JP2009134591A - Vehicle color determination device, vehicle color determination system, and vehicle color determination method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a vehicle color determination device, a vehicle color determination system and a vehicle color determination method, capable of determining precisely a vehicle color at low cost. <P>SOLUTION: A plate recognition part 14 recognizes a number plate. A plate position calculating part 15 calculates moving vectors xm, ym of a plate position, based on coordinates xp, yp of the plate position on each of plate pick-up images at different imaging times. A vehicle color determination area specifying part 17 calculates coordinates xs, ys of the plate position on a color determining pick-up image in the imaging time t+n after a prescribed time lapses from the imaging time t+1, based on the coordinates xp, yp of the plate position, the moving distance (vector) (for example ym) of the plate position, and the like, and specifies a vehicle color determination area, based on the calculated coordinates xs, ys of the plate position. A vehicle color determination part 18 determines the vehicle color, based on a pixel value in the vehicle color determination area. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a vehicle color determination device that determines a color of a vehicle based on a captured image obtained by imaging a vehicle traveling on a road, a vehicle color determination system including the vehicle color determination device, and a vehicle color determination method.

  A technology has been developed to determine the travel time required to travel between predetermined points by identifying the number plates (vehicle numbers) of vehicles traveling on the road at multiple points and collating the identified number plates. ing.

On the other hand, there has been a problem that some characters cannot be recognized due to bending or dirt of the license plate. Therefore, color image data from a camera installed on the road is acquired, a discrimination area for discriminating the color of the vehicle such as a vehicle bonnet is extracted from the obtained color image data, and the color of the extracted discrimination area is identified Discloses a vehicle color discrimination device that discriminates the color of a vehicle (see Patent Document 1).
JP 2000-222673 A

  However, the technique of Patent Document 1 identifies a color of a vehicle different from the character information on the assumption that some of the character information of the license plate cannot be recognized. Both the character information of the license plate and the color of the vehicle Is required to have a high-resolution camera with a wide field of view and a color image. However, such a camera is expensive, and it is not realistic to install many such cameras at a plurality of points on the road. Further, when a discrimination area such as a hood of a vehicle is extracted from image data, there is a possibility that a front window may be erroneously extracted, and a method for accurately identifying the color of a vehicle has been desired.

  The present invention has been made in view of such circumstances, and is a vehicle color determination device that can accurately determine the color of a vehicle at low cost, a vehicle color determination system including the vehicle color determination device, and a vehicle color determination. It aims to provide a method.

  A vehicle color determination device according to a first aspect of the present invention is a vehicle based on a monochrome first captured image and a color second captured image obtained by imaging each of two imaging units that capture a vehicle traveling on a road. A vehicle color determination device for determining a color of a vehicle, wherein the feature part detection unit detects a feature part of the vehicle based on a first captured image, and the first imaging of the feature part detected by the feature part detection unit A determination area specifying means for specifying a determination area for determining the color of the vehicle based on a position on the image on the second captured image, and a color component of each pixel of the determination area specified by the determination area specifying means Vehicle color determination means for determining the color of the vehicle on the basis thereof.

  According to a second aspect of the present invention, in the first aspect of the invention, the vehicle color determination apparatus includes an acquisition unit that acquires a second captured image at a capturing time point after a predetermined time has elapsed from the capturing time point of the first captured image. Is configured to specify a determination region on the second captured image acquired by the acquisition unit.

  The vehicle color determination device according to a third aspect of the present invention is the first aspect or the second aspect, wherein the characteristic part detection means is configured to detect a characteristic part in each of the first captured images at different imaging points, A movement distance calculation unit that calculates a movement distance of the feature part detected by the feature part detection unit on the first captured image, and a second of the feature part according to the movement distance calculated by the movement distance calculation unit Vertical position calculation means for calculating a vertical position on the captured image, wherein the determination area specifying means is configured to specify the determination area based on the vertical position calculated by the vertical position calculation means. Features.

  According to a fourth aspect of the present invention, in the vehicle color determination apparatus according to the second aspect, the feature part detection means is configured to detect a feature part in each of the first captured images at different imaging points, and the feature part detection A moving distance calculating means for calculating a moving distance of the characteristic part detected by the means on the first captured image, a moving distance calculated by the moving distance calculating means, an imaging time point of the second captured image, and the first captured image. Vertical position calculating means for calculating the vertical position of the feature portion on the second captured image according to the imaging time difference from the imaging time point of the imaging, and the determination area specifying means is calculated by the vertical position calculating means The determination area is specified based on the vertical position.

  According to a fifth aspect of the present invention, there is provided the vehicle color determination apparatus according to any one of the first to fourth aspects of the invention, which calculates a horizontal position on the first captured image of the characteristic part detected by the characteristic part detection means. A horizontal position calculating means; and a second horizontal position calculating means for calculating a horizontal position on the second captured image of the characteristic part according to the horizontal position calculated by the first horizontal position calculating means, The area specifying means is configured to specify the determination area based on the horizontal position calculated by the second horizontal position calculating means.

  A vehicle color determination apparatus according to a sixth aspect of the present invention is a vehicle based on a monochrome first captured image and a color second captured image obtained by capturing each of two imaging units that capture a vehicle traveling on a road. A vehicle color determination device for determining the color of a vehicle, characterized by detecting a characteristic part of the vehicle based on each of the first captured images at different imaging points, and of the characteristic part detected by the characteristic part detection unit The movement distance calculation means for calculating the movement distance on the first captured image, and all or a part of the specific part for color determination of the vehicle is the second according to the movement distance calculated by the movement distance calculation means. Determining means for determining an imaging time point of the second captured image so as to be included in a predetermined determination area on the captured image; and each of the determination areas on the second captured image at the imaging time point determined by the determining means The vehicle color is determined based on the color component of the pixel. Characterized in that it comprises a body color determining means for.

