JP5906856B2 - Vehicle image recognition apparatus and vehicle image recognition method - Google Patents

Description

  The present invention relates to a vehicle image recognition device and a vehicle image recognition method.

  Conventionally, for the purpose of vehicle management in parking lots, finding stolen vehicles, etc., there is a technology that uses a camera to photograph a license plate attached to the vehicle, and analyzes the captured image to read numbers and letters. It has been.

  By the way, there is a driver who attaches a cover made of resin or the like (hereinafter referred to as a plate cover) to the license plate for preventing the license plate from being stained or having a design preference. However, some plate covers reduce light transmittance and others absorb or reflect infrared rays, so that the plate cover may prevent reading numbers and characters from images.

  In this connection, a number reading device capable of analyzing both an infrared image using infrared light and a visible image using visible light is known. The number reading device uses the number reading result based on the visible image when it is determined to be daytime based on the determination result of the day / night determination unit, and uses the number reading result based on the infrared image when it is determined to be night. In addition, this number reading device cannot be read with an infrared image, and when a number can be read only with a visible image, a fraudulent plate cover that absorbs or reflects infrared light is attached. Detect as a plate.

  In addition, an infrared camera that captures the license plate and a visible light camera that captures the entire image of the vehicle are provided, and vehicle number information and overall image information are extracted based on images captured by the infrared and visible light cameras. A portable vehicle number recognition device is known. This portable vehicle number recognition apparatus controls the illumination intensity of the infrared illumination apparatus by linking the recognized brightness information of the license plate and the illumination current control mechanism.

JP 2008-299645 A JP 2006-190198 A

  However, the above-described conventional number reading device can only detect the presence of a plate cover that makes infrared light imaging impossible, and cannot monitor any plate cover having a certain degree of light transmission. .

  Further, the conventional portable vehicle number recognition apparatus does not take into account the presence of a plate cover in the first place.

  In one aspect, an object of the present invention is to more accurately determine whether a plate cover is attached to a license plate of a vehicle.

According to an aspect of the present invention, a plate image area extracting unit that extracts a plate image area corresponding to a license plate from an input image including a vehicle, and an average of first areas corresponding to license plate description information of the plate image area A luminance difference calculation unit for calculating a luminance difference between the luminance and an average luminance of the second area corresponding to the background of the license plate in the plate image area; the calculated luminance difference; and a set standard luminance difference A determination unit that determines whether a plate cover is attached to the license plate of the vehicle, and an area dividing unit that divides the plate image area into a plurality of areas, and the brightness difference calculating unit includes: for each of the divided regions and calculating the brightness difference, and outputs the largest luminance difference among the plurality of luminance difference the calculated in the evaluation unit, the vehicle image It is an identification device.

  According to one embodiment, it can be determined more accurately whether or not a plate cover is attached to a license plate of a vehicle.

It is a figure which shows the vehicle image recognition system which concerns on 1st Example of this invention. 2 is a hardware configuration example of the vehicle image recognition device 1. 2 is a functional configuration example of the vehicle image recognition device 1. It is a figure which shows the edge detection image produced | generated from a captured image. It is a figure which shows the structure of a license plate. It is the figure which compared the brightness | luminance difference rate calculated in the number plate with which the plate cover is not attached, and the number plate with which the plate cover was attached with the threshold value (alpha), respectively. It is a figure which shows a mode that a statistical analysis is performed with respect to a some luminance difference rate, and plate cover determination is performed. 4 is an example of data stored as a recognized vehicle database 60. 4 is an example of data stored in a standard average luminance database 62 by a standard average luminance calculation unit 44. It is an example of the flowchart which shows the flow of the process performed by the vehicle image recognition apparatus 1 of 1st Example. It is a functional structural example of the vehicle image recognition apparatus 2 which concerns on 2nd Example of this invention. It is a figure which shows a mode that the brightness difference of a plate image area increases / decreases, and the division | segmentation aspect of a plate image area corresponding to this when the imaged vehicle has turned on a headlight. It is explanatory drawing for demonstrating a mode that it is determined whether the plate image area | region exists in the center part of a vehicle front part, or an edge part. It is a figure which shows a mode that the maximum value of a brightness | luminance difference rate becomes more than threshold value (alpha) as a result of the area | region division part 46 dividing | segmenting a plate image area | region. It is a figure which shows a mode that the brightness | luminance difference of a plate image area increases / decreases by the influence of sunlight, and the division | segmentation aspect of a plate image area corresponding to this when an imaging time slot | zone is the daytime. It is the figure which compared the brightness | luminance difference rate calculated with respect to each of the area | region A and B, and the threshold value (alpha) when the imaging time slot | zone is the daytime. When the imaged vehicle is a large vehicle, it is a figure which shows a mode that the brightness | luminance difference of a plate image area increases / decreases by dirt of a license plate, and the division | segmentation aspect of a plate image area corresponding to this. It is the figure which compared the brightness | luminance difference rate calculated with respect to each of the area | region A, B, and C with the threshold value (alpha) when the imaging time slot | zone is the daytime and the imaged vehicle is a large vehicle. It is an example of the flowchart which shows the flow of the process performed by the vehicle image recognition apparatus 2 of 2nd Example. It is an example of the flowchart which shows the flow of the process performed by the vehicle image recognition apparatus 2 of 2nd Example. It is another example of the flowchart which shows the flow of the process performed by the vehicle image recognition apparatus 2 of 2nd Example. It is a functional structural example of the vehicle image recognition apparatus 3 which concerns on 3rd Example of this invention. It is an example of the flowchart which shows the flow of the process performed by the vehicle image recognition apparatus 3 of 3rd Example. It is an example of the flowchart which shows the flow of the process performed by the vehicle image recognition apparatus 3 of 3rd Example. It is another example of the flowchart which shows the flow of the process performed by the vehicle image recognition apparatus 3 of 3rd Example. It is a functional structural example of the vehicle image recognition apparatus 4 which concerns on 4th Example of this invention. It is a figure which shows the comparison with the threshold value (alpha) and the brightness | luminance difference rate calculated with respect to each area | region AG divided | segmented by the character area division | segmentation part 50, and each area | region. It is an example of the flowchart which shows the flow of the process performed by the vehicle image recognition apparatus 4 of 4th Example. It is a functional structural example of the vehicle image recognition apparatus 5 which concerns on 5th Example of this invention. It is an example of the flowchart which shows the flow of the process performed by the vehicle image recognition apparatus 5 of 5th Example. It is an example of the flowchart which shows the flow of the process performed by the vehicle image recognition apparatus 5 of 5th Example.

  DESCRIPTION OF EMBODIMENTS Hereinafter, embodiments for carrying out the present invention will be described with reference to the accompanying drawings.

<First embodiment>
Hereinafter, a vehicle image recognition device, a vehicle image recognition system, a vehicle image recognition method, and a vehicle image recognition program according to a first embodiment of the present invention will be described with reference to the drawings.

[Hardware configuration]
FIG. 1 is a diagram showing a vehicle image recognition system according to a first embodiment of the present invention. The vehicle image recognition system according to the first embodiment includes a vehicle image recognition device 1 and a camera 100. The vehicle image recognition apparatus 1 is connected to a camera 100 and receives image data of an image captured by the camera 100.

  The camera 100 is a digital camera using an image sensor such as a charge coupled device (CCD) or a complementary metal oxide semiconductor (CMOS). The camera 100 is attached to an arbitrary position where the vehicle can be imaged from the front or the back, such as an upper part of a power pole, a ceiling of a parking lot, a structure suspended on a road, or the like.

  The camera 100 is provided with an infrared light irradiation lamp 110 that irradiates the vehicle with infrared light, and the camera 100 can capture infrared light. For this reason, at night, the infrared light irradiation lamp 110 irradiates the vehicle with infrared light, and the camera 100 captures the infrared light, so that the characters on the license plate in the image can be read. The switching control between the daytime imaging operation and the nighttime imaging operation may be actively performed by the camera 100 and the infrared light irradiation lamp 110, or may be performed by the vehicle image recognition apparatus 1. The camera 100 may include a camera that captures visible light and a camera that captures infrared light as separate bodies, or a single camera switches between an operation of capturing visible light and an operation of capturing infrared light. It may be a thing.

  FIG. 2 is a hardware configuration example of the vehicle image recognition apparatus 1. The vehicle image recognition device 1 includes, for example, a CPU (Central Processing Unit) 10, a drive device 12, an auxiliary storage device 16, a memory device 18, and an interface device 20.

  The CPU 10 is a processor having, for example, a program counter, an instruction decoder, various arithmetic units, an LSU (Load Store Unit), a general-purpose register, and the like. The drive device 12 is a device that can read a program and data from the storage medium 14. When the storage medium 14 on which the program is recorded is loaded into the drive device 12, the program is installed from the storage medium 14 to the auxiliary storage device 16 via the drive device 12. The storage medium 14 is a portable storage medium such as a CD (Compact Disc), a DVD (Digital Versatile Disc), or a USB (Universal Serial Bus) memory. The auxiliary storage device 16 is, for example, an HDD (Hard Disk Drive), a flash memory, a memory card, or an EEPROM (Electrically Erasable and Programmable Read Only Memory). In addition to using the storage medium 14 as described above, the program executed by the CPU 10 can be installed by the interface device 20 being downloaded from another computer via the network and installed in the auxiliary storage device 16. When the CPU 10 executes a program that is installed or stored in advance, the information processing apparatus of the aspect shown in FIG. 2 can function as the vehicle image recognition apparatus 1 of the present embodiment. The memory device 18 is, for example, a RAM (Random Access Memory). The interface device 20 performs processing for acquiring image data from the camera 100 and storing it in the memory device 18 and the like in addition to connecting to the network.

