JP2009070097A - Vehicle length measuring device and vehicle model determination device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a vehicle length measuring device capable of accurately calculating a vehicle length, and a vehicle model determination device using the same. <P>SOLUTION: The vehicle model determination device 10 comprises a vehicle detection means 21 for detecting a vehicle from an image; vehicle shadow extraction means 22 and 23 for separating and extracting a shadow area 52 located just below the vehicle of shadow areas 52 and 54 of the vehicle from the image; a vehicle length calculation means 25 for calculating the length of the vehicle based on the length of the shadow area 52 located just below the vehicle; and a vehicle model determination means 26 for determining the model of the vehicle from the length of the vehicle. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a vehicle length measurement device that processes an image obtained by photographing a vehicle and measures the length of the vehicle, and processes an image obtained by photographing the vehicle to determine a vehicle type of the vehicle. The present invention relates to a vehicle type determination device.

A conventional document (Patent Document 1) shows an apparatus for calculating the length of a vehicle. In the apparatus according to this conventional document, an image of a vehicle is taken with a camera, and the length of the vehicle is calculated using this image.
JP-A-9-35197

  In order to calculate the length of the vehicle using an image obtained by photographing the vehicle, a part of the front, side, back, etc. of the vehicle in the image or the shadow of the vehicle is used. However, when only a part of the front, side, back, etc. of the vehicle is captured in the image, the length of the vehicle that does not appear in the image cannot be calculated accurately.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to provide a vehicle length measurement device that calculates the length of a vehicle more accurately and a vehicle type determination device that uses the vehicle length measurement device.

  In order to achieve the above-described object, a vehicle length measurement device according to the present invention separates and extracts a vehicle detection unit that detects a vehicle from an image and a shadow region directly below the vehicle from the image. Vehicle shadow extracting means, and vehicle length calculating means for calculating the length of the vehicle based on the length of the shadow area directly under the vehicle.

  According to the vehicle length measurement device of the present invention, a vehicle is detected from an image, and a shadow region directly below the vehicle is extracted from the shadow region of the vehicle, and the length of the shadow region directly below the vehicle is extracted. Based on this, the length of the vehicle is calculated. For this reason, even if only a part of the front, side, back, etc. of the vehicle is photographed in the image and the vehicle length cannot be calculated only from the shape of the vehicle in the image, the vehicle from the image In particular, a shadow area directly under the vehicle is extracted from the shadow areas, and the length of the vehicle can be accurately calculated based on the length of the shadow area directly under the vehicle.

  In the vehicle length measurement device according to the present invention, the vehicle shadow extraction means extracts a low luminance area on the lower side of the vehicle as a shadow area of the vehicle, divides the low luminance area into a plurality of divided areas based on luminance, It is preferable to extract a divided area having the lowest luminance among the plurality of divided areas as a shadow area directly under the vehicle. According to this configuration, it is possible to suitably extract the shadow region directly under the vehicle.

  In the vehicle length measuring device according to the present invention, the vehicle shadow extracting means extracts the low luminance area on the lower side of the vehicle as the shadow area of the vehicle, and the shadow area on the lower oblique side of the vehicle based on the height of the vehicle. It is preferable to extract a region obtained by removing the shadow region obliquely below the vehicle from the shadow region of the vehicle as a shadow region directly below the vehicle. According to this configuration, it is possible to suitably extract the shadow region directly under the vehicle.

  In the vehicle length measurement device according to the present invention, the vehicle length calculation means projects the shadow area directly below the vehicle onto an overhead view as viewed from above the vehicle, and the length of the shadow area immediately below the vehicle in the overhead view. Is preferably calculated as the length of the vehicle. According to this configuration, the length of the vehicle can be accurately calculated by calculating the length of the shadow region directly under the vehicle.

  In order to achieve the above-described object, a vehicle type determination device according to the present invention separates and extracts a vehicle detection unit that detects a vehicle from an image and a shadow region directly below the vehicle from the image. Vehicle shadow extraction means, vehicle length calculation means for calculating the length of the vehicle based on the length of the shadow area directly under the vehicle, and vehicle type determination means for determining the vehicle type from the length of the vehicle. Features.