  According to a seventh aspect of the present invention, there is provided a vehicle color determining apparatus according to the sixth aspect, wherein a horizontal position calculating means for calculating a horizontal position on the first captured image of the characteristic part detected by the characteristic part detecting means, and the horizontal position calculating means And a horizontal position determining means for determining a horizontal position on the second captured image of the determination area according to the horizontal position calculated in (5).

  According to an eighth aspect of the present invention, in any one of the first to seventh aspects, the characteristic part is a license plate.

  According to a ninth aspect of the present invention, in any one of the first to eighth aspects, the resolution of the second captured image is smaller than the resolution of the first captured image. .

  A vehicle color determination system according to a tenth aspect of the present invention includes a vehicle color determination apparatus according to any one of the one or more of the above-described inventions, and a reception apparatus that receives predetermined information from the vehicle color determination apparatus. In the determination system, the vehicle color determination device includes a transmission unit that transmits information on a characteristic part of the vehicle and information on the determined vehicle color to the reception device.

  According to an eleventh aspect of the present invention, there is provided a vehicle color determination method based on a monochrome first captured image and a color second captured image obtained by imaging each of two imaging units that capture a vehicle traveling on a road. A vehicle color determination method for determining a color of a vehicle, wherein a characteristic part of a vehicle is detected based on a first captured image, and the color of the vehicle is determined based on a position of the detected characteristic part on the first captured image. A determination area for determining the vehicle is identified on the second captured image, and the color of the vehicle is determined based on the color component of each pixel of the identified determination area.

  A vehicle color determination method according to a twelfth aspect of the invention is based on a monochrome first captured image and a color second captured image obtained by capturing each of two image capturing units that capture a vehicle traveling on a road. A vehicle color determination method for determining a color of a vehicle, wherein a feature part of a vehicle is detected based on each of first captured images at different imaging points, and a movement distance of the detected feature part on the first captured image is determined. The second captured image is calculated so that all or part of the specific part for determining the color of the vehicle is included in a predetermined determination region on the second captured image according to the calculated moving distance. The imaging time point is determined, and the color of the vehicle is determined based on the color component of each pixel in the determination area on the second captured image at the determined imaging time point.

  In the first invention and the eleventh invention, the vehicle color determination device determines the characteristic part of the vehicle based on the first monochrome captured image obtained by imaging with one imaging unit (first imaging unit). To detect. The characteristic part of the vehicle is, for example, a license plate. In addition, it may replace with a number plate and the vehicle head part, a vehicle tail part, the vehicle front or back corner | angular part etc. may be sufficient. The characteristic part of the vehicle is detected, for example, based on the difference between the image when the vehicle is not present and the image when the vehicle is present, and based on the detected vehicle image, It can be detected by performing a cut-out process. The vehicle color determination device detects the vehicle on the second captured image of the color obtained by imaging with the other imaging unit (second imaging unit) based on the position of the detected characteristic part on the first captured image. A determination region (for example, a vehicle hood) for determining the color of the vehicle is specified. Note that the position of the characteristic part includes not only the position at a certain imaging time point but also the displacement of the position between different imaging time points. The determination area is specified by, for example, specifying a corresponding position corresponding to the position of the characteristic part on the first captured image on the second captured image, and using a predetermined size based on the corresponding position on the second captured image. The determination area can be specified.

  The vehicle color determination device determines the color of the vehicle based on the color component of each pixel in the specified determination region. For example, the color component of each pixel in the determination region is converted from RGB to HSI (H: hue, S: saturation, I: luminance), and the color of the vehicle is determined based on the hue H and / or luminance I. it can. By specifying the color determination region with reference to the characteristic part of the vehicle, the color of the vehicle can be determined with higher accuracy than in the conventional method. Further, by using a monochrome first captured image for detecting a characteristic part of the vehicle and a second captured image of a color image for determining the color of the vehicle, the characteristic part of the vehicle can be obtained as in the past. It is not necessary to use a color image and high-resolution imaging unit for detection. For example, a high-resolution, narrow-field and monochrome image is used for detection of a vehicle characteristic part, and a low-resolution is used for vehicle color determination. By using a color image with a wide field of view, the cost of the entire apparatus can be reduced.

  In the second invention, the vehicle color determination device acquires the second captured image at the imaging time point after a predetermined time has elapsed from the imaging time point of the first captured image, and determines the determination region on the acquired second captured image. Is identified. In order to secure a determination area for determining the color of the vehicle in the second captured image at the same time as the first captured image at which the vehicle characteristic part is detected, the vehicle characteristic part and color Because it is different from the determination area for determining the image, the installation conditions of the second imaging unit (for example, the height from the road surface of the second imaging unit, the angle of the optical axis of the lens, the distance to the imaging point, etc.) are adjusted. There is a need to. However, it is difficult to individually adjust the installation conditions of the second imaging unit at different installation positions. By acquiring a second captured image at a capturing time point after a predetermined time has elapsed from the capturing time point of the first captured image in which the characteristic part of the vehicle is detected, the characteristic part of the vehicle moves to the vicinity of the center on the second captured image. In addition, a determination region for determining the color of the vehicle can be secured on the second captured image without adjusting the installation condition of the second imaging unit, and adjustment at the installation location can be made unnecessary. .