[Function configuration]
FIG. 3 is a functional configuration example of the vehicle image recognition apparatus 1. The vehicle image recognition device 1 includes a vehicle image region extraction unit 30, a plate image region extraction unit 32, a number recognition unit 34, a luminance difference calculation unit 36, a luminance difference rate calculation unit 38, a plate cover determination unit 40, and a determination result storage unit 42. And a standard average luminance calculation unit 44. These functional blocks function when the CPU 10 executes a program. Note that these functional blocks do not need to be realized by a clearly separated program, and may be called by other programs as subroutines or functions. Further, a part of the functional blocks may be hardware means such as an LSI (Large Scale Integrated circuit), an IC (Integrated Circuit), or an FPGA (Field Programmable Gate Array).

  Further, the vehicle image recognition device 1 generates a recognized vehicle database 60 and a standard average luminance database 62 on the auxiliary storage device 16 or the memory device 18 and uses them for processing.

  The vehicle image region extraction unit 30 performs edge detection processing on the image data stored in the memory device 18 or the like, and generates an edge detection image. An edge is, for example, a pixel whose luminance difference with an adjacent pixel is a predetermined value or more. The edge may be detected by a filter process such as a known SOBEL filter. FIG. 4 is a diagram illustrating an edge detection image generated from the captured image. FIG. 4A shows a captured image, and FIG. 4B shows an edge detection image. The vehicle image region extraction unit 30 further performs pattern matching processing on the edge detection image and extracts a vehicle image region occupied by the vehicle in the image.

  The plate image region extraction unit 32 extracts a plate image region occupied by the license plate in the vehicle image region extracted by the vehicle extraction unit 30. For example, the plate image region extraction unit 32 first extracts a region where a license plate is likely to exist in the vehicle image region as a search region, and extracts a region where edges are arranged in a rectangle in the search region as a plate image region. When the plate image area cannot be extracted within the search area, the plate image area extraction unit 32 extracts the plate image area outside the search area. The plate image area extraction unit 32 analyzes a portion corresponding to the plate image area in the captured image, acquires the plate size and the plate color, and stores them in the memory device 18 or the like.

  The number recognition unit 34 recognizes characters, numbers, and symbols (−) written on the number plate in the plate image region extracted by the plate image region extraction unit 32. FIG. 5 is a diagram illustrating the configuration of a license plate in Japan. As shown in the figure, on the license plate, a classification number, a kana character, and a series designation number such as a transportation branch office are described in a color different from the background area. On the license plate, for example, letters and the like are described in a color such as green on a white background, white on a green background, yellow on a black background, black on a yellow background, white on a blue background, and the like. Hereinafter, the color type is referred to as “color type”. Since such a color contrast appears as an edge, in the edge detection image generated from the captured image, the edge points are arranged on the outlines of characters, numbers, and symbols.

  The number recognizing unit 34 performs scale reduction, inclination correction, etc. on the plate image area, matches the font pattern held in the auxiliary storage device 16 and the like with the outline recognized from the edge point, and the number plate Recognize the letters, numbers and symbols described in. The number recognizing unit 34 not only recognizes what the characters and numbers written on the license plate are, but also uses a contour recognized from the edge points to indicate a region (hereinafter referred to as a character, number, or symbol) in the image. , Referred to as “character area”).

  Various methods have already been known for vehicle recognition and character recognition in images, and the vehicle image recognition device 1 of the present embodiment is not limited to the above-described method, and any method may be used. .

  The luminance difference calculation unit 36 calculates the average luminance for each of the character etc. region specified by the number recognition unit 34 and the background region obtained by removing the character etc. region from the plate image area. The average luminance is obtained, for example, by summing the luminance values of the pixels included in each region and dividing by the number of pixels included in each region.

  The luminance difference calculation unit 36 further calculates a luminance difference which is a difference between the calculated average luminance of the character area and the average luminance of the background area. The luminance difference calculation unit 36 may use a value obtained by subtracting the average luminance of the background area from the average luminance of the character area or the like, or a value obtained by subtracting the average luminance of the character area from the average luminance of the background area. It is good. In addition, the luminance difference calculation unit 36 may use the absolute value of any of the above values as the luminance difference. Hereinafter, the luminance difference is assumed to be an absolute value.

  The luminance difference rate calculation unit 38 is attached to the license plate by obtaining a ratio between the luminance difference calculated by the luminance difference calculation unit 36 and the standard luminance difference obtained from the standard average luminance stored in the standard average luminance database 62. An index value indicating the light transmittance of the obtained plate cover is calculated. The luminance difference rate calculation unit 38 calculates an index value by, for example, dividing the luminance difference calculated by the luminance difference calculation unit 36 by the standard luminance difference. Hereinafter, the index value is referred to as “luminance difference rate”.

  The standard luminance difference is, for example, as described later, a standard average luminance difference obtained by averaging the average luminance of the character area and the average luminance of the background area for a vehicle that is determined not to have a plate cover attached.

  The plate cover determination unit 40 determines that the plate cover is attached to the license plate of the vehicle imaged by the camera 100 when the luminance difference rate is less than the threshold value. Moreover, the plate cover determination part 40 determines with the plate cover not being attached to the number plate of the vehicle imaged with the camera 100, when a brightness | luminance difference rate is more than a threshold value.

  FIG. 6 is a diagram comparing the luminance difference ratios calculated for the license plate without the plate cover and the license plate with the plate cover, respectively, with the threshold value α. FIG. 6A shows a comparison between the luminance difference rate calculated for the license plate without the plate cover and the threshold value α, and FIG. 6B shows the calculation for the license plate with the plate cover attached. A comparison of the luminance difference rate and the threshold value α is shown. Note that when the standard luminance difference is not set according to the imaging conditions as described later, the plate cover determination unit 40 may set different thresholds according to the imaging conditions.

  In addition, the plate cover determination unit 40 does not simply compare the luminance difference rate with the threshold value. For example, the plate cover attaches a statistically deviated luminance difference among a plurality of luminance difference rates with common imaging conditions. It may be determined that this is a luminance difference of the vehicle. FIG. 7 is a diagram illustrating a state where plate cover determination is performed by performing a statistical analysis on a plurality of luminance difference rates. The plate cover determination unit 40 generates, for example, a histogram of the luminance difference rate and the appearance frequency, and determines that a plate having a luminance difference rate higher than the gap is “no plate cover” when a gap or the like exists in the data distribution. The lower one may be determined as “with plate cover”.

  The determination result storage unit 42 includes a determination result including characters and numbers recognized by the number recognition unit 34 and a luminance difference rate calculated by the luminance difference rate calculation unit 38 for the vehicle determined to have a plate cover attached. Is stored as the recognized vehicle database 60.

  FIG. 8 is an example of data stored as the recognized vehicle database 60. The recognition vehicle database 60 stores a recognition ID for identifying a recognition result, a recognition date and time (may be an imaging date and time), number information, presence / absence of a plate cover, light transmittance (luminance difference rate), and the like. The number information includes, for example, a transportation branch office recognized by the number recognition unit 34, a classification number, kana characters, a series designation number, a plate size and a plate color recognized by the plate image region extraction unit 32.

  The standard average brightness calculation unit 44 calculates a standard average brightness by averaging the average brightness of the character area and the background area related to the vehicle determined by the plate cover determination unit 40 that the plate cover is not attached. Stored in the average luminance database 62. For example, if the average value is n in the past, the existing average value is E (n), the average value calculated this time is E (n + 1), and the value added to the current calculation is A (n + 1), for example, Is obtained by the following equation (1).

  E (n + 1) = {E (n) × n + A (n + 1)} / (n + 1) (1)

  Here, since the standard average luminance is assumed to vary depending on the color of the license plate and the imaging conditions, for example, it is calculated for each color type, for each time zone, and for each weather. Note that data obtained by conducting an experiment or the like at the place where the camera 100 is installed may be used as initial data for a period in which the original data of the standard average luminance is not sufficiently accumulated.

  The standard average luminance calculation unit 44 can recognize the time zone with reference to, for example, a built-in timer. In addition, the standard average luminance calculation unit 44 can acquire information regarding the weather via the interface device 20 and the network.

  In addition, the standard average luminance calculation unit 44 calculates the illuminance of the environment based on the average luminance of the entire visible light image captured by the camera 100, and the standard illuminance is not standard for every time zone or every weather but for every “environmental illuminance”. Average brightness may be calculated. Further, the standard average luminance calculation unit 44 may calculate the standard average luminance for each environmental illuminance based on the illuminance obtained from the output of the illuminance sensor or the like.

  As described above, the standard average luminance calculation unit 44 calculates the standard average luminance difference and the standard luminance difference for each time zone and weather that affect the entire captured image of the camera 100, or for each environment illuminance. In the following description, the vehicle image recognition device 1 calculates and sets the standard average luminance difference and the standard luminance difference for each time zone and each weather. FIG. 9 is an example of data stored in the standard average luminance database 62 by the standard average luminance calculation unit 44.

[flowchart]
Hereinafter, the flow of processing executed by the vehicle image recognition device 1 of the first embodiment will be described. FIG. 10 is an example of a flowchart showing a flow of processing executed by the vehicle image recognition device 1 of the first embodiment. This flowchart is executed each time image data is input from the camera 100, for example.

  First, the vehicle image area extraction unit 30 reads image data stored in the memory device 18 or the like (S200), and extracts a vehicle image area (S202).