  According to the vehicle type determination device of the present invention, a vehicle is detected from an image, and a shadow area directly below the vehicle is extracted from the shadow area of the vehicle, and the length of the shadow area immediately below the vehicle is extracted. The vehicle length is calculated based on the above, and the vehicle type is determined from the vehicle length. For this reason, even if only a part of the front, side, back, etc. of the vehicle is photographed in the image and the vehicle length cannot be calculated only from the shape of the vehicle in the image, the vehicle from the image In particular, a shadow area directly under the vehicle is extracted from the shadow areas, and the length of the vehicle can be accurately calculated based on the length of the shadow area directly under the vehicle. Therefore, the vehicle type of the vehicle can be accurately determined based on the accurately calculated vehicle length.

  In the vehicle type determination device according to the present invention, the vehicle shadow extraction means extracts the low luminance area on the lower side of the vehicle as a shadow area of the vehicle, divides the low luminance area into a plurality of divided areas based on the luminance, It is preferable to extract a divided area having the lowest luminance among the plurality of divided areas as a shadow area directly under the vehicle. According to this configuration, it is possible to suitably extract the shadow region directly under the vehicle.

  In the vehicle type determination device according to the present invention, the vehicle shadow extraction means extracts a low luminance region on the lower side of the vehicle as a shadow region of the vehicle, and determines a shadow region on the lower oblique side of the vehicle based on the height of the vehicle. It is preferable to extract an area obtained by estimating and removing a shadow area obliquely below the vehicle from the shadow area of the vehicle as a shadow area directly below the vehicle. According to this configuration, it is possible to suitably extract the shadow region directly under the vehicle.

  In the vehicle type determination device according to the present invention, the vehicle length calculation means projects a shadow region directly under the vehicle onto an overhead view as viewed from above the vehicle, and calculates the length of the shadow region immediately below the vehicle in the overhead view. It is preferable to calculate the length of the vehicle. According to this configuration, the length of the vehicle can be accurately calculated by calculating the length of the shadow region directly under the vehicle.

  ADVANTAGE OF THE INVENTION According to this invention, the vehicle length measuring apparatus which calculates the length of a vehicle more correctly, and the vehicle type determination apparatus using it can be provided.

  Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the description, the same reference numerals are used for the same elements or elements having the same function, and redundant description is omitted.

  FIG. 1 shows an apparatus (hereinafter referred to as an in-vehicle apparatus) 10 mounted on an automobile according to an embodiment of the present invention. The in-vehicle device 10 according to the embodiment of the present invention is a vehicle length measuring device for measuring the length of the vehicle, and is a vehicle type determining device for determining the type of the vehicle.

  The in-vehicle device 10 includes a camera 12 for photographing the front of the host vehicle 1 and an image processing device 20 for processing an image photographed by the camera 12. The image processing apparatus 20 physically includes a CPU (Central Processing Unit), a RAM (Random Access Memory), a ROM (Read Only Memory), and the like. In addition, the image processing device 20 functionally includes a vehicle detection unit 21, a first shadow region extraction unit 22, a second shadow region extraction unit 23, an overhead image generation unit 24, a vehicle length calculation unit 25, and a vehicle type determination unit 26. It is comprised including.

  The vehicle detection unit 21 processes the image in front of the host vehicle 1 captured by the camera 12 and detects the preceding vehicle 50 existing in front of the host vehicle 1. Here, the vehicle detection unit 21 may use an existing detection technique such as pattern matching in order to detect the preceding vehicle 50. For example, in the image shown in FIG. 2, the preceding vehicle 50 in the center of the image is detected.

  The first shadow region extraction unit 22 extracts the pixel values of the preceding vehicle 50 and surrounding pixels detected by the vehicle detection unit 21 from the image in front of the host vehicle 1 taken by the camera 12, and extracts the plurality of pixels. For each, it is determined whether or not the luminance value is equal to or less than a predetermined threshold value TH1 for shadow area detection. Then, the first shadow region extraction unit 22 extracts the regions 52 and 54 including pixels whose luminance values are equal to or less than the predetermined threshold TH1 for detecting the shadow region as the shadow regions 52 and 54 of the preceding vehicle 50.