  In the third aspect of the invention, the vehicle color determination device calculates the movement distance of the detected characteristic part on the first captured image. For example, the movement distance is calculated based on the difference between the position of the characteristic part on the first captured image at the imaging time t and the position of the characteristic part on the first captured image at the imaging time t + 1. The vehicle color determination device calculates the vertical position on the second captured image of the characteristic part according to the calculated movement distance. In this case, the relationship between the moving distance of the characteristic part and the vertical position of the characteristic part on the second captured image can be set in advance. Accordingly, the position of the feature portion on the second captured image can be accurately obtained in correspondence with the moving distance of the feature portion during the predetermined time, that is, the speed of the vehicle. It is possible to specify the determination area for determining the color of the object with high accuracy.

  In the fourth aspect of the invention, the vehicle color determination device calculates a moving distance of the detected characteristic part on the first captured image. For example, the movement distance is calculated based on the difference between the position of the characteristic part on the first captured image at the imaging time t and the position of the characteristic part on the first captured image at the imaging time t + 1. The vehicle color determination device calculates the vertical position of the characteristic part on the second captured image according to the calculated moving distance and the imaging time difference between the imaging time of the second captured image and the imaging time of the first captured image. To do. In this case, the relationship between the movement distance and imaging time difference of the characteristic part and the vertical position of the characteristic part on the second captured image can be set in advance. Thereby, the position of the characteristic part on the second captured image can be obtained with higher accuracy in correspondence with the moving distance of the characteristic part during the predetermined time, that is, the speed of the vehicle and the difference in imaging time. Regardless of this, the determination region for determining the color of the vehicle can be specified with higher accuracy.

  In the fifth invention, the vehicle color determination device calculates a horizontal position on the first captured image of the detected characteristic part. The vehicle color determination device calculates the horizontal position of the characteristic part on the second captured image in accordance with the calculated horizontal position, and specifies the determination region based on the calculated horizontal position. In this case, the relationship between the horizontal position of the characteristic part on the first captured image and the horizontal position of the characteristic part on the second captured image can be set in advance. As a result, the horizontal position of the characteristic part on the second captured image can be obtained, and the determination region for determining the color of the vehicle can be specified with high accuracy.

  In the sixth invention and the twelfth invention, the vehicle color determination device is obtained by imaging with one imaging unit (first imaging unit), and is based on each monochrome first captured image with different imaging time points. Detect a characteristic part of the vehicle. The characteristic part of the vehicle is, for example, a license plate. In addition, it may replace with a number plate and the vehicle head part, a vehicle tail part, the vehicle front or back corner | angular part etc. may be sufficient. The characteristic part of the vehicle is detected, for example, based on the difference between the image when the vehicle is not present and the image when the vehicle is present, and based on the detected vehicle image, It can be detected by performing a cut-out process. The vehicle color determination device calculates a movement distance of the detected feature portion on the first captured image. For example, the movement distance is calculated based on the difference between the position of the characteristic part on the first captured image at the imaging time t and the position of the characteristic part on the first captured image at the imaging time t + 1. The vehicle color determination device has a predetermined determination region (second determination image) on a second captured image in which all or a part of a specific part (for example, a hood of the vehicle) for determining the color of the vehicle is colored according to the calculated moving distance. For example, the imaging time point of the second captured image is determined so as to be included in the area near the center. In this case, the relationship between the moving distance of the characteristic part and the imaging time of the second captured image (elapsed time from the imaging time of the first captured image) can be set in advance.

  The vehicle color determination device determines the color of the vehicle based on the color component of each pixel in the determination area on the second captured image at the determined imaging time point. For example, the color component of each pixel in the determination region is converted from RGB to HSI (H: hue, S: saturation, I: luminance), and the color of the vehicle is determined based on the hue H and / or luminance I. it can. Since the specific part for determining the color of the vehicle according to the moving distance of the characteristic part of the vehicle, that is, the speed of the vehicle, overlaps the determination region of the second captured image, the vehicle is more accurately compared with the conventional method. The color can be determined. Further, by using a monochrome first captured image for detecting a vehicle characteristic part and a color second captured image for determining the color of the vehicle, the characteristic part of the vehicle is detected as in the past. Therefore, it is not necessary to use a color image and a high-resolution imaging unit.For example, a high-resolution image is used for detection of a vehicle feature, a monochrome image is used with a narrow field of view, and a low-resolution image is used for vehicle color determination. By using a color image with a wide field of view, the cost of the entire apparatus can be reduced.

  In the seventh invention, the vehicle color determination device calculates a horizontal position on the first captured image of the detected characteristic part, and on the second captured image in the determination area according to the calculated horizontal position. Determine the horizontal position. In this case, the relationship between the horizontal position of the characteristic part on the first captured image and the horizontal position of the determination area on the second captured image can be set in advance. Thereby, the horizontal position of the determination area on the second captured image can be determined according to the horizontal position of the characteristic part, and the color of the vehicle can be determined with high accuracy.

  In the eighth invention, the characteristic part is a license plate. As a result, the license plate of the vehicle can be associated with the color of the vehicle, and even if some of the characters on the license plate cannot be recognized, the vehicle identification can be supplemented by the color of the vehicle. The vehicle identification accuracy can be improved.

  In the ninth invention, the resolution of the second captured image is smaller than the resolution of the first captured image. As a result, the first captured image has a high resolution, a narrow field of view and a monochrome image (for example, a black and white image), and the second captured image has a low resolution, a wide field of view, and a color image. In addition, it is not necessary to have an imaging unit with a wide field of view and capable of acquiring a color image, and the cost of the entire apparatus can be reduced.

  In the tenth aspect of the invention, the vehicle color determination device receives information related to the characteristic part of the vehicle (for example, character information of the license plate) and information related to the determined vehicle color (for example, a central device having a reception function, monitoring) Device). As a result, the license plate of the vehicle can be associated with the color of the vehicle, and even if some of the characters on the license plate cannot be recognized, the vehicle identification can be supplemented by the color of the vehicle. The vehicle identification accuracy can be improved.

