KR20050042123A - Vehicle license plate recognition system and the method for improving the definition of vehicle license plate image thereby - Google Patents

Abstract

The present invention relates to a license plate recognition system and a method for improving the sharpness of an image of a license plate using the system, and more particularly, a sensor for detecting illuminance in the sky direction and two for detecting illuminance of sunlight reflected in the license plate direction. And a vehicle number recognition system including two sensors, the illumination values measured by the two detection sensors, that is, the illumination in the sky direction, the illumination of the sunlight reflected in the license plate direction, and the road surface image captured by the camera. Calculate the optimal license plate roughness by using the measured road surface roughness value and control the aperture of the camera in proportion to the calculated optimal license plate illuminance. And always get a clear image of the license plate regardless of changes in illuminance caused by the cloud obscuring the sun from time to time. It relates to a vehicle number recognition system and a method for improving the sharpness of the image of the license plate using the system.

The present invention provides a first illuminance sensor for measuring the illuminance of sunlight reflected in the license plate direction, an A / D converter for converting the output of the first illuminance sensor into a digital signal, and a camera, strobe, and breaker by the output of the vehicle detecting unit. And a computer for recognizing a vehicle number by analyzing a license plate image transmitted from a camera by a trigger signal outputted from the MPU, and an MPU controlling the aperture driving unit, the other input terminal of the A / D converter. Is connected to the second illuminance sensor for sensing a light intensity in the sky direction and outputs a signal proportional to the sensed illuminance value, and reads road roughness using a road surface image connected to an output of the camera and transmitted from the camera. And a road roughness reading unit for outputting the read data to the MPU.

In addition, the present invention is a first illuminance sensor for measuring the sunlight (Sf) of the sunlight reflected in the license plate direction, an A / D converter for converting the output of the first illuminance sensor into a digital signal, and the camera by the output of the vehicle detection unit And a strobe, a circuit breaker, an MPU for controlling the aperture driving unit, a second illuminance sensor for measuring illuminance (St) in the sky direction, a road surface roughness reading unit for measuring road surface roughness using a road surface image, and an output from the MPU. In the vehicle number recognition system consisting of a computer for recognizing the license plate number by analyzing the license plate image transmitted from the camera by the trigger signal, the MPU is the illumination (Sf) in the direction of the license plate, the illumination (St) in the sky direction, The illuminance receiving process of reading the road surface roughness Ct, the illuminance Sf in the direction of the license plate, the illuminance St in the sky direction, and the road surface roughness Ct received during the illuminance receiving process The calculation process of calculating the optimal license plate roughness (Cf) by the equation, and compares the optimal license plate roughness (Cf) and the stored database calculated by the calculation process to the optimal license plate roughness (Cf) It is characterized in that it comprises a data extraction process for extracting the corresponding aperture control data, and an aperture adjustment process for adjusting the aperture by controlling the aperture driver as the data extracted through the data extraction process.

Description

Vehicle license plate recognition system and the method for improving the definition of vehicle license plate image hence}

The present invention relates to a license plate recognition system and a method for improving the sharpness of an image of a license plate using the system, and more particularly, a sensor for detecting illuminance in the sky direction, and two for detecting illuminance of sunlight reflected on the license plate. A vehicle number recognition system including a sensor, and measured by the illumination values measured by the two detection sensors, that is, the illumination in the sky direction, the illumination of the sunlight reflected in the license plate direction, and the road image captured by the camera By using the calculated road surface illuminance value, the optimum license plate illuminance is calculated, and the camera's aperture is controlled in proportion to the optimal license plate illuminance. And you can always get a clear image of the license plate regardless of changes in illumination caused by the cloud obscuring the sun. The present invention relates to a license plate recognition system and a method for improving the sharpness of a license plate image using the system.

Due to the recent increase in vehicles, many problems related to vehicles such as traffic accidents, traffic jams, stolen vehicles, and speeding vehicles have emerged as social issues. Accordingly, vehicle management, vehicle tracking, automatic toll billing, parking management and arrangement Many researches are being conducted to solve problems through license plate recognition in application fields such as vehicle management.

