number recognition verification unit by matching the vehicle number and the image determined by the number recognition verification unit to a control information generation unit for generating enforcement information, the number recognition verification unit receiving the vehicle number recognized by the vehicle number recognition unit ; an omission number determination module for determining whether a missing number exists among the front/rear vehicle numbers input through the vehicle number input module; a vehicle width calculation module that is executed when the missing number determination module determines that there is a missing number, and calculates a vehicle width (D) of a detected vehicle object by analyzing a front image or a rear image photographed of the detected vehicle; The vehicle width D calculated by the vehicle width calculation module is compared with a preset threshold, and if the vehicle width D is greater than or equal to the threshold, the detection vehicle is determined to be a four-wheeled vehicle, but if the vehicle width D is less than the threshold, the detection a two-wheeled vehicle determination module for determining that the vehicle is a two-wheeled vehicle; It is executed when the detection vehicle is determined to be a four-wheeled vehicle by the two-wheeled vehicle determination module, and among the frames of the front image and the rear image obtained by photographing the front photographing camera and the rear photographing camera, in the vehicle number recognition unit After extracting other unused frames, by analyzing the extracted front image frame and rear image frame, the vehicle number re-recognition module for re-recognizing the front vehicle number and the rear vehicle number; an identical determination module for determining whether the front vehicle number and the rear vehicle number re-recognized by the vehicle number re-recognition module are the same; It is executed when the missing number determination module determines that there is no missing number among the front/rear vehicle numbers, and compares the front vehicle number and the rear vehicle number input through the vehicle number input module, and the front vehicle a matching module that determines whether the number and the rear vehicle number match, and if it is determined that the front vehicle number and the rear vehicle number do not match, executes the vehicle number re-recognition module; a failure confirmation module that is executed when it is determined that the front/rear vehicle number is not the same in the sameness determination module, and finally determines that a failure has occurred in vehicle number recognition; 1) The detection vehicle is determined to be a two-wheeled vehicle by the two-wheeled vehicle determination module, 2) The front vehicle number and the rear vehicle number are determined to match in the matching module, or 3) The front/rear vehicle number is determined by the identicalness determination module Executed when it is determined that the rear vehicle number is the same, 1) When the detection vehicle is determined to be a two-wheeled vehicle in the two-wheeled vehicle determination module and executed, the rear vehicle number is finally determined as the vehicle number of the corresponding enforcement vehicle (two-wheeled vehicle); , 2) When it is determined that the front vehicle number and the rear vehicle number match in the matching module, the matching front/rear vehicle number is finally determined as the vehicle number of the intercepted vehicle, and 3) the same determination is made. When it is determined that the front/rear vehicle number is the same in the module and executed, the same front/rear vehicle number is included in the vehicle number determination module to finally determine the vehicle number of the corresponding intercepted vehicle.

delete

delete

delete

In addition, in the present invention, it is preferable that the sensing means is a radar device that receives a reflected signal after transmitting a radar signal toward the front and rear of the installed position.

In addition, in the present invention, the sensing means is a LiDAR scanner that receives a reflected signal after transmitting a LiDAR signal toward the front and rear of the installation position, and the LiDAR scanner includes a polygon mirror forming each side, and the A driving unit that rotates the polygon mirror, and a radar sensor that emits a radar signal toward the rotating polygon mirror and receives a reflected signal, wherein the controller has an optimal sensitivity value matched according to the range of the noise generation rate. a memory in which a reference table, which is data, is preset and stored; a lidar scanner sensitivity optimization unit that is executed at a preset period (T) and detects an optimal sensitivity value of the lidar scanner; a lidar scanner sensitivity control unit for controlling the lidar scanner to operate the lidar scanner according to the optimal sensitivity value detected by the lidar scanner sensitivity optimization unit; a lidar signal analysis unit that analyzes the lidar signal transmitted and received by the lidar sensor of the lidar scanner; a spectrogram modeling unit that utilizes the analysis data detected by the lidar signal analyzer to generate a spectrogram for the lidar signal transmitted and received by the lidar scanner; a profile analyzer for detecting and analyzing profiles of the spectrogram generated by the spectrogram modeling unit; Comparing the profiles detected by the profile analysis unit with preset filtering conditions, detecting a vehicle object, and further comprising a vehicle object detection unit that determines a profile that has passed all the filtering conditions among the detected profiles as a vehicle profile, , the vehicle information generating unit is executed when a vehicle is detected by the vehicle object sensing unit, and generates vehicle information by analyzing the trajectory of the detected vehicle, and the lidar scanner sensitivity optimization unit is a spectrogram during the period (T). a data collection module for collecting them; a sample spectrogram extraction module for extracting sample spectrograms that are spectrograms when a vehicle object is detected from the spectrograms collected by the data collection module; a profile analysis module for analyzing profiles of the sample spectrograms extracted by the sample spectrogram extraction module; In the sample spectrogram extracted by the sample spectrogram extraction module, it is analyzed whether a profile other than a vehicle exists in a preset adjacent region from the vehicle profile, and when a profile other than a vehicle exists in a preset adjacent region, the corresponding profile a noise profile detection module that determines that α is a noise profile; After calculating the sample quantity (L), which is the quantity of sample spectrograms extracted by the sample spectrogram extraction module, the noise which is the quantity of noise spectrograms that are sample spectrograms including the noise profile detected by the noise profile detection module a noise generation ratio calculation module for calculating the quantity (L') and calculating the noise generation ratio (R, R=L'/L) by dividing the calculated noise quantity (L') by the calculated sample quantity (L); Preferably, the method further comprises a sensitivity optimum value calculation module for detecting an optimum sensitivity value according to the noise generation ratio R calculated by the noise generation ratio calculation module by utilizing the reference table stored in the memory.