  In the present invention, the color of the vehicle can be determined with higher accuracy than in the conventional method. In addition, the cost of the entire apparatus can be reduced.

Embodiment 1
Hereinafter, the present invention will be described with reference to the drawings illustrating embodiments. FIG. 1 is a block diagram showing an example of the configuration of a vehicle color determination system according to the present invention. The vehicle color determination system according to the present invention includes a vehicle color determination device 100, a central device 300, and the like. An imaging device 200 is connected to the vehicle color determination device 100. The imaging device 200 may be configured to be built in the vehicle color determination device 100.

  The imaging apparatus 200 includes an illumination unit 210, a plate recognition camera 220, a vehicle body color determination camera 230, and the like. The illumination unit 210 includes a plurality of LEDs, and irradiates the vehicle traveling on the road with near-infrared light at the timing (for example, every 33 ms) when the plate recognition camera 220 images the vehicle. Thereby, for example, even at night, the plate recognition camera 220 can capture the vehicle and recognize the license plate (vehicle number).

  The plate recognition camera 220 images a vehicle traveling on the road, and outputs image data obtained by the imaging to the vehicle color determination device 100. The plate captured image (first captured image) obtained by capturing with the plate recognition camera 220 has high resolution (for example, horizontal x vertical is 1400 x 1050, 1600 x 1200, etc.), and the field of view (view angle). Is narrow (for example, 10 ° to 15 °) and is a black and white image. Here, the black-and-white image is a grayscale image composed of white, black, gray in between, and the like.

  The vehicle color determination camera 230 images a vehicle traveling on the road and outputs image data obtained by the imaging to the vehicle color determination device 100. A color determination captured image (second captured image) obtained by capturing with the vehicle body color determination camera 230 has a low resolution (for example, horizontal × vertical is 720 × 480, 800 × 600, etc.), field of view (view angle). Is a wide (eg, 20 ° -50 °) color image. This eliminates the need for a camera (imaging device) that has a high resolution, a wide field of view, and a color image, and can reduce the cost of the entire device.

  The plate recognition camera 220 and the vehicle body color determination camera 230 are built in the imaging apparatus 200, that is, incorporated in the same housing, and the mounting position of each camera, the optical axis of the lens, and the like are determined when the imaging apparatus 200 is manufactured. Fixed in place. Thus, when the imaging device 200 is installed on the side of the road, it is not necessary to adjust the attachment positions of the plate recognition camera 220 and the vehicle body color determination camera 230.

  FIG. 2 is an explanatory diagram illustrating an example of a plate captured image captured by the plate recognition camera 220. FIG. 3 is an explanatory diagram illustrating an example of a color determination captured image captured by the vehicle body color determination camera 230. As shown in FIG. 2, the plate captured image has a high resolution, and the horizontal axis is represented by xp and the vertical axis is represented by yp with the upper left of the captured image being the center of coordinates. As shown in FIG. 3, the color determination captured image has a resolution lower than the resolution of the plate captured image, and the horizontal axis is represented by xs and the vertical axis is represented by ys with the upper left of the captured image being the center of coordinates.

  By fixing the plate recognition camera 220 and the vehicle body color determination camera 230 at predetermined positions in the image pickup apparatus 200, as shown in FIG. In the case where the image is shown, at the imaging time t + Δt after a predetermined time has elapsed from the imaging time t, the entire vehicle is shown in the color determination captured image as shown in FIG.

  The vehicle color determination apparatus 100 includes a control unit 10 that controls each unit in the apparatus, a plate recognition image input unit 11, a vehicle body color determination image input unit 12, a vehicle detection unit 13, a plate recognition unit 14, and a plate position calculation unit 15. A character recognition unit 16, a vehicle body color determination region specifying unit 17, a vehicle body color determination unit 18, a storage unit 19, a day / night determination unit 20, a clock 21, a communication unit 22, and the like.

  The plate recognition image input unit 11 acquires image data obtained by imaging with the plate recognition camera 220, and stores the acquired image data in the storage unit 19 in association with the imaging time for each frame.

  The vehicle body color determination image input unit 12 acquires image data obtained by imaging with the vehicle body color determination camera 230, and stores the acquired image data in the storage unit 19 in association with the imaging time for each frame.

  The vehicle detection unit 13 takes out image data obtained by imaging with the plate recognition camera 220 from the storage unit 19 and detects a vehicle (including a part of the vehicle) based on the taken-out image data. As the vehicle detection process, for example, the vehicle can be detected based on the difference between the plate-captured image when the vehicle is not present and the plate-captured image when the vehicle is present. The vehicle detection unit 13 outputs the detected vehicle image to the plate recognition unit 14. Note that the vehicle image may be an image of only the vehicle portion obtained by extracting the vehicle portion from the plate-captured image, or may specify the range of the vehicle portion in the plate-captured image.

  Based on the detected vehicle image, the plate recognition unit 14 recognizes a license plate by performing a cutting process of a rectangular license plate part. The plate recognizing unit 14 outputs the recognized plate image of the license plate, the coordinates of the license plate on the captured image for plate, and the like to the plate position calculating unit 15 and the character recognizing unit 16.

  The plate position calculation unit 15 includes an image memory therein, and a predetermined position of the license plate unit (for example, a rectangular license plate unit) from the coordinates on the plate image of the plate image output from the plate recognition unit 14. (Hereinafter referred to as “plate position”), and the calculated coordinates (xp, yp) of the plate position and the imaging time of the captured image for the plate are stored in the image memory in association with each other. In addition, a plate position is not limited to the above-mentioned position, The other location of a license plate part may be sufficient.