  Subsequently, the plate image region extraction unit 32 extracts a plate image region in the vehicle image region (S204), and the number recognition unit 34 recognizes a number (character, number, symbol) in the plate image region (S206). ).

  The luminance difference calculation unit 36 calculates an average luminance for each of the character region and the background region with respect to the plate image region, and calculates a luminance difference between the average luminance of the character region and the average luminance of the background region (S208). .

  Next, the luminance difference rate calculation unit 38 calculates the luminance difference rate (S210), and the plate cover determination unit 40 determines whether the luminance difference rate is less than the threshold value (S212).

  When the luminance difference rate is less than the threshold value, the plate cover determination unit 40 determines that the plate cover is attached to the vehicle license plate included in the image data (with plate cover) (S214). In this case, the determination result storage unit 42 stores the determination result, the luminance difference rate (light transmittance), and other items in the recognized vehicle database 60 (S216). The determination result storage unit 42 may transmit the same data to a monitoring center or the like via a network when storing the data in the recognized vehicle database 60.

  On the other hand, when the plate cover determination unit 40 determines that the luminance difference rate is equal to or greater than the threshold (no plate cover), the standard average luminance calculation unit 44 calculates the standard average luminance and updates the standard average luminance database 62. (S218).

  If the number recognition fails, that is, if the character area and the background area cannot be distinguished, the determination result storage unit 42 stores the reference information in the recognized vehicle database 60. The reference information may include information that can be acquired, such as a recognition date and time, a luminance difference rate within a computable range, and the like.

[Summary]
The vehicle image recognition apparatus 1 according to the first embodiment described above is an index value indicating the light transmittance of the plate cover by dividing the luminance difference between the average luminance of the character portion and the average luminance of the background portion by the standard luminance difference. The luminance difference rate can be calculated.

  Further, the vehicle image recognition device 1 of the first embodiment attaches a plate cover to the license plate of the imaged vehicle by comparing the luminance difference rate with a threshold value or performing statistical processing on the luminance difference rate. It can be determined whether or not.

<Second embodiment>
Hereinafter, a vehicle image recognition device, a vehicle image recognition system, a vehicle image recognition method, and a vehicle image recognition program according to a second embodiment of the present invention will be described with reference to the drawings.

  Since the hardware configuration is the same as that of the first embodiment, the same reference numerals are given to the respective components, and the description thereof is omitted.

[Function configuration]
FIG. 11 is a functional configuration example of the vehicle image recognition device 2 according to the second embodiment of the present invention. The vehicle image recognition device 2 includes an area dividing unit 46 in addition to the components included in the vehicle image recognition device 1 of the first embodiment. Each functional block shown in FIG. 11 functions when the CPU 10 executes a program. Note that these functional blocks do not need to be realized by a clearly separated program, and may be called by other programs as subroutines or functions. Further, a part of the functional blocks may be hardware means such as LSI, IC, FPGA or the like.

  In addition, the vehicle image recognition device 2 generates a recognized vehicle database 60 and a standard average luminance database 62 on the auxiliary storage device 16 or the memory device 18 and uses them for processing.

  In addition to extracting the vehicle image area occupied by the vehicle in the image, the vehicle image area extraction unit 30 according to the second embodiment, based on the size and shape of the vehicle image area, Or not).

  The area dividing unit 46 divides the plate image area extracted by the plate image area extracting unit 32 into a plurality of areas.

  The luminance difference calculation unit 36 according to the second embodiment calculates the average luminance of the character area, the average luminance of the background region, and the luminance difference for each region divided by the region dividing unit 46. And the brightness | luminance difference calculation part 36 which concerns on 2nd Example outputs the largest brightness | luminance difference among the calculated some brightness differences to the brightness | luminance difference rate calculation part 38 as a brightness | luminance difference about a corresponding vehicle.

  Hereinafter, the reason why the vehicle image recognition device 2 of the second embodiment performs such processing will be described. The area dividing unit 46 determines an area dividing mode based on conditions such as an imaging time zone, whether or not the headlight is turned on, the relative position of the license plate and the vehicle, and the size type of the vehicle.

[At night]
FIG. 12 shows how the brightness difference of the plate image area increases and decreases and the plate image area division mode corresponding to this when the imaging time zone is night and the imaged vehicle lights up the headlight. FIG. At night, the brightness difference in the plate image area may be partially reduced due to the influence of light from the headlight.

  FIG. 12A shows a vehicle in which a license plate is attached to the central portion of the front portion of the vehicle, a plate image region in which the luminance difference is partially reduced due to the light of the headlight, and a region corresponding thereto. It is an example of division. As shown in the drawing, in the case of a vehicle in which the license plate is attached to the central portion of the front surface portion of the vehicle, the luminance difference is small at both ends near the headlight in the plate image area. Therefore, in order to obtain the original luminance difference, it is desirable to calculate the luminance difference at the center of the plate image area.

  For this reason, as shown in FIG. 12 (A), the area dividing unit 46 is configured to display the plate when (i) the imaging time zone is nighttime and (ii) the plate image area exists in the center of the vehicle front part. The image area is divided into three areas A, B, and C, for example, with two boundary lines set in the vertical direction. As shown in FIG. 4, the “vehicle front portion” can be identified with the vehicle image area when the camera 100 captures the vehicle from the front. The extraction unit 30 performs pattern matching for specifying the vehicle front portion. The area dividing unit 46 may regard the vertical direction of the image as the horizontal direction when the camera 100 is installed without a roll angle, but if the camera 100 is installed with a roll angle, It is preferable to set a vertical boundary line by performing processing such as correction. Here, the roll angle refers to the rotation angle of the camera 100 around the optical axis of the camera 100.

  On the other hand, FIG. 12B shows a vehicle with a license plate attached to the end, a plate image area in which the brightness difference is partially reduced by the light of the headlight, and an area division corresponding to this. It is an example. As shown in the figure, in the case of a vehicle in which a license plate is attached to the end portion, the luminance difference becomes smaller on the side closer to the headlight in the plate image area. Therefore, in order to obtain the original luminance difference, it is desirable to calculate the luminance difference on the side farther from the headlight in the plate image area.

  For this reason, as shown in FIG. 12B, the area dividing unit 46 is configured such that when (i) the imaging time zone is nighttime and (iii) the plate image area is present at the end of the front surface of the vehicle, For example, the image area is divided into two areas A and B by a boundary line set in the vertical direction.

  The region dividing unit 46 performs, for example, the following processing in order to determine whether the plate image region is present at the center or the end of the front surface of the vehicle. FIG. 13 is an explanatory diagram for explaining a state in which it is determined whether the plate image region is present at the center portion or the end portion of the front surface portion of the vehicle. For example, the region dividing unit 46 calculates the amount of deviation x in which the barycentric point G of the plate image region N deviates from the intermediate point M between both ends of the front surface of the vehicle in the horizontal direction. For example, if the value obtained by dividing the displacement amount x by the distance L between the end portion of the vehicle front portion and the intermediate point is equal to or greater than the reference value, the region dividing unit 46 determines that the displacement amount x exists at the end portion of the vehicle front portion. If the value obtained by dividing x by the distance L is less than the reference value, it is determined that the vehicle exists at the center of the front surface of the vehicle.

  As described above, the vehicle image recognition device 2 according to the second embodiment is an area where the influence of the light of the headlight is small by dividing the plate image area by the boundary line in the vertical direction when the imaging time zone is nighttime. Can produce. As a result, the vehicle image recognition device 2 can suppress the luminance difference calculation unit 36 from calculating a luminance difference smaller than the original value due to the influence of the light of the headlight.

  Further, the vehicle image recognition device 2 of the second embodiment changes the manner in which the plate image area is divided depending on whether the plate image area is present at the center or at the end of the front surface of the vehicle. The influence of the light on the luminance difference rate can be further reduced.

  FIG. 14 is a diagram illustrating a state in which the maximum value of the luminance difference rate is equal to or greater than the threshold value α as a result of the region dividing unit 46 dividing the plate image region.

  FIG. 14A shows the luminance difference ratio and the threshold value calculated for each of the areas A, B, and C when the imaging time zone is nighttime and the plate image area exists in the center of the front surface of the vehicle. It is the figure which compared (alpha). As shown in the figure, the luminance difference rate calculated for the regions A and C may be less than the threshold value α due to the influence of the light of the headlight, but the luminance difference rate calculated for the region B is not less than the threshold value α. Is likely to be. Since the luminance difference calculation unit 36 outputs the maximum luminance difference among the plurality of luminance differences to the luminance difference rate calculation unit 38, the luminance difference rate calculation unit 38 calculates the luminance difference rate for the region B. . As a result, the plate cover determination unit 40 determines “no plate cover”.

  FIG. 14B shows the luminance difference rate and the threshold value α calculated for each of the areas A and B when the imaging time zone is nighttime and the plate image area exists at the end of the front surface of the vehicle. It is the figure compared. As shown in the figure, the luminance difference rate calculated for the region B may be less than the threshold value α due to the influence of the light of the headlight, but the luminance difference rate calculated for the region A may be equal to or greater than the threshold value α. High nature. Since the luminance difference calculation unit 36 outputs the maximum luminance difference among the plurality of luminance differences to the luminance difference rate calculation unit 38, the luminance difference rate calculation unit 38 calculates the luminance difference rate for the region A. . As a result, the plate cover determination unit 40 determines “no plate cover”.

  Thus, the vehicle image recognition device 2 of the second embodiment can reduce the influence of the light of the headlight when determining whether or not the plate cover is attached.