  The shadow areas 52 and 54 of the preceding vehicle 50 extracted by the first shadow area extracting unit 22 include a dark shadow area 52 that is directly below the preceding vehicle 50 and a relatively bright shadow that is obliquely below the preceding vehicle 50. Region 54. The shadow area 52 directly below the preceding vehicle 50 is a shadow area that is formed when sunlight is blocked by the preceding vehicle 50 and receives almost no reflected light from the surroundings, so that it is significantly darker than the surrounding shadow area. It has become. On the other hand, the shadow area 54 obliquely below the preceding vehicle 50 is a shadow area that is formed when sunlight is blocked by the preceding vehicle 50. However, since it receives reflected light from the surrounding area, the shadow area 54 is relatively smaller than the surrounding shadow area. It is getting brighter.

  The second shadow region extraction unit 23 extracts a dark shadow region 52 directly below the preceding vehicle 50 from the shadow regions 52 and 54 extracted by the first shadow region extraction unit 22. The second shadow area extraction unit 23 uses either the first method or the second method described below in order to extract a shadow area 52 directly below the preceding vehicle 50. Note that the second shadow area extraction unit 23 may use both the first method and the second method.

  In the first method, the second shadow region extraction unit 23 divides the shadow regions 52 and 54 extracted by the first shadow region extraction unit 22 into two or more regions based on the luminance value, and provides different luminance value ranges. A plurality of divided regions are generated. Then, the second shadow area extraction unit 23 extracts a divided area having the lowest luminance value range among the plurality of divided areas as the dark shadow area 52 immediately below the preceding vehicle 50. Here, in order to divide the shadow areas 52 and 54, the second shadow area extraction unit 23 uses one or more predetermined threshold values TH2 for dividing the shadow area. For example, when there is one predetermined threshold TH2 for shadow area division, the second shadow area extraction unit 23 has a shadow area 52 whose luminance value is smaller than the predetermined threshold TH2 and a luminance value that is the predetermined threshold TH2. Is divided into a larger shadow area 54.

  Note that the second shadow area extraction unit 23 generates a histogram of the luminance distribution of the shadow areas 52 and 54 when dividing the shadow areas 52 and 54 into a plurality of areas, and uses this histogram for shadow area division. The predetermined threshold TH2 may be determined. That is, as shown in FIG. 3, when the histogram of the luminance distribution of the shadow regions 52 and 54 has a relatively high luminance value peak and a relatively low luminance value peak, the shadow region 52 and Since 54 is assumed to be composed of two separable areas, a predetermined threshold TH2 for shadow area division is set between the two peaks. As a result, the second shadow region extraction unit 23 separates the shadow region 52 having a relatively low luminance value and the shadow region 54 having a relatively high luminance value, so that the dark shadow region 52 directly below the preceding vehicle 50 is obtained. Can be extracted.

  In the second method, the second shadow region extraction unit 23 detects the height of the preceding vehicle 50 from the image captured by the camera 12. Next, the second shadow area extraction unit 23 calculates the length of the shadow according to the height of the preceding vehicle 50 based on the height of the preceding vehicle 50 and the altitude of the sun. Here, the calculated shadow length corresponds to the length of the shadow region 54 that is obliquely below the preceding vehicle 50. Then, the second shadow region extraction unit 23 removes the length of the shadow region 54 due to the height of the preceding vehicle 50 from the shadow regions 52 and 54 extracted by the first shadow region extraction unit 22. As a result, the second shadow area extraction unit 23 can extract the dark shadow area 52 immediately below the preceding vehicle 50.

  The overhead image generation unit 24 performs projective conversion on the dark shadow region 52 immediately below the preceding vehicle 50 extracted by the second shadow region extraction unit 23, and generates an overhead image when the preceding vehicle 50 is viewed from above. For example, the overhead image generation unit 24 performs projective conversion on the image shown in FIG. 2 to generate the overhead image shown in FIG. In FIG. 4, in order to facilitate understanding, in addition to the shadow region 52 after projective transformation, the preceding vehicle 50 after projective transformation is shown.

With reference to FIG. 5, the overhead image generation method by the overhead image generation unit 24 will be described. The bird's-eye view image generation unit 24 captures an image captured by the in-vehicle camera 12 attached to the center upper portion of the host vehicle 1 with the virtual camera 12 ′ virtually existing above the preceding vehicle 50. Projective conversion to a bird's-eye view image. Here, the pixel coordinates of the image photographed by the camera 12 are defined as m _i = (u _i , v _i , 1) ^T, and the pixel coordinates of the overhead image are m _v = (u _v , v _v , 1) ^T Then, the projective transformation by the overhead image generation unit 24 is expressed by the following formula (1).