  The plate position calculation unit 15 calculates a movement vector (xm, ym) of the plate position between times t and t + 1 based on the coordinates of the plate position at different imaging times (for example, time t and time t + 1).

  FIG. 4 is an explanatory view showing an example of the plate position and its moving distance. As shown in FIG. 4, the movement distance xm of the plate position in the xp-axis direction is xm = xp (t + 1) −xp (t), and the movement distance ym of the plate position in the yp-axis direction is ym = yp (t + 1). −yp (t). Here, xp (y) and yp (t) are the plate position coordinate xp and the coordinate yp at time t, respectively. On the plate captured image, the movement distance in the vertical direction (vertical direction) of the plate position can be obtained by ym, and the movement distance in the horizontal direction (horizontal direction) can be obtained by xm.

  The character recognition unit 16 recognizes the character of the number (vehicle number) based on the plate image output from the plate recognition unit 14. The character recognition process indicates, for example, the similarity between the binarized image and the reference image by a matched filter that receives a binarized image obtained by binarizing the plate image and a reference image that is a car number template. A correlation distribution is created, and the character indicated by the maximum value of the correlation distribution can be extracted as a recognized character for each character of the branch name, classification number, vehicle type symbol, and serial number of the license plate.

  The vehicle body color determination region specifying unit 17 includes, for example, the coordinates (xp, yp) of the plate position calculated by the plate-captured image at the imaging time t + 1, the movement distance (for example, ym) of the plate position, and the like. Based on the above, the coordinates (xs, ys) of the plate position on the color determination captured image at the imaging time t + n after the elapse of a predetermined time from the imaging time t + 1 are calculated.

  FIG. 5 is an explanatory diagram illustrating an example of a plate position on the color determination captured image, and FIGS. 6 and 7 are explanatory diagrams illustrating an example of a method for calculating the plate position on the color determination captured image. The example of FIG. 5 shows a state in which the plate position is specified on the color determination captured image at the imaging time t + n based on the plate position specified by the plate captured image at the imaging time t + 1.

  FIG. 6 shows the relationship between the moving distance ym of the plate position in the plate image and the coordinate ys of the plate position in the color determination image. That is, the coordinate ys of the plate position in the color determination captured image can be obtained according to the movement distance ym of the plate position in the plate captured image. The relationship shown in FIG. 6 can be obtained in advance. When calculating the coordinate ys of the plate position in the color determination captured image, the relationship shown in FIG. 6 can be obtained by calculation, or the correspondence between the movement distance ym and the coordinate ys is stored in the LUT or the like. In addition, the movement distance ym can be obtained by converting it into the coordinate ys.

  Thus, the plate position moving distance during the predetermined time (the time difference between the imaging time t and the imaging time t + 1), that is, the vertical position of the plate position on the color determination captured image is accurately matched to the vehicle speed. As a result, a vehicle body color determination area for determining the color of the vehicle, which will be described later, can be specified with high accuracy regardless of the speed of the vehicle.

  In the example of FIG. 6, the coordinate ys corresponding to the moving distance ym can be changed according to the imaging time difference (n−1) between the imaging time t + 1 of the plate imaging image and the imaging time t + n of the color determination imaging image. . For example, when the imaging time difference is small, the coordinate ys can be reduced, and the coordinate ys can be increased as the imaging time difference increases. Thereby, it is possible to accurately determine the vertical position of the plate position on the color-determined captured image corresponding to the moving distance of the plate position during a predetermined time, that is, the speed of the vehicle and the imaging time difference. A vehicle body color determination region for determining the color of the vehicle, which will be described later, can be specified with high accuracy regardless of the speed of the vehicle.

  FIG. 7 shows the relationship between the coordinate xp of the plate position in the captured image for plate and the coordinate xs of the plate position in the captured image for color determination. That is, the plate position coordinate xs in the color determination captured image can be obtained according to the plate position coordinate xp in the plate captured image. The relationship shown in FIG. 7 can be obtained in advance. When calculating the coordinate xs of the plate position in the color determination captured image, the relationship shown in FIG. 7 can be obtained by calculation, or the association between xp and xs is stored in an LUT or the like. It can also be obtained by converting the coordinate xp to the coordinate xs. Accordingly, the horizontal position of the plate position on the color determination captured image can be obtained, and the vehicle body color determination region can be specified with high accuracy.

  8 and 9 are explanatory diagrams showing another example of a method for calculating the plate position on the color determination captured image. FIG. 8 shows the relationship between the movement distance ym of the plate position in the plate image and the coordinate ys of the plate position in the color determination image. In FIG. 9, the plate position coordinate xp in the plate image. And the coordinate xs of the plate position in the color determination captured image are shown. The difference from the example of FIGS. 6 and 7 assumes, for example, a case where the lanes of the road to be imaged are different, and the value of the coordinate ys when the movement distance ym is 0, and the coordinate xs when the coordinate xp is 0. The value is different.

  The vehicle body color determination area specifying unit 17 specifies the vehicle body color determination area on the color determination captured image based on the coordinates (xs, ys) of the plate position on the calculated color determination captured image.

  FIG. 10 is an explanatory diagram showing an example of the vehicle body color determination area. As shown in FIG. 10, the vehicle body color determination area is an area that is a predetermined number of pixels (for example, 10 to 20 pixels) above the coordinates (xs, ys) of the plate position, and is configured by 100 pixels × 100 pixels. It can be a parallelogram area. In this case, the inclination of the sides of the parallelogram can be parallel to the traveling direction of the vehicle or the road direction. It should be noted that the shape, size, distance from the plate position, etc. of the vehicle body color determination area are merely examples, and are not limited to these, and are included in a relatively wide paint color portion such as a vehicle hood or trunk. Any area may be used.