  In addition, instead of or in addition to the condition that “the imaging time zone is at night”, the area dividing unit 46 uses the headlight lighting detection unit 45 to indicate that the imaged vehicle is lighting the headlight. The condition “detected” may be used. For example, the headlight lighting detection unit 45 determines that the vehicle is lighting the headlight when the maximum luminance in the captured image of the camera 100 is equal to or higher than a reference value and the imaging time zone is nighttime. Also in this case, the vehicle image recognition device 2 can achieve the same effect as described above.

[In the case of daytime]
FIG. 15 is a diagram showing how the brightness difference of the plate image area increases and decreases due to the influence of sunlight when the imaging time zone is during the day, and how the plate image area is divided correspondingly. During the daytime, the brightness difference in the plate image area may be partially reduced by the sun light shining from above being blocked by the vehicle bumper to create a shadow.

  As shown in FIG. 15, when the imaging time zone is during the daytime, the luminance difference is small in the upper part of the plate image area. Therefore, in order to obtain the original luminance difference, it is desirable to calculate the luminance difference in the lower part of the plate image area.

  For this reason, as shown in FIG. 15, the area dividing unit 46 divides the plate image area into areas A and B with a border line set in the horizontal direction, for example, when the imaging time zone is during the daytime. Divide into two. Note that, when the camera 100 is installed without a roll angle with respect to the horizontal direction, the area dividing unit 46 may regard the horizontal direction of the image as the horizontal direction, but is installed with the roll angle. If it is, it is preferable to set a horizontal boundary line by performing processing such as roll correction.

  As described above, the vehicle image recognition apparatus 2 according to the second embodiment has a small influence of the shadow produced by sunlight by dividing the plate image region by the horizontal boundary line when the imaging time zone is daytime. An area can be created. As a result, the vehicle image recognition device 2 can suppress the luminance difference calculation unit 36 from calculating a luminance difference smaller than the original value due to the influence of a shadow created by sunlight.

  FIG. 16 is a diagram comparing the luminance difference rate calculated for each of the regions A and B with the threshold value α when the imaging time zone is daytime. As shown in the figure, the luminance difference rate calculated for the region A may be less than the threshold value α due to the influence of the shadow, but the luminance difference rate calculated for the region B is likely to be equal to or higher than the threshold value α. . Since the luminance difference calculation unit 36 outputs the maximum luminance difference among the plurality of luminance differences to the luminance difference rate calculation unit 38, the luminance difference rate calculation unit 38 calculates the luminance difference rate for the region B. . As a result, the plate cover determination unit 40 determines “no plate cover”.

  Thus, the vehicle image recognition device 2 of the second embodiment can reduce the influence of the shadow created by sunlight when determining whether or not the plate cover is attached.

  In addition, when the area dividing unit 46 distinguishes between daytime and nighttime, it is preferable to use different determination criteria depending on the season.

[When the vehicle is a large vehicle]
FIG. 17 is a diagram showing how the brightness difference of the plate image area increases and decreases due to dirt on the lower side of the license plate and the corresponding division of the plate image area when the imaged vehicle is a large vehicle. is there. FIG. 17 is based on the assumption that the imaging time zone is daytime. In the case of a large vehicle, the lower side of the number plate may be particularly stained due to long-time running or the like.

  For this reason, the region dividing unit 46 (v) divides the plate image region into regions by, for example, a boundary set in the horizontal direction when the imaged vehicle is a large vehicle. In the example of FIG. 17, since the imaging time zone is daytime, the region dividing unit 46 divides the plate image region into regions A, B, and C, for example, by two boundary lines set in the horizontal direction.

  As described above, the vehicle image recognition apparatus 2 according to the second embodiment is affected by the contamination of the license plate by dividing the plate image area by the horizontal boundary line when the captured vehicle is a large vehicle. A small area can be created. As a result, the vehicle image recognition device 2 can suppress the luminance difference calculation unit 36 from calculating a luminance difference smaller than the original value due to the influence of the dirt on the license plate.

  FIG. 18 is a diagram comparing the luminance difference rate calculated for each of the regions A, B, and C with the threshold value α when the imaging time zone is daytime and the imaged vehicle is a large vehicle. is there. As shown in the figure, the luminance difference rate calculated for the regions A and C may be less than the threshold value α due to the influence of shadows or dirt, but the luminance difference rate calculated for the region B is greater than or equal to the threshold value α. Probability is high. Since the luminance difference calculation unit 36 outputs the maximum luminance difference among the plurality of luminance differences to the luminance difference rate calculation unit 38, the luminance difference rate calculation unit 38 calculates the luminance difference rate for the region B. . As a result, the plate cover determination unit 40 determines “no plate cover”.

  As described above, the vehicle image recognition device 2 according to the second embodiment can reduce the influence of the shadow created by sunlight and the contamination of the license plate when determining whether or not the plate cover is attached.

[flowchart]
Hereinafter, the flow of processing executed by the vehicle image recognition device 2 of the second embodiment will be described. 19 and 20 are an example of a flowchart showing the flow of processing executed by the vehicle image recognition device 2 of the second embodiment. This flowchart is executed each time image data is input from the camera 100, for example.

  First, the vehicle image area extraction unit 30 reads image data stored in the memory device 18 or the like (S300), and extracts a vehicle image area (S302).

  Subsequently, the plate image region extraction unit 32 extracts a plate image region in the vehicle image region (S304), and the number recognition unit 34 recognizes a number (characters, numbers, symbols) in the plate image region (S306). ).

  Next, the area dividing unit 46 performs area division (S308 to S320; see FIG. 20). The area dividing unit 46 first determines whether or not the imaging time zone is daytime (S308).

  When the imaging time zone is daytime, the area dividing unit 46 determines whether the imaged vehicle is a large vehicle (S310).

  When the imaged vehicle is a large vehicle, the area dividing unit 46 divides the plate image area into three at the boundary line set in the horizontal direction (S312). When the imaged vehicle is not a large vehicle, the area dividing unit 46 divides the plate image area into two at the boundary line set in the horizontal direction (S314).

  On the other hand, when the imaging time zone is nighttime instead of daytime, the region dividing unit 46 determines whether or not the position of the plate image region is present at the center of the vehicle front portion (S316).

  When the position of the plate image area is present at the center of the front surface of the vehicle, the area dividing unit 46 divides the plate image area into three at the boundary line set in the vertical direction (S318). When the position of the plate image area is present at the end of the front surface of the vehicle, the area dividing unit 46 divides the plate image area into two at the boundary line set in the vertical direction (S320).

  The luminance difference calculation unit 36 determines whether or not the outline of the character or number is recognizable in the entire region for each region divided by the processes of S308 to S320 (S322_1). As for the outline of a character or number, for example, the number recognition result by the number recognition unit 34 is used. That is, a line connecting edge points extracted when the number is recognized by the number recognizing unit 34 is recognized as a contour of a character or a number. For example, the luminance difference calculation unit 36 treats an area where the outline is not interrupted in the area as an area where the outline of a character or a number can be recognized.

  When the outline of the character or number is recognizable in the entire area, the brightness difference calculating unit 36 calculates the brightness difference by calculating the average brightness of the character area and the background area for the entire area. The luminance difference that is the maximum among the luminance differences is selected (S322_2).

  If the outline of the character or number is not recognizable in the entire area, the luminance difference calculation unit 36 determines whether the outline of the character or number is recognizable in a partial area (S322_3).

  When the outline of a character or number intersects the boundary line of the region, a situation may occur in which the contour can be recognized in a certain region and not recognized in another region even if the characters or numbers are the same. In addition, although a contour can be recognized for characters and numbers that fall within a certain area, a situation in which a contour cannot be recognized for characters and numbers that fall within another area may occur. Even in such a situation, a significant luminance difference can be calculated by selecting an area where the contour can be recognized.

  When the outline of a character or number is recognizable in a partial area, the luminance difference calculation unit 36 calculates the luminance difference by calculating the average luminance of the character area and the background area for the partial area. Then, the brightness difference that is the maximum value among the calculated brightness differences is selected (S322_4).

  The luminance difference calculation unit 36 stores the reference information in the recognized vehicle database 60 when the outline of the character or number is not recognizable even in a partial region (S322_5).

  Next, based on the luminance difference (maximum value) selected by the luminance difference calculation unit 36, the luminance difference rate calculation unit 38 calculates the luminance difference rate (S324), and the plate cover determination unit 40 determines that the luminance difference rate is a threshold value. It is determined whether it is less than (S326).

  When the luminance difference rate is less than the threshold value, the plate cover determination unit 40 determines that the plate cover is attached to the vehicle license plate included in the image data (with plate cover) (S328). In this case, the determination result storage unit 42 stores the determination result, the luminance difference rate (light transmittance), and other matters in the recognized vehicle database 60 (S330). The determination result storage unit 42 may transmit the same data to a monitoring center or the like via a network when storing the data in the recognized vehicle database 60.

  On the other hand, when the plate cover determination unit 40 determines that the luminance difference rate is equal to or greater than the threshold (no plate cover), the standard average luminance calculation unit 44 calculates the standard average luminance and updates the standard average luminance database 62. (S332).

[Other flowcharts]
In the case where the region dividing unit 46 uses the condition that “the imaged vehicle is turning on the headlight” in addition to the condition that “the imaging time zone is nighttime”, the flowchart of FIG. Changed to FIG. 21 is another example of a flowchart showing a flow of processing executed by the vehicle image recognition device 2 of the second embodiment.

  The area dividing unit 46 first determines whether or not the imaging time zone is daytime (S308).