  In Equation (1), Rv is the rotational component of the motion from the in-vehicle camera 12 to the virtual camera 12 ′, tv is the translational component of the motion from the in-vehicle camera 12 to the virtual camera 12 ′, and K is the in-vehicle camera. 12 are internal variables, and n is a normal vector of the road. In the equation (1), the coordinates of the point in the two-dimensional image are expressed as m = (u, v, 1), and the coordinates of the point in the three-dimensional space are M = (X, Y, Z, 1). It expresses. Before projective transformation, the vehicle width direction is the X axis, the vehicle height direction is the Y axis, and the vehicle longitudinal direction is the Z axis. Further, after projective transformation, the vehicle width direction is taken as the X axis, the vehicle longitudinal direction is taken as the Y axis, and the vehicle height direction is taken as the Z axis.

  The vehicle length calculation unit 25 calculates the length of the preceding vehicle 50 using the overhead view image. As shown in FIG. 4, in the overhead view image, the vertical direction is the vehicle front-rear direction, and the horizontal direction is the vehicle width direction. The vehicle length calculation unit 25 counts the number of pixels in the shadow region 52 arranged in the vehicle width direction for each coordinate value in the vehicle front-rear direction of the overhead view image. As a result, a histogram of the number of pixels in the shadow area 52 shown on the right side in FIG. 4 is obtained. In this histogram, the count value of the number of pixels of the shadow region 52 is shown in the horizontal direction with the length of the bar. In this histogram, the vehicle length calculation unit 25 determines the length of the section in the vehicle front-rear direction in which the count value of the number of pixels in the shadow region 52 is continuously equal to or greater than the vehicle length detection threshold TH3 as the length of the preceding vehicle 50. Calculate as

  The vehicle type determination unit 26 determines the type of the preceding vehicle 50 based on the length of the preceding vehicle 50 calculated by the vehicle length calculation unit 25. That is, the vehicle type determination unit 26 has a database in which the vehicle length is registered for each type of the preceding vehicle 50, and the length of the preceding vehicle 50 calculated by the vehicle length calculation unit 25 is registered in the database. When the vehicle length is approximated, it is determined that the type of the preceding vehicle 50 is the type of the matched vehicle. Thereby, the vehicle type determination unit 26 can determine that the type of the preceding vehicle 50 is any one of a small vehicle, a medium vehicle, a large vehicle, and a large truck.

  Next, a processing procedure performed by the image processing device 20 of the in-vehicle device 10 will be described with reference to the flowchart of FIG. Note that the processing of the flowchart in FIG. 6 is realized by the image processing apparatus 20 executing a program.

  The image processing apparatus 20 captures an image captured by the camera 12 and detects the preceding vehicle 50 from within the image (S601, S602). Next, the image processing apparatus 20 extracts dark shadow regions 52 and 54 on the lower side of the detected preceding vehicle 50 (S603). Further, the image processing apparatus 20 divides the shadow areas 52 and 54 into a plurality of divided areas on the basis of the luminance value of the pixel, and extracts the darkest divided area as the shadow area 52 directly under the vehicle (S604, S605). Next, the image processing device 20 projects the shadow region 52 directly below the vehicle onto the overhead image, and calculates the length of the preceding vehicle 50 from the length of the shadow region 52 in the overhead image (S606, S607).

  According to the in-vehicle device 10 of the present embodiment, the preceding vehicle 50 is detected from the image captured by the camera 12, and the shadow region 52 immediately below the preceding vehicle 50 is separated from the shadow regions 52 and 54 of the preceding vehicle 50. The length of the preceding vehicle 50 is calculated based on the length of the shadow region 52 directly below the preceding vehicle 50, and the vehicle type of the preceding vehicle 50 is determined from the length of the preceding vehicle 50. For this reason, even when only the back surface of the preceding vehicle 50 is photographed in the image photographed by the camera 12, the length of the preceding vehicle 50 cannot be calculated from only the shape of the preceding vehicle 50 in the image. The shadow area 52 that is directly below the preceding vehicle 50 is extracted from the shadow areas 52 and 54 of the preceding vehicle 50, and the length of the preceding vehicle 50 is determined based on the length of the shadow area 52 that is immediately below the preceding vehicle 50. It can be calculated accurately. Further, the vehicle type of the preceding vehicle 50 can be accurately determined based on the accurately calculated length of the preceding vehicle 50.

