JP7640431B2 - Headlight tester beam axis calculation system - Google Patents

Description

The present invention relates to a headlight tester optical axis calculation system. In particular, the present invention relates to improvements to achieve high optical axis calculation accuracy.

Headlight testers have traditionally been used to test vehicle headlamps to see whether their light distribution patterns (specifically, the light distribution patterns of low-beam headlamps for passing vehicles) are within the range of standards set by vehicle safety regulations.

As disclosed in Patent Document 1, for example, a light distribution pattern test using this headlight tester involves facing the vehicle directly to the headlight tester, capturing an image of the light distribution pattern projected from the headlights onto a screen using a camera or the like, and then performing image processing on the captured image to obtain an elbow point of the cutoff line (also called a clear boundary line), and inspecting whether this elbow point is located within a predetermined standard range (for example, the amount of deviation of the elbow point when projecting a predetermined distance ahead is within a predetermined range). As is well known, this elbow point is the intersection of the horizontal cutoff line (a horizontal cutoff line set on the oncoming vehicle side) and the diagonal cutoff line (a diagonal cutoff line set on the vehicle's own side (shoulder side)) in the light distribution pattern. In order to improve the accuracy of calculating the optical axis (the direction of light emitted from the headlights), it is necessary that this elbow point be obtained stably.

JP 2011-79359 A

Calculation of the optical axis of the light distribution pattern requires high accuracy, but in an image of a light distribution pattern captured by a camera or the like, noise exists for each pixel. For this reason, minute changes in the amount of lamp light cause changes in the luminance value, particularly near the edge (the point in the light distribution pattern where the amount of change in luminance value in the y direction (up and down direction) is maximum), making the cutoff line unstable and causing the elbow point to flicker on the image. In such a situation, the elbow point cannot be obtained stably, and therefore high accuracy in calculating the optical axis cannot be obtained.

Although it is possible to reduce the noise by increasing the exposure of the camera, in this case, the luminance of the entire image increases, which makes it more likely that the luminance will reach the upper gradation limit, preventing the application of an algorithm that uses a luminance gradient (an algorithm that finds the position of the maximum amount of change in luminance value as an edge). For this reason, with conventional technology, it has been difficult to reduce noise and apply an algorithm that uses a luminance gradient at the same time.

The present invention was made in consideration of these points, and its purpose is to provide a headlight tester optical axis calculation system that can simultaneously reduce noise and apply an algorithm that uses a luminance gradient.

The solution of the present invention for achieving the above object is premised on an optical axis calculation system for a headlight tester that calculates the optical axis of a vehicle headlamp. The optical axis calculation system for the headlight tester includes an image capture means capable of capturing images with multiple exposures, a composite luminance value calculation unit that calculates a composite luminance value of each pixel from each image acquired by capturing images with multiple exposures with the image capture means, and an optical axis calculation unit that calculates the optical axis of the headlamp based on the composite luminance value calculated by the composite luminance value calculation unit, and the optical axis calculation unit performs edge calculation to detect a position where the amount of change in the composite luminance value becomes large when the area near the cutoff line in the vertical direction is scanned in the captured image. the edge calculation operation, a differentiation operation for performing a second differentiation operation on the line recognized as an edge by the edge calculation operation, and a cutoff line setting operation for performing straight-line approximation processing based on the maximum point and the minimum point in the curve obtained by the differentiation operation to obtain a first horizontal cutoff line extending horizontally to one side of the position corresponding to the maximum point and a second horizontal cutoff line extending horizontally to the other side of the position corresponding to the minimum point, and for obtaining an oblique cutoff line connecting the maximum point and the minimum point, and for determining, as the elbow point, an intersection point of the first horizontal cutoff line and the oblique cutoff line and an intersection point of the second horizontal cutoff line and the oblique cutoff line, respectively .

Due to this specification, the composite luminance value calculated from each image acquired by multiple exposure shooting using the shooting means is obtained as a value in which the noise present in each image, which is exposed differently from the others, is reduced by canceling out the noise. In addition, since the luminance value is calculated by combining multiple images, information on the luminance value with infinite gradations is obtained. Therefore, the cutoff line of the light distribution pattern on the image obtained according to this composite luminance value can be obtained as a stable one, and the adjustment point of the optical axis (e.g., elbow point) can be obtained stably, resulting in high optical axis calculation accuracy.