  When the imaging time zone is daytime, the area dividing unit 46 determines whether the imaged vehicle is a large vehicle (S310).

  When the imaged vehicle is a large vehicle, the area dividing unit 46 divides the plate image area into three at the boundary line set in the horizontal direction (S312). When the imaged vehicle is not a large vehicle, the area dividing unit 46 divides the plate image area into two at the boundary line set in the horizontal direction (S314).

  On the other hand, when the imaging time zone is nighttime instead of daytime, the area dividing unit 46 determines whether or not the headlight of the imaged vehicle is turned on (S315).

  When the imaged vehicle headlight is turned on, the region dividing unit 46 determines whether or not the position of the plate image region is present at the center of the front surface of the vehicle (S316).

  When the position of the plate image area is present at the center of the front surface of the vehicle, the area dividing unit 46 divides the plate image area into three at the boundary line set in the vertical direction (S318). When the position of the plate image area is present at the end of the front surface of the vehicle, the area dividing unit 46 divides the plate image area into two at the boundary line set in the vertical direction (S320).

  When the imaged vehicle headlight is not turned on, the area dividing unit 46 does not perform area division, for example.

  When the process shown in FIG. 21 is completed, the process from S322 onward in FIG. 19 is executed.

[Summary]
The vehicle image recognition apparatus 2 according to the second embodiment described above can calculate the luminance difference rate that is an index value indicating the light transmittance of the plate cover, as in the first embodiment. It can be determined whether or not a plate cover is attached to the plate.

  Furthermore, in the vehicle image recognition apparatus 2 of the second embodiment, since the area dividing unit 46 divides the plate image area based on various conditions, the influence of headlight light, shadows created by sunlight, dirt on the license plate, etc. Can create a small area. For this reason, the vehicle image recognition apparatus 2 of 2nd Example can suppress calculating a luminance difference smaller than an original value. As a result, the vehicle image recognition device 2 of the second embodiment can more accurately determine whether or not the plate cover is attached.

<Third embodiment>
Hereinafter, a vehicle image recognition device, a vehicle image recognition system, a vehicle image recognition method, and a vehicle image recognition program according to a third embodiment of the present invention will be described with reference to the drawings.

  Since the hardware configuration is the same as that of the first embodiment, the same reference numerals are given to the respective components, and the description thereof is omitted.

[Function configuration]
FIG. 22 is a functional configuration example of the vehicle image recognition device 3 according to the third embodiment of the present invention. The vehicle image recognition device 3 includes a region selection unit 48 in addition to the components included in the vehicle image recognition device 1 of the first embodiment. Each functional block shown in FIG. 22 functions when the CPU 10 executes a program. Note that these functional blocks do not need to be realized by a clearly separated program, and may be called by other programs as subroutines or functions. Further, a part of the functional blocks may be hardware means such as LSI, IC, FPGA or the like.

  In addition, the vehicle image recognition device 3 generates a recognized vehicle database 60 and a standard average luminance database 62 on the auxiliary storage device 16 or the memory device 18 and uses them for processing.

  The vehicle image area extraction unit 30 according to the third embodiment determines the size type of the vehicle (particularly whether the vehicle is a large vehicle) as in the second embodiment.

  The region selection unit 48 selects a specific region from the plate image regions extracted by the plate image region extraction unit 32.

  The luminance difference calculation unit 36 according to the third embodiment calculates the average luminance of the character area, the average luminance of the background region, and the luminance difference for the region selected by the region selection unit 48.

  The reason why the vehicle image recognition device 3 of the third embodiment performs such processing is the same as that of the second embodiment. The area selection unit 48 selects a specific area based on conditions such as the imaging time zone, the presence / absence of lighting of the headlight, the relative position of the license plate and the vehicle, and the vehicle size type.

[At night]
The area selection unit 48 is a boundary in which two plate image areas are set in the vertical direction, for example, when (i) the imaging time zone is nighttime and (ii) the plate image area exists in the center of the vehicle front part. Divide the line in three and select the center area. About the reason which performs the process which concerns, FIG. 12 (A) and description of 2nd Example are used, and description is abbreviate | omitted.

  In addition, the area selection unit 48 sets one plate image area, for example, in the vertical direction when (i) the imaging time zone is nighttime and (iii) and the plate image area exists at the end of the front surface of the vehicle. Divide into two at the boundary line, and select the area far from the headlight. Here, the area far from the headlight means an area far from the headlight “closer to the license plate”. About the reason which performs the process which concerns, description is abbreviate | omitted as using FIG. 12 (B) and description of 2nd Example.

  As described above, the vehicle image recognition apparatus 3 according to the third embodiment divides the plate image area by the vertical boundary line when the imaging time zone is nighttime, and selects an area where the influence of the light of the headlight is small. To do. As a result, the vehicle image recognition device 3 can suppress the luminance difference calculation unit 36 from calculating a luminance difference smaller than the original value due to the influence of the light of the headlight.

  Further, the vehicle image recognition device 3 according to the third embodiment changes the mode of dividing the plate image area and the area to be selected depending on whether the plate image area exists at the center or the end of the front surface of the vehicle. . As a result, the vehicle image recognition device 3 of the third embodiment can further reduce the influence of the headlight light on the luminance difference rate.

  By these, the vehicle image recognition apparatus 3 of 3rd Example can reduce the influence which the light of a headlight gives, when determining whether the plate cover is attached.

  In addition, as in the second embodiment, the area selection unit 48 may be replaced with, or in addition to, the condition that “the imaging time zone is at night”. The condition “detected by the headlight lighting detection unit 45” may be used.

[In the case of daytime]
The area selection unit 48 (iv) divides the plate image area into areas A and B, for example, by a border line set in the horizontal direction when the imaging time zone is during the daytime. select. About the reason which performs the process which concerns, description is abbreviate | omitted as using the description of FIG. 15 and 2nd Example.

  As described above, the vehicle image recognition device 3 according to the third embodiment selects the lower region by dividing the plate image region by the horizontal boundary line when the imaging time zone is daytime. It is possible to select a region where the influence of the shadow created by sunlight is small. As a result, the vehicle image recognition device 3 can suppress the luminance difference calculation unit 36 from calculating a luminance difference smaller than the original value due to the influence of a shadow created by sunlight.

  Thereby, the vehicle image recognition device 3 of the third embodiment can reduce the influence of the shadow created by sunlight when determining whether or not the plate cover is attached.

[When the vehicle is a large vehicle]
When the imaged vehicle is a large vehicle, the area selection unit 48 divides the plate image area into areas by, for example, a boundary set in the horizontal direction, and selects an area other than the lowermost area. About the reason for performing such a process, description is abbreviate | omitted as using the description of FIG. 17 and 2nd Example.

  As described above, when the captured vehicle is a large vehicle, the vehicle image recognition device 3 according to the third embodiment divides the plate image region by the horizontal boundary line and selects the region other than the lowermost side. The area where the influence of the dirt on the license plate is small can be selected. As a result, the vehicle image recognition device 3 can suppress the luminance difference calculation unit 36 from calculating a luminance difference smaller than the original value due to the influence of the dirt on the license plate.

  As a result, the vehicle image recognition device 3 of the third embodiment can reduce the influence of shadows created by sunlight and dirt on the license plate when determining whether or not the plate cover is attached.

[flowchart]
Hereinafter, the flow of processing executed by the vehicle image recognition device 3 of the third embodiment will be described. 23 and 24 are an example of a flowchart showing the flow of processing executed by the vehicle image recognition device 3 of the third embodiment. This flowchart is executed each time image data is input from the camera 100, for example.

  First, the vehicle image area extraction unit 30 reads image data stored in the memory device 18 or the like (S400), and extracts a vehicle image area (S402).

  Subsequently, the plate image area extraction unit 32 extracts a plate image area in the vehicle image area (S404), and the number recognition unit 34 recognizes a number (character, number, symbol) in the plate image area (S406). ).

  Next, the region selection unit 48 performs region selection (S408 to S420; see FIG. 24). The area selection unit 48 first determines whether or not the imaging time zone is daytime (S408).

  When the imaging time zone is daytime, the region selection unit 48 determines whether the imaged vehicle is a large vehicle (S410).

  When the imaged vehicle is a large vehicle, the region selection unit 48 divides the plate image region into three by the boundary line set in the horizontal direction, and selects the central region (S412). When the imaged vehicle is not a large vehicle, the region selection unit 48 divides the plate image region into two at the boundary line set in the horizontal direction, and selects the lower region (S414).

  On the other hand, when the imaging time zone is nighttime instead of daytime, the region selection unit 48 determines whether or not the position of the plate image region is present at the center of the front surface of the vehicle (S416).

  When the position of the plate image area is present at the center of the front part of the vehicle, the area selecting unit 48 divides the plate image area into three by the boundary line set in the vertical direction and selects the center area (S418). . When the position of the plate image area is present at the end of the front part of the vehicle, the area selection unit 48 divides the plate image area into two parts by a border line set in the vertical direction, and selects the area far from the headlight. (S420).

  The brightness difference calculation unit 36 calculates the average brightness of the character area and the background area for the area selected by the processing of S408 to S420 to calculate the brightness difference (S422).

  The luminance difference calculation unit 36 determines whether or not the outline of the character or number is recognizable in the region selected by the processing of S408 to S420 (S422_1). As for the outline of a character or number, for example, the number recognition result by the number recognition unit 34 is used. For example, the luminance difference calculation unit 36 treats an area where the outline is not interrupted in the area as an area where the outline of a character or a number can be recognized.