In addition, in the present invention, it is preferable that the lidar scanner transmits and receives a lidar signal through 16 channels, and the controller detects a vehicle by analyzing the lidar signal for each channel.

According to the present invention having the above problems and solutions, when detecting an interception vehicle, it is configured to recognize the vehicle number by using the rear license plate of the intercept vehicle, so that the vehicle number of the vehicle such as a two-wheeled vehicle to which a front license plate is damaged or a rear license plate is attached By accurately recognizing , it is possible to increase service reliability and enforcement rate.

In addition, according to the present invention, in normal times, when the license plate recognition of the intercept vehicle is normally performed, the front and rear images are matched using the vehicle location as a medium, but in an unexpected situation in which the license plate of the intercept vehicle is not recognized, it is based on the width of the intercept vehicle object By first determining whether the enforcement vehicle is a two-wheeled vehicle with only a rear license plate attached, the front license plate is damaged or the two-wheeled vehicle with only the rear license plate attached by matching the image of the unrecognized front license plate with the image of the relevant enforcement vehicle object. It can not only accurately recognize the vehicle number, but also increase the video matching rate.

1 is a block diagram illustrating a multi-lane-based vehicle photographing apparatus disclosed in Korean Patent Registration No. 10-1584105 (Title of the Invention: Multi-lane-based vehicle photographing apparatus and control method thereof).
2 is a configuration diagram showing a vehicle number recognition system according to an embodiment of the present invention.
3 is an exemplary view showing FIG. 2 .
4 is a block diagram illustrating the controller of FIG. 2 .
5 is a block diagram illustrating a second vehicle number recognition system, which is a second embodiment of the present invention.
6 is an exemplary view of FIG. 5 .
7 (a) is an exemplary view for explaining a license plate obtained when the front camera and rear camera of FIG. It is an exemplary diagram for explaining the license plate obtained when shooting a two-wheeled vehicle.
8 is a block diagram illustrating a second controller of FIG. 5 .
9 is a block diagram illustrating the number recognition verification unit of FIG. 8 .
10 is a configuration diagram illustrating a third vehicle number recognition system, which is a third embodiment of the present invention.
11 is a block diagram illustrating a third controller of FIG. 10 .
12 is a block diagram illustrating the lidar scanner sensitivity optimization unit of FIG. 11 .

Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings.

2 is a configuration diagram illustrating a vehicle number recognition system according to an embodiment of the present invention, and FIG. 3 is an exemplary view showing FIG. 2 .

2 and 3, the vehicle number recognition system 1, which is an embodiment of the present invention, is configured to recognize the vehicle number by using the rear license plate of the intercepted vehicle when detecting an intercepted vehicle. This is to maximize the accuracy and precision of vehicle detection as well as increase service reliability and accuracy by allowing number recognition.

In addition, as shown in FIGS. 2 and 3, the vehicle number recognition system 1 of the present invention is installed in the first gantry G1 and transmits a radar signal toward the road R, and then collects the reflected signal. a radar device 7, a rear camera 6 for acquiring a rear image by photographing the rear of the detected interception vehicle C when an interception vehicle is detected by a controller 3 to be described later, and a radar device 7 After determining whether the detection vehicle is an intercept vehicle by analyzing the radar signal transmitted and received by The controller 3 that analyzes and recognizes the license plate number of the enforcement vehicle, and the traffic control center server 9 that is connected to the controller 3 and the communication network 10 and receives enforcement information from the controller 3 and performs follow-up actions and a communication network 10 that provides a data movement path between the controller 3 and the traffic control center server 9 .

At this time, in the present invention, for convenience of explanation, the camera 6 for rear view has been described as an example to be installed on the first gantry G1, but the structure for fixing the camera 6 for rear view is not limited to the gantry, It may be configured to be installed in a variety of known structures.

The communication network 10 supports data communication between the controller 3 and the traffic control center server 9, and in detail, consists of a wired/wireless network such as a wide area network (WAN), 3G, 4G, 5G, LTE, etc. can be

The traffic control center server 9 stores and monitors the traffic information, enforcement information, and the image of the enforcement vehicle received from the controller 3, and when receiving the enforcement information, performs subsequent processing such as issuing a fine for the enforcement vehicle.

4 is a block diagram illustrating the controller of FIG. 2 .

The controller 3 of FIG. 4 manages and controls the operations of the rear camera 6 and the radar device 7, and analyzes the rear image obtained by the rear camera 6 when detecting the intercept vehicle. After recognizing the vehicle number of (C), after generating enforcement information, the generated enforcement information is transmitted to the traffic control center server 9 .

4, the controller 3 includes a control unit 30, a memory 31, a communication interface unit 32, a radar signal analysis unit 34, a vehicle object detection unit 37, and vehicle information. Generation unit 38, control whether or not determination unit 39, trigger signal generation unit 40, rear image input unit 41, rear image analysis unit 42, vehicle number recognition unit 43, enforcement information generation unit ( 45).