In the present invention, a composite luminance value of each pixel is calculated from each image acquired by multiple exposure photography using a photography means, and the optical axis of the headlamp is calculated based on the composite luminance value. This makes it possible to achieve both noise reduction and the application of an algorithm that uses a luminance gradient.

FIG. 1 is a schematic diagram showing a schematic configuration of an optical axis calculation system. FIG. 2 is a control block diagram of the optical axis calculation system. 1A and 1B are diagrams showing an example of each image acquired by multiple exposure shooting and an example of a composite image of the images; FIG. 13 is a diagram showing an example of the luminance value for each exposure in one pixel of each image acquired by multiple exposure shooting. 11A and 11B are diagrams for explaining a data narrowing-down operation in the composite luminance value calculation process. 11A and 11B are diagrams for explaining a linear model creation operation in the composite luminance value calculation process. 11A and 11B are diagrams for explaining a reference exposure calculation operation in the composite brightness value calculation process. 13A to 13C are diagrams for explaining an edge calculation operation of a composite image in the optical axis calculation process. 13A to 13C are diagrams for explaining a differential operation for an edge of a composite image in the optical axis calculation process. 11A and 11B are diagrams for explaining a cutoff line setting operation in the optical axis calculation process.

The following describes an embodiment of the present invention with reference to the drawings.

-Outline of the optical axis calculation system-
1 is a schematic diagram showing a schematic configuration of an optical axis calculation system 1 of a headlight tester according to this embodiment. Although not shown, this optical axis calculation system 1 is supported by a support pillar erected on a dolly of the headlight tester. The dolly is free to move in the horizontal direction (for example, horizontal movement on a rail). Moreover, the optical axis calculation system 1 is free to move in the vertical direction relative to the support pillar (up and down movement by a lifting mechanism). This allows the optical axis calculation system 1 to be directly faced to the headlamp HL of the vehicle to be tested.

The optical axis calculation system 1 is configured such that a Fresnel lens 3, a screen 4, and an optical axis camera (imaging means) 5 are housed in a casing 2, and a facing camera 6 is mounted on the top of the casing 2.

The Fresnel lens 3 is disposed on the front surface of the casing 2 (the surface facing the vehicle headlamps HL) and has a central optical axis extending horizontally. It focuses the light (illuminated light) emitted from the vehicle headlamps HL and transmits it toward the inside of the casing 2 (toward the screen 4).

The screen 4 receives the irradiated light collected by the Fresnel lens 3. As a result, the light distribution pattern of the light irradiated from the headlamp HL of the vehicle is projected onto the screen 4. In this light distribution pattern, a large difference in brightness occurs at the boundary between the part where the light hits and becomes bright and the part where the light does not hit and becomes dark, resulting in a cutoff line. As mentioned above, this cutoff line consists of a horizontal cutoff line and an oblique cutoff line. The horizontal cutoff line is set on the oncoming vehicle side and extends horizontally at a predetermined height so as not to dazzle the driver of the oncoming vehicle. The oblique cutoff line is set on the vehicle side (shoulder side) and extends upward toward the shoulder side to increase the visibility of pedestrians walking on the shoulder. The intersection point of these horizontal cutoff lines and oblique cutoff lines is called the elbow point, and in testing the light distribution pattern with a headlight tester, it is checked whether this elbow point is within a predetermined standard range (for example, the amount of deviation of the elbow point when irradiating a specified distance ahead is within a predetermined range). In addition, a horizontal cutoff line may also be set closer to the shoulder than the diagonal cutoff line.

The optical axis camera 5 is, for example, a CCD (Charge Coupled Device) camera or a CMOS (Complementary Metal Oxide Semiconductor) camera, and is disposed on the rear side of the screen 4 so that its optical axis coincides with the optical axis of the Fresnel lens 3. The optical axis camera 5 is also capable of multiple exposure photography, and is capable of photographing the light distribution pattern of the irradiated light projected onto the screen 4 while changing the exposure in multiple steps (for example, 15 steps). The number of multiple exposure steps is not limited to those mentioned above. Exposure adjustment parameters for realizing this multiple exposure include the shutter speed, the aperture value of the photographing lens, and the ISO sensitivity value.