In general, the application fields of the vehicle identification system are used in speeding enforcement systems (fixed, portable, mobile), signal violation enforcement systems, toll evasion prevention systems, parking lot management systems, digital cameras, strobes, etc. It consists of Grab Board, Main Number Recognition Controller, Auto IRIS Lens, and Vehicle Detection Sensor (Loop Coil, etc.).

An important core technology in such a vehicle identification system is to obtain a good image for the main controller to recognize the number. The key to the recognition rate depends on whether the obtained vehicle image is a good image. The accuracy and reliability of the number recognition algorithm and the state of the license plate affect the success rate.

In general, in order to obtain a better image of the license plate, by adjusting the intensity of the strobe, by adopting an IRIS lens, by setting the illumination of the seasonal time zone in advance, or feedback control of the aperture with the brightness of the captured image Although many attempts have been made, they failed to satisfy conditions for variables such as weather conditions, sun movement conditions, and cloud movement conditions.

Accordingly, in order to obtain a better license plate image, Korean Patent No. 416252 (Registration Date: January 12, 2004) extracts data corresponding to preset license plate color information by converting color coordinates of a captured image to obtain license plate image data. At the same time, the license plate image data is detected according to the license plate color information, and the coordinates of the license plate image are extracted by projecting the detected license plate image data onto a two-dimensional plane. Even if there is a license plate image can be extracted accurately.

In addition, the Republic of Korea Patent Publication No. 2003-51557 (published date: June 25, 2003), the system intelligently identifies whether there is a vehicle in a continuous image taken in real time, and if the vehicle image is determined to determine the vehicle number By recognizing and extracting, the vehicle number with the highest accuracy is selected from several images, thus reducing the probability of being unrecognized and misidentified.

Conventional Patent Nos. 416252 and 2003-51557 as described above can obtain a clear image of a license plate in a state of constant illumination, but in the case of backlight and pure light caused by the movement of the sun, or clouds If the illumination is changed from time to time because the sun obscurs repeatedly, clear images of license plates cannot be obtained.

That is, as shown in (a) of FIG. 1, when the backlight is backlit according to the movement path of the sun, the image of the license plate becomes dark overall, so that it is difficult to obtain a clear image. On the contrary, as shown in (b) of FIG. In this case, the image of the license plate becomes bright overall, and there is a problem in that a clear image cannot be obtained due to the reflection of light.

In addition, there is a problem that the shooting brightness is different from the clear day and rainy weather, especially in the case of a clear day because the shooting illuminance changes due to the phenomenon that the cloud often obscures the sun, there is a problem in the sharpness.

The present invention is provided with a vehicle number recognition system including a sensor for detecting the illumination of the sky direction and two sensors for detecting the illumination of the sunlight reflected on the license plate in order to solve the above problems, the two detection sensors The optimal license plate illuminance is calculated by using the illuminance value measured in the sky direction, that is, the illuminance in the sky direction, the sunlight intensity reflected in the license plate direction, and the road surface illuminance value measured by the road image taken by the camera. By controlling the aperture of the camera in proportion to the optimal license plate illuminance, the license plate is always vivid regardless of the change in illumination caused by the sun's movement, as well as on the sunny and cloudy days, and the cloud obscuring the sun from time to time. It is a technical task to be able to obtain an image.

Hereinafter, a vehicle number recognition system according to the present invention and a method for improving the sharpness of a license plate image using the system will be described in detail with reference to FIGS. 2 to 6.

2 is a view illustrating an installation state of a vehicle number recognition system according to the present invention. On one side of the housing 1, a camera 2 for photographing a license plate image and a strobe 3 for supplying lighting to a license plate at night are provided. On the other side of the housing (1), the first illuminance sensor (4) for measuring the illuminance of the sunlight reflected in the vehicle license plate direction and outputs an electrical signal proportional to the measured illuminance value is respectively installed, the housing On the upper surface of (1), the second illuminance sensor 5 which measures illuminance in the sky direction during the day and outputs an electric signal proportional to the measured high illuminance value, and measures the dark low illuminance at night A third illuminance sensor 6 for outputting an electrical signal proportional to the illuminance value is provided.