The control unit 30 is an operating system (OS) of the controller 3, and the control target 31, 32, 33, 34, 35, 36, 37, 38 , (39), (40), (41), (42), (43), (45), and (46) manage and control the operations.

In addition, the control unit 30 inputs the transmitted and received radar signal information output from the radar device 7 through the communication interface unit 32 to the radar signal analysis unit 34 .

In addition, when vehicle information is generated by the vehicle information generating unit 38 , the control unit 30 controls the communication interface unit 32 to transmit the generated vehicle information to the traffic control center server 9 .

In addition, the control unit 30 executes the trigger signal generating unit 40 when the detection vehicle is determined to be an intermittent vehicle by the interception determination unit 39 .

In addition, when a trigger signal is generated by the trigger signal generating unit 40 , the control unit 30 controls the communication interface unit 32 to output the generated trigger signal to the rear photographing camera 6 .

In addition, the control unit 30 inputs the rear image obtained by the photographing of the rear photographing camera 6 to the rear image input unit 41 .

In addition, when the enforcement information is generated by the enforcement information generating unit 45 , the control unit 30 controls the communication interface unit 32 so that the generated enforcement information is transmitted to the traffic control center server 9 .

In addition, the control unit 30 executes the lidar scanner sensitivity optimization unit 46 every preset period (T).

The memory 31 temporarily stores the rear image obtained by the photographing of the rear photographing camera 6 .

In addition, the vehicle information generated by the vehicle information generating unit 38 is stored in the memory 31 .

In addition, the intermittent information generated by the intermittent information generating unit 45 is stored in the memory 31 .

In addition, the memory 31 stores communication identification information and installation location information of the camera 6 and the radar device 7 for rear imaging.

Also, pixel information corresponding to each lane on the rear image is preset and stored in the memory 31 .

In addition, the detection area of the radar device 7 is preset and stored in the memory 31 .

The communication interface unit 32 transmits and receives data to and from the traffic control center server 9 .

In addition, the communication interface unit 32 inputs and outputs data with the rear camera 6 and the radar device 7 .

The radar signal analysis unit 34 analyzes the radar signal transmitted and received by the radar device 7 .

The vehicle object detection unit 37 detects the vehicle object by comparing the detected profiles with preset filtering conditions by using the analysis data detected by the radar signal analysis unit 34 .

The vehicle information generating unit 38 is executed when a vehicle is detected by the vehicle object detecting unit 37, and after analyzing the trajectory of the detected vehicle, calculating the lane, position and speed V, the calculated detection vehicle The vehicle information is generated by matching the lane, location, and speed (V) information of

At this time, the vehicle information generated by the vehicle information generation unit 38 is transmitted to the traffic control center server 9 through the communication interface unit 32 under the control of the control unit 30 .

The intermittent vehicle determination unit 39 extracts the speed (V) information of the vehicle information generated by the vehicle information generation unit 38, and compares the extracted speed (V) with a preset speed limit (V'), If the speed (V) of the detected vehicle is below the speed limit (V'), it is determined that the detected vehicle is a general vehicle that does not violate the speed limit, and if the speed (V) of the detected vehicle exceeds the speed limit (V') Then, it is determined that the detection vehicle is an enforcement vehicle that violates the speed limit.

At this time, when the detection vehicle is determined as the intermittent vehicle by the intermittent vehicle determining unit 39 , the control unit 30 executes the trigger signal generating unit 39 .

The trigger signal generating unit 40 is executed when the intermittent vehicle determination unit 39 determines that the detected vehicle is an intermittent vehicle, and extracts vehicle information of the intermittent vehicle.

In addition, the trigger signal generating unit 40 generates a trigger signal including lane information of the intermittent vehicle by referring to and utilizing the vehicle information of the intermittent vehicle.

At this time, the control unit 30 controls the communication interface unit 32 so that the trigger signal generated by the trigger signal generating unit 40 is output to the rear camera 6, respectively.

On the other hand, when the rear camera 6 receives a trigger signal from the controller 3, it acquires a rear image by photographing a lane included in the input trigger signal, and then outputs the acquired rear image to the controller 3 .

The rear image input unit 41 receives a rear image obtained by photographing of the rear photographing camera 6 .

The rear image analysis unit 42 analyzes the rear image input through the rear image input unit 41, and in detail, after noise removal and pre-processing of the input image, edge detection or image improvement information is used or , using a color and aspect ratio, or using a Hough transform may be configured to detect a license plate.

At this time, the rear license plate image detected by the rear image analysis unit 42 is input to the vehicle number recognition unit 43 .

The license plate recognition unit 43 analyzes the input rear license plate image using a preset number recognition algorithm, and recognizes the license plate number of the intercepted vehicle (C).

The enforcement information generating unit 45 generates the enforcement information including the vehicle number recognized by the vehicle number recognition unit 43, the enforcement time, the rear image, the speed (V) and the enforcement contents.

At this time, the control unit 30 controls the communication interface unit 32 so that the enforcement information generated by the enforcement information generating unit 45 is transmitted to the traffic control center server 9 .

5 is a block diagram illustrating a second vehicle number recognition system, which is a second embodiment of the present invention, and FIG. 6 is an exemplary view of FIG.