  By using the color determination captured image at a capturing time after a predetermined time has elapsed from the image capturing time of the plate captured image, the plate position of the vehicle moves to the vicinity of the center on the color determining captured image. As a result, the vehicle color determination region can be secured on the color determination captured image, so that it is not necessary to adjust the camera installation conditions at the installation location.

  The vehicle body color determination unit 18 converts the pixel value of the specified vehicle body color determination region from, for example, RGB to HSI (H: hue, S: saturation, I: luminance), and according to the value of the saturation S. The color of the vehicle is determined based on the hue H and / or the luminance I. The type of color to be determined can be determined as appropriate. For example, according to the value of hue H, it can be classified into red, yellow, green, water, blue, purple, and the value of luminance I. Depending on the situation, it can be divided into white, gray, black, etc.

  More specifically, the vehicle body color determination unit 18 obtains the hue H and the luminance I for all the pixels in the vehicle body color determination region, determines the color for each pixel, and totals the number of pixels of each color. The vehicle body color determination unit 18 determines the color of the vehicle body based on the total number of pixels for each color. For example, the color with the largest number of pixels can be the color of the vehicle body. Or it is good also as the 1st candidate of the color of a vehicle body, and a 2nd candidate in order with the number of pixels more than a predetermined threshold. Or you may determine the color which mixed each color of the pixel number more than a predetermined threshold value as the color of a vehicle body.

  The vehicle body color determination unit 18 performs color determination regardless of chromatic or achromatic color during daytime, for example, according to the determination result from the day / night determination unit 20 described later, and only achromatic color at night. Judgment of color is performed.

  The day / night determination unit 20 determines whether the day / night is different (for example, whether the mode is the day mode) according to the output (time) from the clock 21, and outputs the determination result to the vehicle body color determination unit 18.

  Under the control of the control unit 10, the communication unit 22 associates the license plate (vehicle number) recognized by the character recognition unit 16 with the character information and the color of the vehicle determined by the vehicle body color determination unit 18, and sends it to the central device 300. Send. As a result, the license plate of the vehicle can be associated with the color of the vehicle, and even if some of the characters on the license plate cannot be recognized, the vehicle identification can be supplemented by the color of the vehicle. The vehicle identification accuracy can be improved.

  Next, the operation of the vehicle color determination device 100 will be described. FIG. 11 is a flowchart showing a processing procedure of the vehicle color determination apparatus 100. The control part 10 acquires the picked-up image for plates (S11), and detects a vehicle (S12). The control unit 10 extracts a rectangular license plate portion from the vehicle image, thereby recognizing the plate (S13) and recognizing characters from the recognized plate (S14).

  The control unit 10 calculates the plate position based on the extracted position of the license plate unit (S15), and calculates the moving distance of the plate position based on the plate position on the plate captured image at different imaging times (S15). S16).

  The control unit 10 acquires a color determination captured image at an imaging time after a predetermined time has elapsed from the imaging time of the plate captured image (S17), and obtains the plate position on the color determination captured image, thereby determining the body color. An area is specified (S18). The controller 10 determines the vehicle body color (S19). The details of the vehicle body color determination process will be described later.

  The control unit 10 associates the recognized plate information (character information of the car number) and the vehicle body color and transmits them to the external central device (S20), and determines whether or not there is a processing end instruction (S21). When there is no process end instruction (NO in S21), the control unit 10 continues the process from step S11, and when there is a process end instruction (YES in S21), the process ends.

  FIG. 12 is a flowchart showing a processing procedure for vehicle body color determination. The control unit 10 converts the pixel value (color component of each pixel) in the vehicle body color determination region from RGB to HSI (H: hue, S: saturation, I: luminance) (S191), and the saturation S is a predetermined value. It is determined whether it is larger than a threshold value (for example, 0.4) (S192). When the saturation S is greater than the predetermined threshold (YES in S192), the control unit 10 calculates the hue H as a chromatic color (S193), and when the saturation S is not greater than the predetermined threshold (NO in S192). ), The luminance I is calculated as an achromatic color (S194).

  The control unit 10 determines the color of the pixel based on the hue H or the luminance I (S195), and determines whether the color determination for all the pixels in the vehicle body color determination region is completed (S196). If the color determination for all the pixels has not been completed (NO in S196), the control unit 10 continues the process from step S192.

  When the color determination of all the pixels is completed (YES in S196), the control unit 10 adds up the determined number of pixels of each color (S197) and determines whether the day mode is set (S198). If it is the daytime mode (YES in S198), the control unit 10 determines the vehicle body color by chromatic or achromatic color (S199), and if it is not the daytime mode (NO in S198), the control unit 10 determines the vehicle body color by the achromatic color ( S200), the process is terminated.

  As described above, by specifying the vehicle body color determination region based on the position of the license plate (characteristic part) of the vehicle, the color of the vehicle can be determined with higher accuracy than in the conventional method. Further, by using a plate-captured image for detecting the plate position of the vehicle and a color-determined captured image for determining the color of the vehicle, a color image having a high resolution and a wide field of view is obtained. This eliminates the need for a camera capable of reducing the overall cost of the apparatus.

Embodiment 2
In the first embodiment described above, a predetermined position from the imaging time t + 1 is determined based on the coordinates (xp, yp) of the plate position calculated from the captured image for the plate at the imaging time t + 1, the movement distance (for example, ym) of the plate position, and the like. Although the coordinates (xs, ys) of the plate position on the color determination captured image at the imaging time t + n after the lapse of time have been calculated, the imaging time of the plate captured image and the imaging time of the color determination captured image The relationship with is not limited to this. For example, the imaging time of the color determination captured image can be determined based on the coordinates (xp, yp) of the plate position calculated from the captured image for plate, the movement distance (for example, ym) of the plate position, and the like.