The facing camera 6 is used to face the optical axis calculation system 1 (more specifically, the Fresnel lens 3 and the optical axis camera 5) to the headlamp HL of the vehicle, and is used to determine whether the headlamp HL is facing forward by photographing a vehicle stopped in front of the optical axis calculation system 1. In other words, the headlight tester performs optical axis calculation when it is determined that the headlamp HL is facing forward based on the image photographed by this facing camera 6. Note that a known facing laser sight or the like can be used instead of this facing camera 6.

- Configuration of the control system for the optical axis calculation system -
Next, the configuration of the control system of the optical axis calculation system 1 will be described. Fig. 2 is a control block diagram of the optical axis calculation system 1. The control system of the optical axis calculation system 1 includes, as a hardware configuration, a processor such as a CPU (Central Processing Unit), a ROM (Read-Only Memory) that stores a control program, a RAM (Random-Access Memory) that temporarily stores data, and an input/output port. These hardware components may be built into the optical axis camera 5, or may be configured as a control unit separate from the optical axis camera 5.

The control system of the optical axis calculation system 1 has a function of calculating a composite brightness value of each pixel from each image acquired by multiple exposure shooting with the optical axis camera 5, and calculating the optical axis of the headlamp HL (the irradiation direction of the light emitted from the headlamp HL) based on this calculated composite brightness value. Figure 3 is a diagram showing an example of each image acquired by multiple exposure shooting and their composite image. Figures 3(a) and 3(b) show examples of each image captured with different exposures. Although only two images are shown here, images corresponding to the number of stages of multiple exposure will be captured. For example, 15 images with different exposures will be captured. Also, Figure 3(c) shows an example of a composite image created based on these images (each image with a different exposure).

To realize the above-mentioned function of calculating the composite brightness value, the control system of the optical axis calculation system 1 includes an image receiving unit 71, a composite brightness value calculation unit 72, and an optical axis calculation unit 73 as functional units realized by the control program, as shown in FIG. 2.

The image receiving unit 71 acquires image data (image data by exposure) taken by the optical axis camera 5 at different exposures for the same test object (light distribution pattern of the light emitted by the headlamp HL). The image data is temporarily stored in the RAM.

The composite luminance value calculation unit 72 uses the data of each image to calculate a composite luminance value using the change in luminance value in response to a change in exposure for each pixel on the image.

The composite brightness value calculation process performed by the composite brightness value calculation unit 72 will be described below. In this composite brightness value calculation process, a data narrowing operation, a linear model creation operation, and a reference exposure calculation operation are performed in this order.

First, when the composite brightness value calculation process starts, brightness value data corresponding to the exposure at each pixel in each image acquired by multiple exposure shooting is acquired. This means that when shooting is performed with a certain exposure, the brightness value of each pixel on the captured image is acquired, and the exposure and brightness value are acquired as information corresponding to each pixel. Therefore, when shooting with all exposures is completed, information corresponding to the number of exposures and brightness values corresponding to the product of the number of exposures and the number of pixels in the image is acquired. Figure 4 is a diagram showing an example of the brightness value (vertical axis) for each exposure (horizontal axis) in one pixel of each image acquired by multiple exposure shooting. As shown in Figure 4, most brightness values increase as the exposure increases. Note that in Figure 4, there is data in which the brightness value appears small even when the exposure increases due to the influence of disturbances, etc.

Figure 5 is a diagram for explaining the data narrowing down operation. This data narrowing down operation is intended to eliminate data from the data shown in Figure 4 where the luminance value for the exposure may not be appropriately obtained due to the influence of disturbances, etc. In Figure 5, data in the range where the exposure is less than a predetermined value ex1 and in the range where the exposure is more than a predetermined value ex2 (data in the shaded range in Figure 5) are removed, so that only information with high reliability in the relationship between exposure and luminance value remains. The predetermined values ex1 and ex2 can be set arbitrarily.

Figure 6 is a diagram for explaining the linear model creation operation. For the data narrowed down by the data narrowing down operation described above (data showing the relationship between exposure and luminance value), a linear model "y = ax + b" is obtained by the least squares method. The coefficient a here represents the slope of the linear model, and indicates the rate of increase in luminance value relative to an increase in exposure.