The second vehicle number recognition system 200 of FIGS. 5 and 6 is a second embodiment of the present invention.

In addition, the second license plate recognition system 200 includes a rear-view camera 6, a radar device 7, and a traffic control center server 9 that perform the same technology and operation as in FIG. 2 described above.

In addition, the second license plate recognition system 200 further includes a front camera 205 for capturing a front image by photographing the front of the enforcement vehicle C, and a second controller 203 .

The front camera 205 is coupled to the second gantry G2 installed to be spaced apart from the front of the first gantry G1.

In addition, when receiving a trigger signal from the second controller 203, the front camera 205 captures the front of the intermittent vehicle passing between the first and second gantry G1 and G2 to obtain a front image. and output the acquired front image to the second controller 203 .

7 (a) is an exemplary view for explaining a license plate obtained when the front camera and rear camera of FIG. It is an exemplary diagram for explaining the license plate obtained when shooting a two-wheeled vehicle.

The front camera 5 constituting the second vehicle number recognition system 200 of the present invention is a four-wheeled vehicle C1 having a license plate attached to the front and rear of the vehicle body, as shown in FIGS. 7 (a) and (b). ), a front image including the front license plate is acquired, and when shooting a two-wheeled vehicle (C2) with a license plate attached only to the rear of the body, a front image not including the front license plate is acquired. .

In contrast, the rear camera 6 of the present invention, as shown in FIGS. A rear image including It is possible to accurately recognize the vehicle number of a vehicle such as a two-wheeled vehicle, thereby increasing service reliability and enforcement rate.

8 is a block diagram illustrating a second controller of FIG. 5 .

The second controller 203 of FIG. 8 includes a communication interface unit 32, a radar signal analysis unit 34, a vehicle object detection unit 37, and a vehicle information generation unit that perform the same techniques and operations as those of FIG. 4 described above. 38), an intermittent determination unit 39, and a trigger signal generation unit 40.

In addition, the second controller 203 includes a second control unit 230 , a second memory 231 , a front/rear image input unit 241 , a front/rear image analysis unit 242 , and a second vehicle number recognition unit 243 . ), a number recognition verification unit 244 , and a second control information generation unit 245 .

At this time, the communication interface unit 32 constituting the second controller 203 , the radar signal analysis unit 34 , the vehicle object detection unit 37 , the vehicle information generation unit 38 , the enforcement determination unit 39 , the trigger Since the signal generator 40 is configured to perform the same technology and operation as described above with reference to FIG. 4 , a detailed description thereof will be omitted.

The second control unit 230 has the same configuration as the control unit 30 of FIG. 4 described above, but when a trigger signal is generated by the trigger signal generation unit 40, the generated trigger signal is transmitted to the front camera 205 and Controls the communication interface unit 32 so as to be output to the camera 6 for rear view.

In addition, when the front image and the rear image are inputted from the front camera 205 and the rear camera 6 through the communication interface unit 32 , the second control unit 230 converts the input front/rear images into a front/rear image. It is inputted to the input unit 241 .

In addition, when the vehicle number is recognized by the second vehicle number recognition unit 243, the second control unit 230 executes the number recognition verification unit 244, and when the number recognition is verified by the number recognition verification unit 244 The second control information generating unit 245 is executed.

The second memory 231 stores the same data as the memory 31 of FIG. 4 described above, but the front image obtained by the photographing of the front photographing camera 203 is temporarily stored.

In addition, communication identification information and installation location information of the front photographing camera 205 are preset and stored in the second memory 231 .

Also, pixel information corresponding to each lane on the front image is preset and stored in the second memory 231 .

The front/rear image input unit 241 receives a front image and a rear image obtained by photographing the front camera 205 and the rear camera 6 .

The front/rear image analysis unit 242 analyzes the front image and the rear image input through the front/rear image input unit 241, and in detail, after noise removal and pre-processing of the input image, edge detection Alternatively, the license plate is detected using image improvement information, using color and aspect ratio, or using Hough transform.

At this time, the front/rear license plate images detected by the front/rear image analysis unit 242 are input to the second license plate recognition unit 243 .

The second license plate recognition unit 243 analyzes the input front and rear license plate images using a preset number recognition algorithm, and recognizes the front license plate number from the front license plate image, and the rear license plate number from the rear license plate image. recognize

9 is a block diagram illustrating the number recognition verification unit of FIG. 8 .

The number recognition verification unit 244 of FIG. 9 is executed when at least one of a front vehicle number and a rear vehicle number is recognized by the second vehicle number recognition unit 243 .

In addition, as shown in FIG. 9, the number recognition verification unit 244 includes a vehicle number input module 2441, a missing number determination module 2442, a vehicle width calculation module 2443, and a two-wheeled vehicle determination module 2444. , a vehicle number re-recognition module 2445 , the same or not determining module 2446 , a matching module 2447 , an error occurrence confirmation module 2448 , and a vehicle number determining module 2449 .

The vehicle number input module 2441 receives the front vehicle number and the rear vehicle number recognized by the second vehicle number recognition unit 243 .

The missing number determination module 2442 determines whether both the front vehicle number and the rear vehicle number have been input through the vehicle number input module 2441, that is, whether there is a missing number among the front/rear vehicle numbers.

At this time, the control unit 30 in the missing number determination module 2442, 1) if it is determined that the missing number exists, executes the vehicle width calculation module 2443, 2) if it is determined that the missing number does not exist, match The determination module 2447 is executed.