  FIG. 13 is an explanatory diagram illustrating an example of a method for specifying a vehicle body color determination region according to the second embodiment. As illustrated in FIG. 13, the control unit 10 determines the color based on the coordinates (xp, yp) of the plate position calculated from the captured image for the plate at the imaging time t + 1, the movement distance (for example, ym) of the plate position, and the like. The imaging time t + m of the captured image is determined. That is, all or a part of a specific part for determining the color of the vehicle (for example, a vehicle hood, trunk, etc.) is included in a predetermined vehicle color determination region (for example, a region near the center) of the color determination captured image. Thus, the imaging time t + m is determined. In this case, the relationship between the movement distance of the plate position, the coordinates (xp, yp) of the plate position and the imaging time t + m of the color determination captured image (the elapsed time from the imaging time t + 1 of the captured image for plate) is set in advance. Can be kept.

  Further, the horizontal position on the color determination captured image of the vehicle body color determination region can be determined according to the plate position coordinate xp (horizontal position) on the plate captured image. In this case, the relationship between the coordinate xp (horizontal position) of the plate position on the plate image and the horizontal position on the color determination image in the vehicle body color determination area can be set in advance. Accordingly, the horizontal position of the vehicle body color determination region on the color determination captured image can be determined according to the horizontal position of the plate position, and the color of the vehicle can be determined with high accuracy.

  FIG. 14 is a flowchart illustrating a processing procedure of the vehicle color determination apparatus 100 according to the second embodiment. The control part 10 acquires the picked-up image for plates (S31), and detects a vehicle (S32). The controller 10 recognizes the plate by extracting a rectangular license plate portion from the vehicle image (S33), and recognizes characters from the recognized plate (S34).

  The control unit 10 calculates the plate position based on the extracted position of the license plate part (S35), and calculates the movement distance of the plate position based on the plate position on the plate captured image at different imaging times (S35). S36).

  The control unit 10 determines the imaging time of the color determination captured image based on the movement distance and coordinates of the plate position (S37), and acquires the color determination captured image at the determined imaging time (S38). The controller 10 determines the vehicle body color (S39). The vehicle body color determination process is the same as in the example of FIG.

  The control unit 10 associates the recognized plate information (character information of the car number) and the vehicle body color and transmits them to the external central device (S40), and determines whether or not there is a processing end instruction (S41). When there is no process end instruction (NO in S41), the control unit 10 continues the process after step S31, and when there is a process end instruction (YES in S41), the process ends.

  As described above, the color determination is performed so that the specific part for determining the color of the vehicle according to the moving distance of the plate position of the vehicle, that is, the speed of the vehicle overlaps the predetermined vehicle body color determination region of the color determination captured image. By determining the imaging time of the captured image for the vehicle, the color of the vehicle can be determined more accurately than in the conventional method. Further, by using a plate-captured image for detecting the plate position of the vehicle and a color-determined captured image for determining the color of the vehicle, a color image having a high resolution and a wide field of view is obtained. This eliminates the need for a camera capable of reducing the overall cost of the apparatus.

  In the above-described embodiment, the size is 100 pixels × 100 pixels and the parallelogram-shaped vehicle body color determination region is used. However, the size and shape of the vehicle body color determination region are fixed. It is not limited. For example, the size of the vehicle is detected from the plate captured image or the color determination captured image, and the size, shape, distance from the plate position, etc. of the vehicle body color determination region is changed according to the detected vehicle size. You can also. Further, the vehicle type or the like may be detected instead of the size of the vehicle.

  In the above-described embodiment, the vehicle is imaged from the front of the vehicle traveling on the road. However, the present invention is not limited to this, and may be configured to image from the rear of the vehicle. In this case, the vehicle body color determination region may be included in a paint color portion such as a trunk or a roof of the vehicle.

  In the above-described embodiment, the case where the camera is installed so as to read the license plate on the front side of the vehicle has been described, but the present invention also applies to the case where the camera is installed so as to read the license plate on the rear side of the vehicle. Can be applied. In this case, since the traveling direction of the vehicle is reversed, in the examples of FIGS. 6 to 9, the slope of the straight line is inverted from that of the front side (that is, the slope of the straight line is negative). Moreover, the straight lines illustrated in FIGS. 6 to 9 can move up, down, left, and right depending on the installation conditions of the camera.

  The disclosed embodiments are to be considered in all respects as illustrative and not restrictive. The scope of the present invention is defined by the terms of the claims, rather than the description above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

It is a block diagram which shows an example of a structure of the vehicle color determination system which concerns on this invention. It is explanatory drawing which shows an example of the picked-up image for plates imaged with the plate recognition camera. It is explanatory drawing which shows an example of the picked-up image for color determination imaged with the vehicle body color determination camera. It is explanatory drawing which shows an example of a plate position and its moving distance. It is explanatory drawing which shows an example of the plate position on the captured image for color determination. It is explanatory drawing which shows an example of the calculation method of the plate position on the picked-up image for color determination. It is explanatory drawing which shows an example of the calculation method of the plate position on the picked-up image for color determination. It is explanatory drawing which shows the other example of the calculation method of the plate position on the captured image for color determination. It is explanatory drawing which shows the other example of the calculation method of the plate position on the captured image for color determination. It is explanatory drawing which shows an example of a vehicle body color determination area | region. It is a flowchart which shows the process sequence of a vehicle color determination apparatus. It is a flowchart which shows the process sequence of vehicle body color determination. FIG. 10 is an explanatory diagram illustrating an example of a method for specifying a vehicle body color determination region according to a second embodiment. 6 is a flowchart illustrating a processing procedure of the vehicle color determination device according to the second embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 Car color determination apparatus 10 Control part 11 Image input part for plate recognition 12 Image input part for vehicle body color determination 13 Vehicle detection part 14 Plate recognition part 15 Plate position calculation part 16 Character recognition part 17 Car body color determination area specification part 18 Car body color Determination unit 19 Storage unit 20 Day / night determination unit 21 Clock 22 Communication unit 200 Imaging device 210 Illumination unit 220 Plate recognition camera 230 Car body color determination camera 300 Central device