For example, the area above the cutoff line (the area above the area near the cutoff line) is a dark area where light does not reach, so the brightness value hardly increases even if the exposure is increased. For this reason, the pixels in this area are obtained as having a small slope (coefficient a) in the linear model. Also, the area below the cutoff line (the area below the area near the cutoff line; for example, a hot zone with the maximum brightness value and its surrounding area) is a bright area where light reaches, so in this case too, the increase in brightness value is small even if the exposure is increased. For this reason, the pixels in this area are obtained as having a small slope in the linear model. In contrast, in the area near the cutoff line, the brightness value increases significantly as the exposure is increased. For this reason, the pixels in this area are obtained as having a large slope in the linear model.

Figure 7 is a diagram for explaining the reference exposure calculation operation. In this reference exposure calculation operation, an exposure corresponding to a preset reference value Y0 of the luminance value is calculated as a reference exposure X0 for the data (linear model data) obtained in the linear model creation operation. This reference value Y0 of the luminance value can be set arbitrarily.

After the reference exposure X0 is calculated for each pixel in this way, the reciprocal (1/X0) is calculated and set as the composite luminance value corresponding to each pixel. In other words, the reciprocal of each reference exposure X0 (element of the set) in the set (X0∈R ⁺ ) of the reference exposures X0 calculated for each pixel is set as the composite luminance value v (v=1/X0).

As described above, in the area above the cutoff line and the area below the cutoff line, the slope of the linear model is small, so the reference exposure X0 calculated for these pixels is relatively large, and the composite luminance value (1/X0), which is the reciprocal of this reference exposure X0, is obtained as a small value. In contrast, in the area near the cutoff line, the slope of the linear model is large, so the reference exposure X0 calculated for these pixels is relatively small, and the composite luminance value (1/X0), which is the reciprocal of this reference exposure X0, is obtained as a large value. As a result, when the area near the cutoff line is scanned in the vertical direction, there are positions where the change in luminance value (composite luminance value) is large, and by recognizing this position as the edge of the light distribution pattern, it is possible to stably obtain the edge.

Next, the optical axis calculation process performed by the optical axis calculation unit 73 will be described. In this optical axis calculation process, a composite image is created from the composite luminance value for each pixel calculated by the composite luminance value calculation unit 72, and an edge calculation operation is performed to calculate edges on this composite image, a differentiation operation is performed on the edges of the composite image, and a cutoff line setting operation is performed to set a cutoff line, in that order.

The edge calculation operation is performed by cutting out a specific area (the area where the elbow point is estimated to exist) from the composite image, and then detecting the position where the amount of change in the luminance value (composite luminance value) becomes large when scanning the area near the cutoff line in the vertical direction as described above. The line ED in Figure 8 represents the edge set by this edge calculation operation.

In the differentiation operation, a second differentiation operation is performed on the line ED recognized as an edge by the edge calculation operation. Figure 9 is a diagram (a diagram showing an enlarged portion of the image) for explaining this differentiation operation. The dashed dotted line in Figure 9 is an example of a curve obtained when a first differentiation operation is performed on the line ED recognized as an edge. The dashed two dotted line in Figure 9 is an example of a curve obtained when a second differentiation operation is performed on the line ED recognized as an edge. The curve obtained by this second differentiation operation has a maximum point (maximum point) P1 and a minimum point (minimum point) P2.

In the cutoff line setting operation, as shown in FIG. 10(a), a cutoff line is set by performing a linear approximation process with a BOX based on the maximum point P1 and the minimum point P2 in the curve obtained by the second differentiation operation. As shown in FIG. 10(b), the first horizontal line (horizontal cutoff line) CL1 extending horizontally to the left of the position corresponding to the maximum point P1 and the second horizontal line CL2 extending horizontally to the right of the position corresponding to the minimum point P2 are obtained as the cutoff line. In this case, the second horizontal line CL2 is located a predetermined dimension above the first horizontal line CL1 on the image. Then, a diagonal line (diagonal cutoff line) CL3 connecting the maximum point P1 and the minimum point P2 is obtained by the linear approximation process with the BOX. Then, the intersection point EL1 between the first horizontal line CL1 and the diagonal line CL3 and the intersection point EL2 between the second horizontal line CL2 and the diagonal line CL3 are obtained as elbow points. As mentioned above, in the area near the cutoff line, a stable edge is obtained based on the composite brightness value, so the cutoff line of the light distribution pattern on the image can also be obtained stably, the elbow points EL1 and EL2 can be obtained stably, and high optical axis calculation accuracy can be achieved.