The vehicle width calculation module 2443 is executed when it is determined that the missing number exists in the missing number determination module 2442, and analyzes the front image or rear image of the detected vehicle to determine the vehicle width (D) of the detected vehicle object. ) is calculated.

The two-wheeled vehicle determination module 2444 compares the vehicle width D calculated by the vehicle width calculation module 2443 with a preset threshold.

In this case, the threshold means the minimum value of the vehicle width D at which the detection vehicle can be determined to be a two-wheeled vehicle.

In addition, the two-wheeled vehicle determination module 2444 determines that the detected vehicle is not a two-wheeled vehicle when the vehicle width D is equal to or greater than the threshold, but determines that the detected vehicle is a two-wheeled vehicle when the vehicle width D is less than the threshold.

At this time, in the two-wheeled vehicle determination module 2444, 1) if it is determined that the detection vehicle is a two-wheeled vehicle, the control unit 30 executes the vehicle number determination module 2449, and 2) if it is determined that the detection vehicle is not a two-wheeled vehicle, The vehicle number re-recognition module 2445 is executed.

The vehicle number re-recognition module 2445 1) determines whether the detection vehicle is not a two-wheeled vehicle in the two-wheeled vehicle determination module 2444, or 2) matches the front vehicle number and the rear vehicle number in the matching determination module 2447 to be described later It is executed when it is judged not to do so.

In addition, the vehicle number re-recognition module 2445 is not utilized in the front/rear image analysis unit 242 among the frames of the front image and the rear image obtained by shooting the front camera 5 and the rear camera 6 After extracting other frames that are not, the extracted front image frame and rear image frame are analyzed to re-recognize the front vehicle number and the rear vehicle number.

The same determination module 2446 determines whether the front vehicle number and the rear vehicle number re-recognized by the vehicle number re-recognition module 2445 are the same.

At this time, the control unit 30 executes the vehicle number determination module 2449 if 1) the front/rear vehicle number is determined to be the same in the same determination module 2446, 2) the front/rear vehicle number is not the same If it is determined not to, the failure confirmation module 2448 is executed.

The match determination module 2447 is executed when the missing number determination module 2442 determines that there is no missing number among the front/rear vehicle numbers.

In addition, the matching module 2447 compares the front vehicle number and the rear vehicle number input through the vehicle number input module 2441, and determines whether the front vehicle number and the rear vehicle number match.

At this time, the second control unit 230 executes the vehicle number determination module 2449 when it is determined that 1) the front vehicle number and the rear vehicle number match in the matching module 2447, 2) the front vehicle number and the rear If it is determined that the vehicle number does not match, the vehicle number re-recognition module 2445 is executed.

The fault occurrence confirmation module 2448 is executed when it is determined that the front/rear vehicle number is not the same in the sameness determination module 2446, and finally determines that a failure has occurred in the recognition of the vehicle number.

The vehicle number determination module 2449 is configured to 1) determine whether the detection vehicle is a two-wheeled vehicle in the two-wheeled vehicle determination module 2444, or 2) determine whether the front vehicle number and the rear vehicle number match in the matching determination module 2447, or or 3) when the front/rear vehicle number is determined to be the same in the same determination module 2446, it is executed according to the control of the control unit 30.

In addition, the vehicle number determination module 2449 1) finally determines the rear vehicle number as the vehicle number of the intermittent vehicle (two-wheeled vehicle) when the detection vehicle is determined to be a two-wheeled vehicle by the two-wheeled vehicle determination module 2444 and executed.

In addition, the vehicle number determination module 2449 2) When it is determined that the front vehicle number and the rear vehicle number match in the match determination module 2447 and is executed, the matching front/rear vehicle number is finalized as the vehicle number of the intercepted vehicle. decide

In addition, the vehicle number determination module 2449 3) determines the same front/rear vehicle number as the vehicle number of the intermittent vehicle when it is determined that the front/rear vehicle number is the same and executed by the same determination module 2446. .

Referring back to FIG. 8 and looking at the second enforcement information generating unit 245 , the second enforcement information generating unit 245 is the vehicle number finally determined by the number recognition verification unit 244 , the control time, the front/rear image, and the speed. (V) and generate enforcement information including enforcement details.

At this time, the second control unit 230 controls the communication interface unit 32 so that the enforcement information generated by the second enforcement information generating unit 245 is transmitted to the traffic control center server 9 .

10 is a configuration diagram illustrating a third vehicle number recognition system, which is a third embodiment of the present invention.

The third vehicle number recognition system 300 of FIG. 10 is a third embodiment of the present invention.

In addition, the third license plate recognition system 300 includes a front camera 205, a rear camera 6, and a traffic control center having the same configuration and function as the second license plate recognition system 300 of FIG. 10 described above. It includes a server (9).

In addition, the third license plate recognition system 300 further includes a lidar scanner 17 and a third controller 303 .

The lidar scanner 17 is installed in the first gantry G1, transmits a LiDAR signal forward and backward with respect to the first gantry G1, and then receives a reflected signal.

At this time, although not shown in the drawing, the lidar scanner 17 emits a LiDAR signal toward the polygon mirror, a driving unit that rotates the polygon mirror, and the rotating polygon mirror, and then receives a reflected signal. As it is composed of sensors, it is possible to analyze the lidar signal for 16 channels, so it is possible to effectively detect vehicle objects and remove pedestrians based on various collected data.