Then, the circuit breaker 7 controls the entry of the vehicle at a predetermined interval from the housing 1, and calculates the optimum license plate roughness by a computer program to control the aperture of the camera 2 and other vehicle number recognition systems. The control box 8 is installed to control the equipment 10 as a whole.

3 is a block diagram of a vehicle identification system according to the present invention, the first to third illuminance to measure the illumination of sunlight in the direction of the license plate, the illumination of the sky direction and the dark low light at night to output an electrical analog signal An output terminal of the sensors 4, 5 and 6 is connected with an A / D converter 27 which converts the digital signal in proportion to the sensed illuminance.

The output terminal of the limit switch 9a which is contacted as the vehicle detecting unit 9 and the circuit breaker 7 rise and fall sensing the entry of the vehicle, and the input terminal of the blocking drive unit 7a that raises and lowers the circuit breaker 7. The first amplifier 28 is connected to the.

At the output terminal of the A / D converter 27 and the input / output terminal of the first amplifier 28, an image of the road surface transmitted from the camera 2 is measured at regular intervals, and the road surface roughness is measured. After calculating the optimal license plate illumination value by using the illumination of the sunlight reflected in the direction of the license plate transmitted from the first and second illumination sensor 4, 5 and the sky direction, the corresponding camera ( 2) A microprocessor unit (hereinafter referred to as 'MPU') 26 which transmits data for controlling the iris and controls all the devices of the system is connected.

A second amplifier 25 for driving the camera 2 and the strobe 3 by various information input to one output terminal of the MPU 26 and transmitting the information to the computer 23 via the communication port 21. Is connected to the other output terminal of the MPU 26, a third amplifier 24 for transmitting a trigger signal to the computer 23 when the vehicle is detected by the vehicle detection unit 9 is connected, the MPU (26) The other end of the) is connected to the aperture driving unit 11 for adjusting the aperture of the camera 2, the other side of the computer 23 to read the road surface roughness as a data signal transmitted from the camera 2 as a data signal A road roughness reading unit 22 for processing and transmitting is connected.

Referring to the conventional operation process already known in the vehicle number recognition system according to the present invention configured as described above are as follows.

First, when a vehicle entry is sensed by the vehicle detection unit 9, the vehicle detection signal is transmitted to the MPU 26 through the first amplifier 28, and the MPU 26 transmits the second amplifier 25. At the same time as driving the camera 2 to photograph the license plate, the MPU 26 transmits a vehicle detection signal, that is, a trigger signal, to the computer 23 through the third amplifier 24.

At this time, the image of the license plate photographed by the camera 2 is converted into data by the road roughness reading unit 22 and transmitted to the computer 23, the computer 23 is the road roughness reading unit by the trigger signal Read the data received from 22, read the vehicle number, and process the various information (vehicle number, time of entry, passing area, type of car, etc.) by the program and issue a parking ticket (passport) immediately after the vehicle number is read. Or you will collect a parking fee.

When the parking ticket (toll) is issued or the parking fee (toll) is collected, the MPU 26 controls the shutoff mechanism 7a to raise and lower the breaker 7, and at the upper and lower limits of the lowering. When reaching, the limit switch 9a is activated to prevent excessive lifting of the breaker 7.

In addition, when the MPU 26 reads the output signal value of the third illuminance sensor 6 and recognizes it as a dark night, the vehicle detection unit 9 operates the strobe 3 whenever the vehicle entry is sensed. It is supposed to photograph the license plate.

Next, a method for improving the sharpness of a license plate image in a vehicle number recognition system according to the present invention will be described in detail with reference to FIGS. 4A to 6.