  The brightness difference calculating unit 36 calculates the brightness difference by calculating the average brightness of the character area and the background area in the selected area when the outline of the character or number is recognizable in the selected area. (S422_2).

  If the outline of the character or number is not recognizable in the selected area, the luminance difference calculation unit 36 determines whether the outline of the character or number is recognizable in another area (S422_3).

  If the outline of the character or number is recognizable in another area, the brightness difference calculation unit 36 calculates the average brightness of the character area and the background area in the other area to calculate the brightness difference ( S422_4).

  The brightness difference calculation unit 36 stores the reference information in the recognized vehicle database 60 when the character or number outline is not recognized in the selected region or other regions (S422_5).

  Next, based on the brightness difference calculated by the brightness difference calculation unit 36, the brightness difference rate calculation unit 38 calculates the brightness difference rate (S424), and the plate cover determination unit 40 determines whether the brightness difference rate is less than the threshold value. It is determined whether or not (S426).

  When the luminance difference rate is less than the threshold value, the plate cover determination unit 40 determines that the plate cover is attached to the license plate of the vehicle included in the image data (with plate cover) (S428). In this case, the determination result storage unit 42 stores the determination result, the luminance difference rate (light transmittance), and other items in the recognized vehicle database 60 (S430). The determination result storage unit 42 may transmit the same data to a monitoring center or the like via a network when storing the data in the recognized vehicle database 60.

  On the other hand, when the plate cover determination unit 40 determines that the luminance difference rate is equal to or greater than the threshold (no plate cover), the standard average luminance calculation unit 44 calculates the standard average luminance and updates the standard average luminance database 62. (S432).

[Other flowcharts]
In the case where the region selection unit 48 uses the condition that “the imaged vehicle is turning on the headlight” in addition to the condition that the imaging time zone is nighttime, the flowchart of FIG. Changed to FIG. 25 is another example of a flowchart showing a flow of processing executed by the vehicle image recognition device 3 of the third embodiment.

  The area selection unit 48 first determines whether or not the imaging time zone is daytime (S408).

  When the imaging time zone is daytime, the region selection unit 48 determines whether the imaged vehicle is a large vehicle (S410).

  When the imaged vehicle is a large vehicle, the region selection unit 48 divides the plate image region into three by the boundary line set in the horizontal direction, and selects the central region (S412). When the imaged vehicle is not a large vehicle, the region selection unit 48 divides the plate image region into two at the boundary line set in the horizontal direction, and selects the lower region (S414).

  On the other hand, when the imaging time zone is nighttime instead of daytime, the region selection unit 48 determines whether or not the headlight of the imaged vehicle is turned on (S415).

  When the imaged vehicle headlight is lit, the region selection unit 48 determines whether or not the position of the plate image region is present at the center of the front surface of the vehicle (S416).

  When the position of the plate image area is present at the center of the front part of the vehicle, the area selecting unit 48 divides the plate image area into three by the boundary line set in the vertical direction and selects the center area (S418). . When the position of the plate image area is present at the end of the front part of the vehicle, the area selection unit 48 divides the plate image area into two parts by a border line set in the vertical direction, and selects the area far from the headlight. (S420).

  When the imaged vehicle headlight is not turned on, the region selection unit 48 selects, for example, the entire plate image region (S421).

  When the process shown in FIG. 25 is completed, the process from S422 onward in FIG. 23 is executed.

[Summary]
The vehicle image recognition device 3 of the third embodiment described above can calculate the luminance difference rate that is an index value indicating the light transmittance of the plate cover, as in the first embodiment. It can be determined whether or not a plate cover is attached to the plate.

  Furthermore, in the vehicle image recognition device 3 of the third embodiment, since the region selection unit 48 selects a specific region from the plate image region based on various conditions, the light of the headlight, the shadow created by the sunlight, the license plate It is possible to select an area that is less affected by dirt or the like. For this reason, the vehicle image recognition apparatus 3 of 3rd Example can suppress calculating a luminance difference smaller than an original value. As a result, the vehicle image recognition device 3 of the third embodiment can more accurately determine whether or not the plate cover is attached.

<Fourth embodiment>
Hereinafter, a vehicle image recognition device, a vehicle image recognition system, a vehicle image recognition method, and a vehicle image recognition program according to a fourth embodiment of the present invention will be described with reference to the drawings.

  Since the hardware configuration is the same as that of the first embodiment, the same reference numerals are given to the respective components, and the description thereof is omitted.

[Function configuration]
FIG. 26 is a functional configuration example of the vehicle image recognition device 4 according to the fourth embodiment of the present invention. The vehicle image recognition device 4 includes a character area dividing unit 50 in addition to the components included in the vehicle image recognition device 1 of the first embodiment. Each functional block shown in FIG. 26 functions when the CPU 10 executes a program. Note that these functional blocks do not need to be realized by a clearly separated program, and may be called by other programs as subroutines or functions. Further, a part of the functional blocks may be hardware means such as LSI, IC, FPGA or the like.

  In addition, the vehicle image recognition device 4 generates a recognized vehicle database 60 and a standard average luminance database 62 on the auxiliary storage device 16 or the memory device 18 and uses them for processing.

  The character area dividing unit 50 divides the plate image area extracted from the plate image area into a plurality of areas including characters and numbers.

  The luminance difference calculation unit 36 according to the fourth embodiment calculates the average luminance of the character region, the average luminance of the background region, and the luminance difference for each region divided by the character region dividing unit 50. Then, the luminance difference calculation unit 36 according to the fourth embodiment outputs the maximum luminance difference among the calculated plurality of luminance differences to the luminance difference rate calculation unit 38.

  FIG. 27 is a diagram illustrating comparison between the regions A to G divided by the character region dividing unit 50 and the luminance difference rate calculated for each region and the threshold value α. As shown in the figure, when the number recognition in the areas A and B fails due to the influence of a shadow or the like, the luminance difference rate for the areas A and B may be an error value, but for the other areas C to G, An appropriate luminance difference rate is calculated. The character area dividing unit 50 divides the plate image area into, for example, an area A indicating a transportation branch office, an area B indicating a classification number, an area C indicating a kana character, and areas D to G indicating a series designation number. To divide. Hereinafter, such an area is referred to as a character area. The character area is set, for example, as an area that includes characters and numbers recognized by the recognition processing of the number recognition unit 34 one by one. The character area may be set by an enlargement / reduction process or the like based on the relative position in the license plate that has been previously determined.

[flowchart]
Hereinafter, the flow of processing executed by the vehicle image recognition device 4 of the fourth embodiment will be described. FIG. 28 is an example of a flowchart showing a flow of processing executed by the vehicle image recognition device 4 of the fourth embodiment. This flowchart is executed each time image data is input from the camera 100, for example.

  First, the vehicle image region extraction unit 30 reads image data stored in the memory device 18 or the like (S500), and extracts a vehicle image region (S502).

  Subsequently, the plate image region extraction unit 32 extracts a plate image region in the vehicle image region (S504), and the number recognition unit 34 recognizes a number (characters, numbers, symbols) in the plate image region ( S506).

  Next, the character area dividing unit 50 divides the plate image area into character areas (S508).

  Next, the number recognizing unit 34 determines whether or not the recognition is successful for all the characters and numbers in the recognition processing of S506 (S510).

  When the recognition is not successful for all the characters, numbers, and symbols, the number recognition unit 34 determines whether the recognition is successful for some characters, numbers, and symbols (S512). When the recognition is successful for some characters and numbers, the number recognition unit 34 designates a region where the recognition has failed as a target for calculating the average luminance and the luminance difference (S514).

  The brightness difference calculation unit 36 calculates the brightness difference by calculating the average brightness of the character area and the background area for each area divided in S508 (excluding the area specified in S514). The luminance difference that is the maximum value among the plurality of luminance differences is selected (S516).

  Next, the luminance difference rate calculation unit 38 calculates the luminance difference rate based on the luminance difference (maximum value) selected by the luminance difference calculation unit 36 (S518), and the plate cover determination unit 40 determines that the luminance difference rate is less than the threshold value. It is determined whether or not (S520).

  If the luminance difference rate is less than the threshold value, the plate cover determination unit 40 determines that the plate cover is attached to the license plate of the vehicle included in the image data (with plate cover) (S522). In this case, the determination result storage unit 42 stores the determination result, the luminance difference rate (light transmittance), and other items in the recognized vehicle database 60 (S524). The determination result storage unit 42 may transmit the same data to a monitoring center or the like via a network when storing the data in the recognized vehicle database 60.

  On the other hand, when the plate cover determination unit 40 determines that the luminance difference rate is equal to or greater than the threshold (no plate cover), the standard average luminance calculation unit 44 calculates the standard average luminance and updates the standard average luminance database 62. (S526).

  In S512, when it is determined that recognition has failed for all characters and numbers, the determination result storage unit 42 stores the reference information in the recognized vehicle database 60 (S528). The reference information may include information that can be acquired, such as a recognition date and time, a luminance difference rate within a computable range, and the like.

[Summary]
The vehicle image recognition device 4 of the fourth embodiment described above can calculate the luminance difference rate that is an index value indicating the light transmittance of the plate cover, as in the first embodiment, and the number of the imaged vehicle It can be determined whether or not a plate cover is attached to the plate.

  Furthermore, in the vehicle image recognition device 4 of the fourth embodiment, since the character area dividing unit 50 divides the plate image area, an area that is less affected by headlight light, shadows produced by sunlight, dirt on the license plate, etc. is produced. be able to. For this reason, the vehicle image recognition apparatus 4 of 4th Example can suppress calculating a luminance difference smaller than an original value. As a result, the vehicle image recognition device 4 of the fourth embodiment can more accurately determine whether or not the plate cover is attached.