After irradiating an optical signal to an external object, the LiDAR signal measures the distance by detecting the time of flight or phase difference of the received optical signal, and is a three-dimensional point cloud of x, y, and z. Cloud) can be used to model a three-dimensional image in real time, so the trajectory of the detected object can be tracked and detected.

The lidar scanner 17 configured in this way outputs the transmitted and received lidar signal information to the third controller 303 , and the third controller 303 receives each of the 16 channels received from the lidar scanner 17 . By analyzing the multi-signal information, it detects the vehicle and calculates the speed of the detected vehicle at the same time to determine whether the vehicle is intercepted. In lane information and a trigger signal are output.

11 is a block diagram illustrating a third controller of FIG. 10 .

The third controller 303 of FIG. 11 includes the communication interface unit 32, the vehicle object detection unit 37, the vehicle information generation unit 38, the intermittent vehicle determination unit 39, and the trigger signal generation unit of FIG. (40), front/rear image input unit 241, front/rear image analysis unit 242, second vehicle number recognition unit 243, number recognition verification unit 244, second enforcement information generation unit 245 includes

In addition, the third controller 303 includes a third control unit 330 , a third memory 331 , a lidar scanner sensitivity control unit 333 , a lidar signal analysis unit 334 , a spectrogram modeling unit 335 , and a profile. It further includes an analysis unit 336 and a lidar scanner sensitivity optimization unit 346 .

The third control unit 330 inputs 16-channel lidar signal information output from the lidar scanner 17 through the communication interface unit 32 to the lidar signal analysis unit 334 .

Also, when the spectrogram is generated by the spectrogram modeling unit 335 , the third control unit 330 stores the generated spectrogram in the third memory 331 .

In addition, the third control unit 330 executes the lidar scanner desensitization optimization unit 346 at every preset period (T).

The third memory 331 stores communication identification information and installation location information of the lidar scanner 17 .

In addition, the detection area of the lidar scanner 17 is preset and stored in the third memory 331 .

Also, a preset reference table is stored in the third memory 331 . In this case, the reference table means data in which the optimum sensitivity value is matched according to the range of the noise generation ratio, and the optimal sensitivity value is set to a lower size as the range of the noise generation ratio increases.

The lidar scanner sensitivity control unit 333 controls the sensitivity (S′, sensitivity) of the lidar signal transmitted from the lidar scanner 17 . At this time, since the technique and method for controlling the sensitivity of the lidar signal are known techniques, a detailed description thereof will be omitted.

In addition, the lidar scanner sensitivity control unit 333 controls the sensitivity of the lidar scanner 17 according to the optimal sensitivity value detected by the lidar scanner sensitivity optimization unit 346 every preset period T.

The lidar signal analysis unit 334 analyzes the lidar signal for each channel transmitted and received by the lidar scanner 17 .

The spectrogram modeling unit 335 uses the analysis data detected by the lidar signal analysis unit 334 to generate a spectrogram for each channel. In this case, the spectrogram is a technique for visualizing and expressing the spectrum of a passing vehicle, and since the spectrogram is a technique commonly used in signal analysis, a detailed description thereof will be omitted.

The profile analyzer 336 detects and analyzes the spectrogram profiles generated by the spectrogram modeling unit 335 .

In this case, the vehicle object detecting unit 37 compares the profiles detected by the profile analyzing unit 336 with preset filtering conditions to detect the vehicle object.

12 is a block diagram illustrating the lidar scanner sensitivity optimization unit of FIG. 11 .

The lidar scanner sensitivity optimization unit 346 of FIG. 12 is executed under the control of the third control unit 330 at every preset period T.

In addition, as shown in FIG. 12, the lidar scanner sensitivity optimization unit 346 includes a data collection module 3461, a sample spectrogram extraction module 3462, a profile analysis module 3463, and a noise profile detection module 3464. , a noise generation ratio (R) calculation module 3465 and a sensitivity optimum value calculation module 3466 .

The data collection module 3461 collects spectrograms stored in the third memory 331 during the period T.

The sample spectrogram extraction module 3462 extracts sample spectrograms that are spectrograms when a vehicle object is detected from the spectrograms collected by the data collection module 3461 .

At this time, the lidar scanner not only generates 16 channels of spectrogram per cycle, but it also takes several seconds for one vehicle to pass through the area. Accordingly, the sample spectrogram extraction module 3462 of the present invention does not extract all spectrograms related to vehicle object detection, but is configured to extract only n spectrograms among them.

The profile analysis module 3463 analyzes the profiles of the sample spectrograms extracted by the sample spectrogram extraction module 3462 .

In the sample spectrogram, the noise profile detection module 3464 analyzes whether a profile other than a vehicle exists in a preset adjacent region from the vehicle profile. It is considered a profile.

For example, when it rains, when the lidar sensor is set to high sensitivity, a noise profile is detected by water splashing by the vehicle body and tires, and since this noise profile reduces the accuracy of object detection, in the present invention, the The noise profile phenomenon was reduced by adjusting the sensitivity of the Ida sensor. As another example, when no noise profile is detected at a point adjacent to the vehicle body and tire, the sensitivity of the lidar sensor needs to be increased by a predetermined degree as the sensitivity is set excessively low.