First, in order to realize a method for improving the sharpness of the license plate image according to the present invention, three types of illuminance should be recognized. The first is illuminance (Sf) of sunlight reflected in the direction of the license plate, and the second is illuminance in the sky ( St), and the third is road surface roughness (Ct), and by recognizing the above three roughnesses, the optimum license plate roughness (Cf) can be accurately measured, and the measured license plate roughness (Cf) of the camera By adjusting the aperture of (2), it is possible to obtain a clear image of the license plate.

Hereinafter, a process of obtaining the three kinds of roughness described above will be described in detail with reference to FIGS. 4A and 4B.

FIG. 4A shows the first illuminance sensor 4 and 4a installed at the entrance / exit side housing 1 installed in a parking lot of a specific place to measure the illuminance (Sf) of sunlight reflected in the direction of the first license plate. It records the output waveform according to the illuminance from to sunrise.

That is, (a) of FIG. 4A is an output waveform diagram of the first illuminance sensor 4 installed on the inlet side of the cloudy day, and (B) is an output waveform diagram of the first illuminance sensor 4a installed on the outlet side of the cloudy day. , (C) is an output waveform diagram of the first illuminance sensor 4 installed on the sunny day entry, (d) is an output waveform diagram of the first illuminance sensor 4a installed on the exit day on the sunny day.

As shown in (a), (b), (c), and (d) of FIG. 4a, the output waveforms of the first illuminance sensor 4 and 4a on the exit side of the first illumination sensor are on a cloudy day compared to a sunny day. This irregularity can be seen, which is a phenomenon caused by the change in illuminance due to the cloud obscuring the sun as well as the factor of the change in illuminance according to the degree of cloudyness.

In addition, as shown in (a), (c), (b), and (d) of FIG. 4a, it can be seen that the output waveforms at the inlet and outlet sides are symmetrical with each other, which is caused by the backlight of the sun. And phenomenon caused by pure light.

As described above, the first illuminance sensor 4 senses the illuminance Sf in the direction of the license plate, which changes from time to time due to weather, clouds, and pure light and backlight, and outputs an electrical signal proportional to the first illuminance sensor. The illuminance (Sf) signal in the direction of the license plate output from (4) is converted into a digital signal through the A / D converter 27 and transmitted to the MPU 26.

Next, the second illuminance sensor 5 is mounted on the upper surface of the housing 1 in order to measure the second sky illuminance St. The second illuminance sensor 5 measures the illuminance in the sky direction. Because it is only affected by the weather, ie cloudy and sunny days, and the flow of clouds, it is not affected by pure or backlit.

As such, the sky illuminance St, which is frequently changed by cloudy days, sunny days, and cloud flows, is sensed by the second illuminance sensor 5 to output an electrical signal proportional to the second illuminance sensor 5. The sky illuminance St signal outputted from (5) is converted into a digital signal through the A / D converter 27 and transmitted to the MPU 26.

Next, in order to measure the third road surface roughness Ct, the camera 2 photographs the road surface in the vehicle entry direction at regular intervals, and measures the road surface roughness with the captured image.

That is, the MPU 26 transmits a photographing signal to the camera 2 at regular intervals through the second amplifier 25, and the camera 2 receives a photographing signal from the MPU 26 in a vehicle entry direction. The road surface photographed, and the photographed road surface image is transmitted to the road surface roughness reading unit 22.

The road surface roughness reading unit 22 receives a road surface image transmitted from the camera 2 to measure road surface roughness Ct, and converts the measured road surface roughness Ct into a digital signal and transmits it to the MPU 26. do.

The road surface roughness reading unit 22 receives a road surface image and measures road surface roughness Ct using a GetPixel function, passing the horizontal and vertical position values of the corresponding pixel as a factor value. When this function is called, the color value of the pixel represented at the position is returned as hexadecimal value. When the color value is obtained, it is classified into three colors of R, G, and B, and also classified by color temperature. Determination of the illuminance, which corresponds to the publicly known art, and thus detailed description thereof will be omitted.