  In the embodiment described above, the in-vehicle device 10 processes an image obtained by capturing only the back surface of the preceding vehicle 50 to calculate the length of the preceding vehicle 50 in the vehicle front-rear direction and determines the vehicle type. The invention is not limited to this. In another embodiment, the in-vehicle device 10 may calculate an length of the preceding vehicle 50 in the vehicle front-rear direction or determine a vehicle type by processing an image obtained by capturing only the front surface of the preceding vehicle 50. In another embodiment, the in-vehicle device 10 may process an image obtained by capturing only the side surface of the preceding vehicle 50 to calculate the length in the vehicle width direction of the preceding vehicle 50 or determine the vehicle type.

It is a block diagram which shows the structure of the vehicle-mounted apparatus which concerns on embodiment of this invention. It is a figure which shows the image containing the preceding vehicle image | photographed with the camera. It is a graph of the luminance value-frequency in the shadow area of the preceding vehicle. It is a figure which shows the image after projective transformation. It is a schematic diagram for demonstrating projective transformation. It is a flowchart which shows the process of a vehicle-mounted apparatus.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Own vehicle, 10 ... In-vehicle apparatus (vehicle length measuring device, vehicle type determination apparatus), 12 ... Camera, 12 '... Virtual camera, 20 ... Image processing device, 21 ... Vehicle detection part, 22 ... 1st shadow area extraction part , 23 ... second shadow region extraction unit, 24 ... overhead view image generation unit, 25 ... vehicle length calculation unit, 26 ... vehicle type determination unit, 50 ... preceding vehicle, 52, 54 ... shadow region.

Claims (8)

Vehicle detection means for detecting the vehicle from the image;
Vehicle shadow extraction means for separating and extracting a shadow area directly below the vehicle from the image of the shadow area of the vehicle;
Vehicle length calculating means for calculating the length of the vehicle based on the length of the shadow region directly under the vehicle;
A vehicle length measuring device comprising:   The vehicle shadow extraction means extracts a low luminance area below the vehicle as a shadow area of the vehicle, divides the low luminance area into a plurality of divided areas based on luminance, and The vehicle length measuring device according to claim 1, wherein a divided area having the lowest luminance is extracted as a shadow area directly below the vehicle.   The vehicle shadow extraction means extracts a low luminance area on the lower side of the vehicle as a shadow area of the vehicle, estimates a shadow area on an oblique lower side of the vehicle based on the height of the vehicle, The vehicle length measurement device according to claim 1, wherein an area obtained by removing a shadow area obliquely below the vehicle from a shadow area of the vehicle is extracted as a shadow area immediately below the vehicle.   The vehicle length calculation means projects a shadow area directly below the vehicle onto an overhead view as viewed from above the vehicle, and calculates the length of the shadow area directly below the vehicle in the overhead view as the length of the vehicle. The vehicle length measuring device according to claim 1, wherein the vehicle length measuring device is calculated as follows. Vehicle detection means for detecting the vehicle from the image;
Vehicle shadow extraction means for separating and extracting a shadow area directly below the vehicle from the image of the shadow area of the vehicle;
Vehicle length calculating means for calculating the length of the vehicle based on the length of the shadow region directly under the vehicle;
Vehicle type determination means for determining the vehicle type from the length of the vehicle;
A vehicle type determination device comprising:   The vehicle shadow extraction means extracts a low luminance area below the vehicle as a shadow area of the vehicle, divides the low luminance area into a plurality of divided areas based on luminance, and 6. The vehicle type determination device according to claim 5, wherein a divided area having the lowest luminance among the divided areas is extracted as a shadow area directly below the vehicle.   The vehicle shadow extraction means extracts a low luminance area on the lower side of the vehicle as a shadow area of the vehicle, estimates a shadow area on an oblique lower side of the vehicle based on the height of the vehicle, The vehicle type determination device according to claim 5, wherein an area obtained by removing a shadow area obliquely below the vehicle from a shadow area of the vehicle is extracted as a shadow area directly below the vehicle.   The vehicle length calculation means projects a shadow area directly below the vehicle onto an overhead view as viewed from above the vehicle, and calculates the length of the shadow area directly below the vehicle in the overhead view as the length of the vehicle. The vehicle type determination device according to claim 5, wherein the vehicle type determination device is calculated as follows.

Images