The noise generation ratio (R) calculation module 465 calculates the sample quantity L, which is the quantity of the sample spectrogram extracted by the sample spectrogram extraction module 462 .

In addition, the noise generation ratio R calculates the noise quantity L', which is the quantity of the noise spectrogram, which is a sample spectrogram including the noise profile detected by the noise profile detection module 464 .

In addition, the noise generation ratio (R) calculation module 465 calculates the noise generation ratio (R, R = L'/L) by dividing the noise quantity (L') by the sample quantity (L).

The sensitivity optimal value calculation module 466 detects an optimal sensitivity value according to the noise generation ratio R by using a preset reference table.

In this case, the reference table means data in which the optimum sensitivity value is matched according to the range of the noise generation ratio, and the optimal sensitivity value is set to a lower size as the range of the noise generation ratio increases.

In addition, the optimal sensitivity value detected by the sensitivity optimal value calculation module 466 is input to the lidar scanner sensitivity control module 333 , and the lidar scanner 17 is controlled by the lidar scanner sensitivity control module 333 . , the lidar signal is transmitted according to the optimal sensitivity value during the next period (T).

As such, the vehicle number recognition system 1, which is an embodiment of the present invention, is configured to recognize the vehicle number using the rear number plate as well as the front license plate of the intercepted vehicle when the intercepted vehicle is detected, so that the front license plate is damaged or the rear license plate is attached By accurately recognizing the license plate number for vehicles such as two-wheeled vehicles, it is possible to increase service reliability and enforcement rate.

In addition, the vehicle number recognition system 1 matches the front and rear images using the vehicle location as a medium in normal times when the license plate recognition of the intercept vehicle is normally performed. Based on this, it is primarily determined whether the enforcement vehicle is a two-wheeled vehicle with only the rear license plate attached, and the front license plate is damaged or only the rear license plate is attached by matching the image of the front of which the front license plate is not recognized to the image of the object of the enforcement vehicle. It can not only accurately recognize the license plate number for a two-wheeled vehicle, but also increase the image matching rate.

In addition, the vehicle number recognition system 1 is configured to optimize the sensitivity of the lidar scanner 17 at every preset period T, thereby maximizing the accuracy and reliability of vehicle detection, thereby increasing service reliability.

1: Front and rear camera-based vehicle number recognition system
3: Controller 6: Rear view camera
7: Radar 9: Traffic Control Center Server
10: communication network 30: control unit
31: memory 32: communication interface unit
34: radar signal analysis unit 37: vehicle object detection unit
38: vehicle information generation unit 39: intercepted vehicle determination unit
40: trigger signal generating unit 41: rear image input unit
42: rear image analysis unit 43: vehicle number recognition unit
45: Enforcement information generation unit

Claims (7)