The road surface roughness reading unit 22 may be embedded in the computer 23, and may be implemented as software by installing a road surface roughness reading program in the computer 23.

Figure 4b is a road surface roughness reading unit 22 receives the road surface image transmitted from the camera 2 to measure the road surface roughness (Ct), the sunrise at the entrance and exit side installed in the parking lot of a particular place sunrise at sunset The output waveform according to the measured road roughness is recorded.

That is, Figure 4b (a) is the output waveform of the road surface roughness on the cloudy day, (b) is the output waveform of road surface roughness on the cloudy day, (c) is the output waveform of the road surface roughness on the sunny day (D) is an output waveform diagram of road surface roughness on the exit side on a clear day.

As shown in (a), (b), (c), and (d) of FIG. 4b, it can be seen that cloudy days are irregular compared to sunny days, and the output waveform of the road surface roughness (Ct) is irregular. This is a phenomenon caused by the change in illuminance due to the cloud obscuring the sun as well as the factor of the change in illuminance according to the degree of cloud quantity.

In addition, as shown in (a), (c), (b), and (d) of FIG. 4b, it can be seen that the output waveforms at the in and out sides have a similar shape, and the road surface roughness (Ct) It can be seen that it is not influenced by the backlight and pure light due to the movement of the sun, but only by the influence of the weather, ie, cloudy and sunny days, and the flow of clouds.

As mentioned above, the MPU 26 receives three types of illuminance, that is, illuminance Sf in the license plate direction, illuminance St in the sky direction, and road surface roughness Ct, thereby obtaining an optimal license plate illuminance Cf value. A process of acquiring a clear image of the license plate by calculating and adjusting the aperture of the camera 2 in proportion to the determined optimal license plate illuminance Cf will be described in detail with reference to FIG. 5.

First, in order to calculate the optimum license plate roughness Cf, the relationship between the solar light intensity Sf reflected in the license plate direction and the sky light intensity St and the road surface roughness Ct will be described.

That is, measured by the illuminance (Sf) of the sunlight reflected in the direction of the license plate output from the first illuminance sensor (4), the illuminance (St) in the sky direction output from the second illuminance sensor (5), and the road surface image The relationship between the road roughness Ct and the optimum license plate roughness Cf is equal to the following equation (1).

Ct: Cf = St: Sf ....... (Equation 1)

(Ct is the illuminance of the road surface, Cf is the optimal license plate illuminance, St is the illuminance in the sky direction, and Sf is the illuminance of the sunlight reflected in the license plate direction.)

The following proportional expression is established by Equation 1.

Cf × St ∝ Ct × Sf ....... (Equation 2)

In addition, Equation 2 is equal to Equation 3 below.

....... (Equation 3)

That is, as shown in Equation 3, the optimum license plate roughness (Cf) is inversely proportional to the roughness (St) in the sky direction, and is proportional to the product of the roughness (Sf) and the road surface roughness (Ct) of sunlight reflected in the license plate direction. It can be seen.

(In this case, k is a constant, and the ideal value is '1', but it may vary depending on the type and type of the camera 2, and it is a value that hardly affects the entire equation.)

Accordingly, in the MPU 26, illuminance Sf of the sunlight reflected from the direction of the vehicle license plate output from the first illuminance sensor 4, illuminance St of the sky direction output from the second illuminance sensor 5, By calculating the road surface roughness Ct output from the road surface roughness reading unit 22 according to Equation 3, an optimum license plate roughness Cf can be obtained.

In addition, the MPU 26 stores data for controlling the aperture of the camera 2 in various stages according to the calculated optimal license plate illuminance Cf as a database as shown in Table 1 below.

Table 1

Calculated optimal license plate roughness (Cf) Aperture Adjustment a≤Cf≤b k1 b <Cf≤c k2 c <Cf≤d k3 .... .... y <Cf≤z kn

FIG. 5 is a flowchart illustrating an operation of the control unit in the vehicle number recognition system according to the present invention. When power is supplied to the vehicle number recognition system 10, the MPU 26 initializes the system, and then the first illuminance sensor 4 every certain period. ) Is read through the A / D converter 27 to read the illuminance Sf of the sunlight reflected in the direction of the license plate output from the light and the sky direction illuminance St output from the second illuminance sensor 5 (step 101, 102).