Claims (12)

A vehicle color determination device that determines a color of a vehicle based on a monochrome first captured image and a color second captured image obtained by imaging each of two imaging units that capture a vehicle traveling on a road. And
Feature part detecting means for detecting a feature part of the vehicle based on the first captured image;
Determination area specifying means for specifying on the second captured image a determination area for determining the color of the vehicle based on the position on the first captured image of the characteristic part detected by the characteristic part detecting means;
A vehicle color determination device comprising: vehicle color determination means for determining a color of the vehicle based on a color component of each pixel in the determination region specified by the determination region specification means. An acquisition unit that acquires a second captured image at a capturing time after a predetermined time has elapsed from the capturing time of the first captured image;
The determination area specifying means includes
The vehicle color determination device according to claim 1, wherein the determination region is specified on the second captured image acquired by the acquisition unit. The characteristic part detecting means includes
It is configured to detect a feature part in each of the first captured images at different imaging time points,
A moving distance calculating means for calculating a moving distance on the first captured image of the characteristic part detected by the characteristic part detecting means;
Vertical position calculating means for calculating a vertical position of the feature portion on the second captured image according to the moving distance calculated by the moving distance calculating means;
The determination area specifying means includes
The vehicle color determination device according to claim 1 or 2, wherein a determination region is specified based on the vertical position calculated by the vertical position calculation means. The characteristic part detecting means includes
It is configured to detect a feature part in each of the first captured images at different imaging time points,
A moving distance calculating means for calculating a moving distance on the first captured image of the characteristic part detected by the characteristic part detecting means;
The vertical position of the characteristic part on the second captured image is determined according to the travel distance calculated by the travel distance calculating means and the imaging time difference between the imaging time of the second captured image and the imaging time of the first captured image. Vertical position calculating means for calculating,
The determination area specifying means includes
The vehicle color determination device according to claim 2, wherein the determination region is specified based on the vertical position calculated by the vertical position calculation means. First horizontal position calculation means for calculating a horizontal position on the first captured image of the characteristic part detected by the characteristic part detection means;
Second horizontal position calculating means for calculating a horizontal position on the second captured image of the characteristic part according to the horizontal position calculated by the first horizontal position calculating means,
The determination area specifying means includes
The vehicle color determination device according to any one of claims 1 to 4, wherein the determination region is specified based on the horizontal position calculated by the second horizontal position calculation means. . A vehicle color determination device that determines a color of a vehicle based on a monochrome first captured image and a color second captured image obtained by imaging each of two imaging units that capture a vehicle traveling on a road. And
Feature part detection means for detecting a feature part of the vehicle based on each of the first captured images at different imaging points;
A moving distance calculating means for calculating a moving distance on the first captured image of the characteristic part detected by the characteristic part detecting means;
In accordance with the movement distance calculated by the movement distance calculation means, the second imaging so that all or a part of the specific part for determining the color of the vehicle is included in a predetermined determination area on the second imaging image. Determining means for determining an image capturing time;
A vehicle color determination device comprising: vehicle color determination means for determining a color of the vehicle based on a color component of each pixel in the determination region on the second captured image at the time of image capture determined by the determination means. . Horizontal position calculating means for calculating a horizontal position on the first captured image of the characteristic part detected by the characteristic part detecting means;
The vehicle color according to claim 6, further comprising: a horizontal position determining unit that determines a horizontal position of the determination region on the second captured image according to the horizontal position calculated by the horizontal position calculating unit. Judgment device. The characteristic part is
The vehicle color determination device according to any one of claims 1 to 7, wherein the vehicle color determination device is a license plate.   The vehicle color determination device according to any one of claims 1 to 8, wherein the resolution of the second captured image is smaller than the resolution of the first captured image. A vehicle color determination system comprising: the vehicle color determination device according to any one of claims 1 to 9; and a reception device that receives predetermined information from the vehicle color determination device.
The vehicle color determination device
A vehicle color determination system comprising: a transmission unit configured to transmit information regarding a characteristic part of a vehicle and information regarding a determined vehicle color to the receiving device. A vehicle color determination method for determining a color of a vehicle based on a monochrome first captured image and a color second captured image obtained by capturing each of two imaging units that capture a vehicle traveling on a road. And
Detecting a characteristic part of the vehicle based on the first captured image;
A determination area for determining the color of the vehicle based on the position of the detected characteristic part on the first captured image is specified on the second captured image,
A vehicle color determination method characterized by determining a color of a vehicle based on a color component of each pixel of a specified determination region. A vehicle color determination method for determining a color of a vehicle based on a monochrome first captured image and a color second captured image obtained by capturing each of two imaging units that capture a vehicle traveling on a road. And
Detecting a characteristic part of the vehicle based on each of the first captured images at different imaging points;
Calculating the movement distance of the detected feature part on the first captured image;
According to the calculated moving distance, the imaging time point of the second captured image is determined so that all or a part of the specific part for determining the color of the vehicle is included in the predetermined determination region on the second captured image. Decide
A vehicle color determination method, comprising: determining a color of a vehicle based on a color component of each pixel in the determination region on the second captured image at the determined capturing time.

Images