<Fifth embodiment>
Hereinafter, a vehicle image recognition device, a vehicle image recognition system, a vehicle image recognition method, and a vehicle image recognition program according to a fifth embodiment of the present invention will be described with reference to the drawings.

  Since the hardware configuration is the same as that of the first embodiment, the same reference numerals are given to the respective components, and the description thereof is omitted.

[Function configuration]
FIG. 29 is a functional configuration example of the vehicle image recognition device 5 according to the fifth embodiment of the present invention. The vehicle image recognition device 5 includes a character region selection unit 52 in addition to the components included in the vehicle image recognition device 1 of the first embodiment. Each functional block shown in FIG. 29 functions when the CPU 10 executes a program. Note that these functional blocks do not need to be realized by a clearly separated program, and may be called by other programs as subroutines or functions. Further, a part of the functional blocks may be hardware means such as LSI, IC, FPGA or the like.

  Similar to the character area dividing unit 50 of the fourth embodiment, the character area selecting unit 52 divides the image into a plurality of areas including characters and numbers, selects an area according to the condition, and calculates the average luminance and the luminance difference. Specify the calculation target. The character area selection unit 52 divides the plate image area into an area A indicating a transportation branch office, an area B indicating a classification number, an area C indicating a kana character, and areas D to G indicating a series designation number. (See FIG. 27).

[flowchart]
Hereinafter, the flow of processing executed by the vehicle image recognition device 5 of the fifth embodiment will be described. 30 and 31 are an example of a flowchart showing a flow of processing executed by the vehicle image recognition device 5 of the fifth embodiment. This flowchart is executed each time image data is input from the camera 100, for example.

  First, the vehicle image area extraction unit 30 reads image data stored in the memory device 18 or the like (S600), and extracts a vehicle image area (S602).

  Subsequently, the plate image region extraction unit 32 extracts a plate image region in the vehicle image region (S604), and the number recognition unit 34 recognizes a number (characters, numbers, symbols) in the plate image region (S606). ).

  Next, the character area selection unit 52 performs area selection (S608 to S618). The character area selection unit 52 first determines whether or not the imaging time zone is daytime (S608).

  When the imaging time zone is daytime, the character area selection unit 52 selects areas D to G in which the series designation numbers are indicated (S610).

  On the other hand, when the imaging time zone is nighttime instead of daytime, the character area selection unit 52 determines whether the position of the plate image area exists on the right side, the central part, or the left side of the vehicle front part ( S612). In such determination, the character region selection unit 52 uses the method described with reference to FIG. 13 in the second embodiment, for example.

  When the position of the plate image area is present on the right side of the front part of the vehicle (on the left side in an actual vehicle), the character area selection unit 52 includes an area A indicating a transportation branch office, an area C indicating a kana character, and a series of designations. Areas D and E in which the previous two digits are indicated are selected from the numbers (S614).

  When the position of the plate image area is present at the center of the front surface of the vehicle, the character area selection unit 52 selects the areas E and F where the center two digits are indicated among the series designation numbers (S616).

  When the position of the plate image area is present on the left side of the front part of the vehicle (on the right side in an actual vehicle), the character area selection unit 52 indicates the area B where the classification number is indicated and the last two digits of the series designation number. Regions F and G are selected (S618).

  The luminance difference calculation unit 36 calculates the luminance difference by calculating the average luminance of the character region and the background region for the region selected by the processing of S608 to S618 (S620).

  Next, the luminance difference rate calculation unit 38 calculates a luminance difference rate based on the luminance difference calculated by the luminance difference calculation unit 36 (S622), and the plate cover determination unit 40 determines whether the luminance difference rate is less than the threshold value. Is determined (S624).

  When the luminance difference rate is less than the threshold, the plate cover determination unit 40 determines that the plate cover is attached to the vehicle license plate included in the image data (with plate cover) (S626). In this case, the determination result storage unit 42 stores the determination result, the luminance difference rate (light transmittance), and other items in the recognized vehicle database 60 (S628). The determination result storage unit 42 may transmit the same data to a monitoring center or the like via a network when storing the data in the recognized vehicle database 60.

  On the other hand, when the plate cover determination unit 40 determines that the luminance difference rate is equal to or greater than the threshold (no plate cover), the standard average luminance calculation unit 44 calculates the standard average luminance and updates the standard average luminance database 62. (S630).

  Note that the vehicle image recognition device 5 of the fifth embodiment may perform a process of designating a region where the recognition of characters and numbers has failed as a target for calculating the average luminance and the luminance difference, as in the fourth embodiment. .

  Further, the vehicle image recognition device 5 of the fifth embodiment is similar to the second and third embodiments, in place of or in addition to the condition that “the imaging time zone is at night”. The condition that “the light is turned on” may be used.

[Summary]
The vehicle image recognition apparatus 5 according to the fifth embodiment described above can calculate the luminance difference rate that is an index value indicating the light transmittance of the plate cover, as in the first embodiment. It can be determined whether or not a plate cover is attached to the plate.

Further, in the vehicle image recognition device 5 of the fifth embodiment, since the character region selection unit 52 selects a specific region from the plate image region based on various conditions, the light of the headlight, the shadow produced by the sunlight, the license plate It is possible to select an area where the influence of dirt or the like is small.
For this reason, the vehicle image recognition apparatus 5 of 5th Example can suppress calculating a luminance difference smaller than an original value. As a result, the vehicle image recognition device 5 of the fifth embodiment can more accurately determine whether or not the plate cover is attached.

  The “first area” in the claims corresponds to, for example, the character area of each embodiment, and the “second area” corresponds to the background area of each embodiment, for example. Further, the “determination unit” in the claims corresponds to, for example, the luminance difference rate calculation unit 38 and the plate cover determination unit 40 of each embodiment, and the “index value calculation unit” includes, for example, the luminance difference calculation of each embodiment. This corresponds to the part 36. Further, the “vehicle size type determination unit” in the claims corresponds to, for example, the vehicle image region extraction unit 30 of each embodiment.

  The best mode for carrying out the present invention has been described above with reference to the embodiments. However, the present invention is not limited to these embodiments, and various modifications can be made without departing from the scope of the present invention. And substitutions can be added.

  For example, in each embodiment, the region is selected depending on whether the position of the plate image region exists in the center or the end of the front portion of the vehicle. You may change a division | segmentation aspect according to the quantity away from the center part. For example, the vehicle image recognition device may increase the number of divisions as the amount of the position of the plate image region that is separated from the center increases. A large amount of the plate image area away from the center means that the position of the license plate is close to the headlight. In this case, it is expected that the area where the luminance difference rate cannot be accurately calculated will increase. Because it is done.