At the same time, the MPU 26 transmits a photographing signal to the camera 2 through the second amplifier 25 at predetermined intervals, and the camera 2 receives a photographing signal from the MPU 26 in a vehicle entry direction. The road surface photographed, and the road surface image is taken to be transmitted to the road surface roughness reading unit 22, the road surface roughness reading unit 22 receives and reads the road surface image transmitted from the camera (2) Ct) is measured, and the measured road roughness Ct is converted into a digital signal and outputted, and the road roughness Ct is read from the MPU 26 (step 103) to perform an illumination reception process L1. Done.

When the illuminance receiving process (L1) is completed, using the illuminance (Sf) of the sunlight reflected in the direction of the vehicle license plate in the MPU 26, the illuminance (St) and the road surface roughness (Ct) in the sky direction by the above-described Equation 3 The calculation process L2 of calculating the license plate roughness Cf of the vehicle is performed.

Thereafter, the optimum vehicle license plate illuminance Cf calculated through the calculation process L2 and the stored database are compared with each other, and the aperture control data corresponding to the optimum license plate illuminance Cf calculated as shown in Table 1 below. It performs a data extraction process (L3) for extracting, and performs the aperture adjustment process (4) for adjusting the aperture to be proportional to the received data by transmitting the extracted data to the aperture driver (11).

The aperture controlled through the above process is not significantly affected by the backlight and the pure light according to the movement path of the sun, and the light amount (in response to the illumination changes due to the weather, that is, cloudy days and sunny days, and cloud flow) By allowing the light to be optimally received, a clear image of the license plate can be obtained, and the adjustment of the aperture is repeatedly performed at regular intervals.

Figure 6 (a), (b) is a state diagram of the license plate photographed in the backlight and forward light by the method of the present invention, (a) is photographed in the backlight state, (b) is photographed in the backlight state Even in daylight and backlight due to the sun's path, clear images are always obtained.

As such, when the vehicle enters the vehicle number recognition system 10 in a state in which the aperture of the camera 2 is adjusted at regular intervals under the control of the MPU 26, the vehicle detection unit 9 outputs a detection signal. The signal is transmitted to the MPU 26 through the first amplifier 28.

When the vehicle detection signal is received, the MPU 26 drives the camera 2 through the second amplifier 25 to photograph the license plate of the entering vehicle and transmits the license plate to the computer 23. The trigger signal is transmitted to the computer 23 through 24.

When the trigger signal is received from the MPU 26, the computer 23 receives and analyzes the image of the vivid license plate as shown in FIG. 6 transmitted from the camera 2 to recognize the number of the vehicle entering by the program. In addition, by processing the various information (entry and departure time, passing area, car model, etc.) based on the recognized vehicle number, the parking ticket (toll ticket) is issued or the parking fee (toll fee) is collected.

Therefore, the present invention provides the illuminance Sf of the sunlight reflected by the first illuminance sensor 4 in the direction of the license plate, the illuminance St of the sky direction by the second illuminance sensor 5, and the camera 2. Using the road surface roughness (Ct) by the road surface roughness reading unit 22 for analyzing the road surface image taken by the calculation according to the above equation 3 calculates the optimal license plate roughness (Cf), the camera By controlling the aperture of (2) at regular intervals, the image of the license plate is always clear regardless of the change in the illumination caused by the sun's movement as well as on the clear and cloudy days and the cloud obscuring the sun from time to time. You can get it.

As described above, the present invention includes a vehicle number recognition system including a sensor for detecting illuminance in a sky direction and two sensors for detecting illuminance of sunlight reflected in a vehicle license plate direction. The brightness of the license plate is calculated using the illuminance value measured in the sky direction, that is, the illuminance of the sunlight reflected in the direction of the license plate, and the road surface illuminance value measured by the image of the photographed road surface. By controlling the aperture of the camera in proportion to the optimal license plate illuminance, the license plate is always vivid regardless of the change in the illumination caused by the sun's movement as well as the bright and cloudy days and clouds obstructing the sun. The effect is to get the image.