sensing means for detecting a vehicle;
a controller that analyzes the detection signal detected by the detection means and determines whether the detection vehicle is an intermittent vehicle;
a rear camera for photographing the rear of the intermittent vehicle to obtain a rear image according to the control of the controller;
and a front camera for photographing the front of the intercepted vehicle to obtain a front image under the control of the controller,
the controller is
A vehicle generating vehicle information including the lane, position, and speed (V) of the detected vehicle after analyzing the detection signal detected by the detection means to calculate the lane, position, and speed (V) of the detection vehicle information generating unit;
By comparing the speed (V) information of the vehicle information generated by the vehicle information generating unit with the speed limit (V'), a detection vehicle having a speed (V) exceeding the speed limit (V') is determined as the intermittent vehicle an interception vehicle determination unit;
a trigger signal generating unit for outputting a trigger signal including lane information to the rear camera and the front camera when the intercept vehicle is detected by the interception vehicle determining unit;
a vehicle number recognition unit for recognizing a vehicle number by analyzing the rear image and the front image obtained by the photographing of the rear camera and the front camera;
It is executed when at least one of the front/rear vehicle numbers is recognized by the vehicle number recognition unit, and if the recognized front vehicle number and the rear vehicle number are the same, it is determined that the intermittent vehicle is a four-wheeled vehicle and the same front/rear vehicle number The vehicle number is determined as the vehicle number of the intercepted vehicle, and if the front vehicle number is omitted from the front/rear vehicle number, the vehicle width (D) of the intercepted vehicle is calculated by analyzing the front image or rear image, and then the calculated vehicle width (D) ) is less than the threshold, a number recognition verification unit for determining the enforcement vehicle as a two-wheeled vehicle and at the same time determining the rear vehicle number as the vehicle number of the enforcement vehicle;
and an enforcement information generating unit for generating enforcement information by matching the vehicle number and the image determined by the number recognition verification unit,
The number recognition verification unit
a vehicle number input module for receiving the vehicle number recognized by the vehicle number recognition unit;
an omission number determination module for determining whether a missing number exists among the front/rear vehicle numbers input through the vehicle number input module;
a vehicle width calculation module that is executed when the missing number determination module determines that there is a missing number, and calculates a vehicle width (D) of a detected vehicle object by analyzing a front image or a rear image photographed of the detected vehicle;
The vehicle width D calculated by the vehicle width calculation module is compared with a preset threshold, and if the vehicle width D is greater than or equal to the threshold, the detection vehicle is determined to be a four-wheeled vehicle, but if the vehicle width D is less than the threshold, the detection a two-wheeled vehicle determination module for determining that the vehicle is a two-wheeled vehicle;
It is executed when the detection vehicle is determined to be a four-wheeled vehicle by the two-wheeled vehicle determination module, and among the frames of the front image and the rear image obtained by photographing the front photographing camera and the rear photographing camera, in the vehicle number recognition unit After extracting other unused frames, by analyzing the extracted front image frame and rear image frame, the vehicle number re-recognition module for re-recognizing the front vehicle number and the rear vehicle number;
an identical determination module for determining whether the front vehicle number and the rear vehicle number re-recognized by the vehicle number re-recognition module are the same;
It is executed when the missing number determination module determines that there is no missing number among the front/rear vehicle numbers, and compares the front vehicle number and the rear vehicle number input through the vehicle number input module, and the front vehicle a matching module that determines whether the number and the rear vehicle number match, and if it is determined that the front vehicle number and the rear vehicle number do not match, executes the vehicle number re-recognition module;
a failure confirmation module that is executed when it is determined that the front/rear vehicle number is not the same in the sameness determination module, and finally determines that a failure has occurred in vehicle number recognition;
1) The detection vehicle is determined to be a two-wheeled vehicle by the two-wheeled vehicle determination module, 2) The front vehicle number and the rear vehicle number are determined to match in the matching module, or 3) The front/rear vehicle number is determined by the identicalness determination module Executed when it is determined that the rear vehicle number is the same, 1) When the detection vehicle is determined to be a two-wheeled vehicle in the two-wheeled vehicle determination module and executed, the rear vehicle number is finally determined as the vehicle number of the corresponding enforcement vehicle (two-wheeled vehicle); , 2) When it is determined that the front vehicle number and the rear vehicle number match in the matching module, the matching front/rear vehicle number is finally determined as the vehicle number of the intercepted vehicle, and 3) the same determination is made. When it is determined that the front/rear vehicle number is the same in the module and executed, the vehicle number recognition system comprising a vehicle number determination module that finally determines the same front/rear vehicle number as the vehicle number of the corresponding enforcement vehicle. delete delete delete According to claim 1, wherein the sensing means is
A vehicle number recognition system, characterized in that the radar device receives a reflected signal after transmitting a radar signal toward the front and rear of the installed position. According to claim 1, wherein the sensing means is
It is a lidar scanner that transmits a LiDAR signal toward the front and rear of the installation location and receives the reflected signal,
The lidar scanner
A polygon mirror forming each surface, a driving unit rotating the polygon mirror, and a radar sensor receiving a reflected signal after emitting a radar signal toward the rotating polygon mirror,
the controller is
a memory in which a reference table is preset and stored in which the optimal sensitivity value is matched according to the range of the noise generation rate;
a lidar scanner sensitivity optimization unit that is executed at a preset period (T) and detects an optimal sensitivity value of the lidar scanner;
a lidar scanner sensitivity control unit for controlling the lidar scanner to operate the lidar scanner according to the optimal sensitivity value detected by the lidar scanner sensitivity optimization unit;
a lidar signal analysis unit that analyzes the lidar signal transmitted and received by the lidar sensor of the lidar scanner;
a spectrogram modeling unit that utilizes the analysis data detected by the lidar signal analyzer to generate a spectrogram for the lidar signal transmitted and received by the lidar scanner;
a profile analyzer for detecting and analyzing profiles of the spectrogram generated by the spectrogram modeling unit;
Comparing the profiles detected by the profile analysis unit with preset filtering conditions, detecting a vehicle object, and further comprising a vehicle object detection unit that determines a profile that has passed all the filtering conditions among the detected profiles as a vehicle profile, ,
The vehicle information generating unit
It is executed when a vehicle is detected by the vehicle object detection unit, and generates vehicle information by analyzing the trajectory of the detected vehicle,
The lidar scanner sensitivity optimization unit
a data collection module for collecting spectrograms during the period (T);
a sample spectrogram extraction module for extracting sample spectrograms that are spectrograms when a vehicle object is detected from the spectrograms collected by the data collection module;
a profile analysis module for analyzing profiles of the sample spectrograms extracted by the sample spectrogram extraction module;
In the sample spectrogram extracted by the sample spectrogram extraction module, it is analyzed whether a profile other than a vehicle exists in a preset adjacent region from the vehicle profile, and when a profile other than a vehicle exists in a preset adjacent region, the corresponding profile a noise profile detection module that determines that α is a noise profile;
After calculating the sample quantity (L), which is the quantity of sample spectrograms extracted by the sample spectrogram extraction module, the noise which is the quantity of noise spectrograms that are sample spectrograms including the noise profile detected by the noise profile detection module a noise generation ratio calculation module for calculating the quantity (L') and calculating the noise generation ratio (R, R=L'/L) by dividing the calculated noise quantity (L') by the calculated sample quantity (L);
By utilizing the reference table stored in the memory, the vehicle number recognition characterized in that it further comprises a sensitivity optimum value calculation module for detecting the optimum sensitivity value according to the noise generation rate (R) calculated by the noise generation rate calculation module system. The method of claim 6, wherein the lidar scanner is
A vehicle number recognition system, characterized in that transmitting and receiving a lidar signal through 16 channels, and the controller detects a vehicle by analyzing the lidar signal for each channel.

Images