Regarding the above description, the following items are further disclosed.
(Appendix 1)
A plate image area extraction unit for extracting a plate image area corresponding to a license plate from an input image including a vehicle;
A luminance difference for calculating a luminance difference between an average luminance of the first area corresponding to the license plate description information of the plate image area and an average luminance of the second area corresponding to the background of the license plate in the plate image area. A calculation unit;
A determination unit for determining whether a plate cover is attached to the license plate of the vehicle based on the calculated luminance difference and a set standard luminance difference;
A vehicle image recognition device.
(Appendix 2)
The vehicle image recognition device according to attachment 1, wherein
An area dividing unit for dividing the plate image area into a plurality of areas;
The luminance difference calculation unit calculates the luminance difference for each of the divided areas, and outputs the largest luminance difference among the calculated plurality of luminance differences to the determination unit.
The vehicle image recognition device according to appendix 1.
(Appendix 3)
The vehicle image recognition device according to attachment 2, wherein
The area dividing unit divides the plate image area by a boundary set in a horizontal direction or a vertical direction according to an imaging time zone.
(Appendix 4)
The vehicle image recognition device according to attachment 3, wherein
From the input image, a headlight lighting detection unit that detects whether or not the vehicle headlights are lit,
When lighting of the headlight is detected by the headlight lighting detection unit, the region dividing unit divides the plate image region by a boundary set in a vertical direction,
Vehicle image recognition device.
(Appendix 5)
The vehicle image recognition device according to attachment 3, wherein
A vehicle image region extraction unit that extracts a vehicle image region corresponding to the vehicle from the input image;
When the area dividing unit divides the plate image area with a boundary set in the vertical direction,
A vehicle image recognition apparatus, wherein a division method of the plate image area is changed according to a positional relationship between the vehicle image area and the plate image area.
(Appendix 6)
The vehicle image recognition device according to attachment 3, wherein
A vehicle size type determining unit for determining the size type of the vehicle included in the input image;
When the size classification is a large vehicle, the area dividing unit divides the plate image area by a boundary line set in a horizontal direction.
(Appendix 7)
The vehicle image recognition device according to attachment 2, wherein
The region dividing unit divides the plate image region into a plurality of regions each including one character or number.
(Appendix 8)
The vehicle image recognition device according to attachment 1, wherein
An area selection unit that divides the plate image area into a plurality of areas and selects a part of the plurality of areas,
The brightness difference calculating unit calculates the brightness difference for the selected region.
(Appendix 9)
The vehicle image recognition device according to appendix 8, wherein
The region selection unit divides the plate image region with a boundary set in a horizontal direction or a vertical direction in accordance with an imaging time zone.
(Appendix 10)
The vehicle image recognition device according to appendix 9, wherein
A vehicle image region extraction unit that extracts a vehicle image region corresponding to the vehicle from the input image;
When lighting of a headlight is detected by the headlight lighting detection unit, the region selection unit selects a part from a plurality of regions divided by a boundary line set in the vertical direction. A vehicle image recognition device.
(Appendix 11)
The vehicle image recognition device according to appendix 9, wherein
A vehicle image region extraction unit that extracts a vehicle image region corresponding to the vehicle from the input image;
When the area selection unit divides the plate image area with a boundary set in the vertical direction,
A vehicle image recognition apparatus, wherein a division method of the plate image area is changed according to a positional relationship between the vehicle image area and the plate image area.
(Appendix 12)
The vehicle image recognition device according to appendix 9, wherein
A vehicle size type determining unit for determining the size type of the vehicle included in the input image;
When the size type is a large vehicle, the area selection unit selects the plate image area other than the lowermost area among a plurality of areas divided by a boundary set in the horizontal direction. A vehicle image recognition device.
(Appendix 13)
The vehicle image recognition device according to appendix 8, wherein
The vehicle image recognition apparatus, wherein the region selection unit divides the plate image region into a plurality of regions each including one letter or number, and selects a part of the plurality of regions.
(Appendix 14)
The vehicle image recognition device according to any one of appendices 1 to 13,
A vehicle image provided with an index value calculation unit that calculates an index value indicating a light transmittance of a plate cover attached to the license plate by obtaining a ratio between the calculated brightness difference and a set standard brightness difference. Recognition device.
(Appendix 15)
The vehicle image recognition device according to attachment 3, wherein
The area dividing unit divides the plate image area by a boundary set in a horizontal direction when the imaging time zone is daytime.
(Appendix 16)
The vehicle image recognition device according to attachment 3, wherein
The area dividing unit divides the plate image area by a boundary set in a vertical direction when the imaging time zone is nighttime.
(Appendix 17)
The vehicle image recognition device according to appendix 5,
When the area dividing unit divides the plate image area with a boundary set in the vertical direction,
When the plate image area of the vehicle is present in the center portion in the horizontal direction of the vehicle front part included in the image, the plate image area is divided into three by setting two boundary lines,
Vehicle image recognition characterized in that, if the plate image area does not exist in the central portion of the front surface of the vehicle included in the image in the horizontal direction, the plate image area is divided into two by setting one boundary line. apparatus.
(Appendix 18)
The vehicle image recognition device according to attachment 3, wherein
The region dividing unit divides the plate image region into a plurality of regions each including one letter or number.
(Appendix 19)
The vehicle image recognition device according to appendix 8, wherein
The region selection unit selects a lower region among a plurality of regions obtained by dividing the plate image region by a boundary line set in a horizontal direction when the imaging time zone is daytime. Recognition device.
(Appendix 20)
The vehicle image recognition device according to appendix 8, wherein
The area selection unit selects a part from a plurality of areas obtained by dividing the plate image area by a boundary set in the vertical direction when the imaging time zone is nighttime.
(Appendix 21)
The vehicle image recognition device according to appendix 11,
When the area selection unit divides the plate image area of the vehicle with a boundary set in the vertical direction,
When the plate image area is present in the central portion of the vehicle front portion included in the image in the horizontal direction, the two boundary lines are set, the plate image area is divided into three, and the center area is selected.
When the plate image area does not exist in the central portion in the horizontal direction of the front part of the vehicle included in the image, an area on the side farther from the headlight by setting one boundary line and dividing the plate image area into two A vehicle image recognition apparatus, wherein:
(Appendix 22)
The vehicle image recognition device according to appendix 8, wherein
The area selection unit divides the plate image area into a plurality of areas each including one letter or number, and selects a part of the plurality of areas.
(Appendix 23)
The vehicle image recognition device according to any one of appendices 1 to 22,
A camera for imaging a vehicle license plate and transmitting an image to the vehicle image recognition device;
A vehicle image recognition system.
(Appendix 24)
Computer
Extract the plate image area corresponding to the license plate from the input image including the vehicle,
Calculating the luminance difference between the average luminance of the first area corresponding to the license plate description information of the plate image area and the average luminance of the second area corresponding to the background of the license plate in the plate image area;
Based on the calculated brightness difference and the set standard brightness difference, it is determined whether a plate cover is attached to the license plate of the vehicle.
Vehicle image recognition method.
(Appendix 25)
On the computer,
Extract the plate image area corresponding to the license plate from the input image including the vehicle,
Calculating the luminance difference between the average luminance of the first area corresponding to the license plate description information of the plate image area and the average luminance of the second area corresponding to the background of the license plate in the plate image area;
Based on the calculated brightness difference and the set standard brightness difference, it is determined whether or not a plate cover is attached to the license plate of the vehicle.
Vehicle image recognition program

1, 2, 3, 4, 5 Vehicle image recognition device 30 Vehicle image region extraction unit 32 Plate image region extraction unit 34 Number recognition unit 36 Luminance difference calculation unit 38 Luminance difference rate calculation unit 40 Plate cover determination unit 42 Determination result storage unit 44 Standard average luminance calculation unit 45 Headlight lighting detection unit 46 Region division unit 48 Region selection unit 50 Character region division unit 52 Character region selection unit 60 Recognition vehicle database 62 Standard average luminance database 100 Camera 110 Infrared light irradiation lamp

Claims (13)

A plate image area extraction unit for extracting a plate image area corresponding to a license plate from an input image including a vehicle;
A luminance difference for calculating a luminance difference between an average luminance of the first area corresponding to the license plate description information of the plate image area and an average luminance of the second area corresponding to the background of the license plate in the plate image area. A calculation unit;
A determination unit for determining whether a plate cover is attached to the license plate of the vehicle based on the calculated luminance difference and a set standard luminance difference;
An area dividing unit for dividing the plate image area into a plurality of areas;
The luminance difference calculation unit calculates the luminance difference for each of the divided areas, and outputs the largest luminance difference among the calculated plurality of luminance differences to the determination unit.
Vehicle image recognition device. The vehicle image recognition device according to claim 1 ,
The area dividing unit divides the plate image area by a boundary set in a horizontal direction or a vertical direction according to an imaging time zone. The vehicle image recognition device according to claim 2 ,
From the input image, a headlight lighting detection unit that detects whether or not the vehicle headlights are lit,
When lighting of the headlight is detected by the headlight lighting detection unit, the region dividing unit divides the plate image region by a boundary set in a vertical direction,
Vehicle image recognition device. The vehicle image recognition device according to claim 2 ,
A vehicle image region extraction unit that extracts a vehicle image region corresponding to the vehicle from the input image;
When the area dividing unit divides the plate image area with a boundary set in the vertical direction,
A vehicle image recognition apparatus, wherein a division method of the plate image area is changed according to a positional relationship between the vehicle image area and the plate image area. The vehicle image recognition device according to claim 2 ,
A vehicle size type determining unit for determining the size type of the vehicle included in the input image;
When the size classification is a large vehicle, the area dividing unit divides the plate image area by a boundary line set in a horizontal direction. The vehicle image recognition device according to claim 1 ,
The region dividing unit divides the plate image region into a plurality of regions each including one character or number. A plate image area extraction unit for extracting a plate image area corresponding to a license plate from an input image including a vehicle;
A luminance difference for calculating a luminance difference between an average luminance of the first area corresponding to the license plate description information of the plate image area and an average luminance of the second area corresponding to the background of the license plate in the plate image area. A calculation unit;
A determination unit for determining whether a plate cover is attached to the license plate of the vehicle based on the calculated luminance difference and a set standard luminance difference;
An area selection unit that divides the plate image area into a plurality of areas and selects a part of the plurality of areas,
The brightness difference calculating unit calculates the brightness difference for the selected region. The vehicle image recognition device according to claim 7 ,
The area selection unit divides the plate image area by a boundary set in a horizontal direction or a vertical direction according to an imaging time zone, and selects a part of the plurality of areas. Image recognition device. The vehicle image recognition device according to claim 8 ,
From the input image, a headlight lighting detection unit that detects whether or not the vehicle headlights are lit,
When lighting of a headlight is detected by the headlight lighting detection unit, the region selection unit selects a part from a plurality of regions divided by a boundary line set in the vertical direction. A vehicle image recognition device. The vehicle image recognition device according to claim 8 ,
A vehicle image region extraction unit that extracts a vehicle image region corresponding to the vehicle from the input image;
When the area selection unit divides the plate image area with a boundary set in the vertical direction,
A vehicle image recognition apparatus, wherein a division method of the plate image area is changed according to a positional relationship between the vehicle image area and the plate image area. The vehicle image recognition device according to claim 8 ,
A vehicle size type determining unit for determining the size type of the vehicle included in the input image;
When the size type is a large vehicle, the area selection unit selects the plate image area other than the lowermost area among a plurality of areas divided by a boundary set in the horizontal direction. A vehicle image recognition device. The vehicle image recognition device according to claim 7 ,
The vehicle image recognition apparatus, wherein the region selection unit divides the plate image region into a plurality of regions each including one letter or number, and selects a part of the plurality of regions. Computer
Extract the plate image area corresponding to the license plate from the input image including the vehicle,
Calculating the luminance difference between the average luminance of the first area corresponding to the license plate description information of the plate image area and the average luminance of the second area corresponding to the background of the license plate in the plate image area;
Based on the calculated brightness difference and the set standard brightness difference, it is determined whether or not a plate cover is attached to the license plate of the vehicle ,
Dividing the plate image area into a plurality of areas;
The process of calculating the luminance difference calculates the luminance difference for each of the divided areas, and supplies the largest luminance difference among the calculated plurality of luminance differences to the determination process.
Vehicle image recognition method.