Figure 1 (a), (b) is a photographic image attitude of the license plate plate in the backlight and pure light by the conventional method.

2 is an installation state diagram of a vehicle number recognition system according to the present invention.

3 is a block diagram of a vehicle number recognition system according to the present invention.

4A is an illuminance waveform diagram in the direction of a license plate output from sunrise to sunset from a first illuminance sensor installed at an entry / exit side;

4B is a waveform diagram of road roughness measured by analyzing a road image transmitted from a camera installed at an entry / exit side from sunrise to sunset.

5 is an operation flowchart of a control unit in a vehicle number recognition system according to the present invention.

Figure 6 (a), (b) is a photographing state of the license plate plate in the backlight and pure light by the method of the present invention.

<Explanation of symbols for main parts of drawing>

 1: housing 2,2a: camera 3,3a: strobe

 4,4a: 1st illuminance sensor 5,5a: 2nd illuminance sensor 6,6a: 3rd illuminance sensor

 7: Breaker 7a: Breaker mechanism 8: Control box

 9: vehicle detection unit 10: vehicle number recognition system 11: aperture driving unit

21: communication port 22: road roughness reading unit 23: computer

24: third amplifier 25: second amplifier 26: MPU

27: A / D converter 28: the first amplifier

Claims (2)

A first illuminance sensor 4 for measuring illuminance of sunlight reflected to the license plate direction, an A / D converter 27 for converting the output of the first illuminance sensor 4 into a digital signal, and a vehicle detecting unit 9 MPU 26 controlling the camera 2, the strobe 3, the circuit breaker 7, the aperture driver 11 by the output of the camera, and the camera 2 by the trigger signal output from the MPU 26. In the vehicle number recognition system consisting of a computer 23 for recognizing the license plate number by analyzing the license plate image transmitted from, A second illuminance sensor 5 connected to the other input terminal of the A / D converter 27 for sensing illuminance in the sky and outputting a signal proportional to the sensed illuminance value; A road roughness reading unit 22 connected to an output terminal of the camera 2 to read a road roughness using a road image transmitted from the camera 2, and outputs the read data to the MPU 26. Vehicle number recognition system, characterized in that it is included. A first illuminance sensor 4 for measuring illuminance Sf of sunlight reflected in the license plate direction, an A / D converter 27 for converting the output of the first illuminance sensor 4 into a digital signal, and a vehicle detection The second illuminance for measuring the illuminance St in the sky direction and the MPU 26 for controlling the camera 2, the strobe 3, the breaker 7, and the aperture driver 11 by the output of the unit 9. The license plate image transmitted from the camera 2 by the sensor 5, the road surface roughness reading unit 22 for measuring the road surface roughness (Ct) using the road surface image, and the trigger signal output from the MPU (26) In a vehicle number recognition system consisting of a computer 23 for analyzing and recognizing a vehicle number, The MPU 26 is an illuminance (Sf) of the sunlight reflected in the license plate direction, an illuminance receiving process (L1) for reading the illuminance (St) in the sky direction, road surface roughness (Ct), The illuminance (Sf) in the direction of the license plate received in the illuminance receiving process (L1), the illuminance (St) in the sky direction, the road surface roughness (Ct) A calculation process L2 for calculating an optimum license plate illuminance Cf by the equation; A data extraction process (L3) for extracting aperture adjustment data corresponding to the optimal license plate illuminance (Cf) by comparing the optimal license plate illuminance (Cf) and the stored database calculated through the calculation process (L2); , Method for improving the clarity of the license plate image using a vehicle number recognition system, characterized in that it comprises an aperture adjustment process (4) for controlling the aperture by controlling the aperture driving unit as the data extracted through the data extraction process (L3).