When determining whether or not the elbow point is within a specified standard range during an actual test (a test of the light distribution pattern using a headlight tester), the elbow point used in the test will differ depending on the destination of the vehicle. For example, in Japan, etc., it will be determined whether or not the intersection (elbow point) EL1 of the first horizontal line CL1 and the diagonal line CL3 is within the standard range, and in the United States, etc., it will be determined whether or not the intersection (elbow point) EL2 of the second horizontal line CL2 and the diagonal line CL3 is within the standard range.

Furthermore, in a destination where traffic is on the left side of the road, the first horizontal line CL1 is located on the right side (oncoming vehicles side) when viewed from the vehicle side, and the diagonal line CL3 and the second horizontal line CL2 are located on the left side (shoulder side). In this embodiment, the optical axis camera 5 is disposed on the rear side of the screen 4, so that the acquired image has the first horizontal line CL1 located on the left side, and the diagonal line CL3 and the second horizontal line CL2 located on the right side, as shown in Figures 8 and 10. In contrast, in destinations where traffic is on the right side of the road, the first horizontal line CL1 is located on the left side (oncoming vehicles) as viewed from the vehicle side, and the diagonal line CL3 and the second horizontal line CL2 are located on the right side (shoulder side). In this embodiment, the image captured by the optical axis camera 5 is inverted left to right, so that the first horizontal line CL1 is located on the left side, and the diagonal line CL3 and the second horizontal line CL2 are located on the right side, as in the above-mentioned case (destinations where traffic is on the left side). This makes it possible to treat the same optical axis calculation system 1 equally in destinations where traffic is on the left side and in destinations where traffic is on the right side of the road, thereby increasing the versatility of the optical axis calculation system 1.

--Effects of the embodiment--
As described above, in this embodiment, the composite luminance value of each pixel is calculated from each image acquired by multiple exposure photography using the optical axis camera 5, and the elbow points EL1 and EL2 of the light distribution pattern of the headlamp HL are calculated based on the calculated composite luminance value. Therefore, the composite luminance value is obtained as a value in which noises that existed in each image with different exposures are canceled out by each other, and the noise is reduced. In addition, since the luminance value is calculated by combining multiple images, information on the luminance value with infinite gradation is obtained. Therefore, the cutoff line of the light distribution pattern on the image obtained according to this composite luminance value can be obtained as a stable one, and the adjustment point (elbow point) of the optical axis can be obtained stably, and high optical axis calculation accuracy can be obtained. In the conventional technology, when noise is reduced by increasing the exposure of the camera, the luminance easily reaches the upper gradation limit due to the increase in the luminance of the entire image, which prevents the application of an algorithm using a luminance gradient. However, in this embodiment, there is no need to use a means such as increasing the exposure of the camera to reduce noise, and therefore it is possible to achieve both noise reduction and application of an algorithm using a luminance gradient.

-Other embodiments-
The present invention is not limited to the above-described embodiment, and all modifications and applications within the scope of the claims and equivalents thereto are possible.

For example, in the above embodiment, the light distribution pattern projected on the screen 4 is photographed by the optical axis camera 5 from the rear side of the screen 4. The present invention is not limited to this, and can also be applied to an optical axis calculation system for a headlight tester in which the light emitted from the headlamp is split by a half mirror and the split light is photographed by the optical axis camera.

The present invention can be applied to a headlight tester optical axis calculation system that can achieve both noise reduction and the application of an algorithm using a luminance gradient.

1 Optical axis calculation system 5 Optical axis camera (photographing means)
72 Composite luminance value calculation unit 73 Optical axis calculation unit

Claims (1)

A headlight tester optical axis calculation system for calculating an optical axis of a headlamp of a vehicle, comprising:
A photographing means capable of photographing with multiple exposures;
a composite luminance value calculation unit that calculates a composite luminance value of each pixel from each image acquired by multiple exposure photography using the photography means;
an optical axis calculation unit that calculates an optical axis of the headlamp based on the composite luminance value calculated by the composite luminance value calculation unit ,
The optical axis calculation unit
an edge calculation operation of extracting an area where an elbow point, which is an intersection point between a horizontal cutoff line and an oblique cutoff line in a light distribution pattern, is estimated to exist from a composite image created from the composite luminance value for each pixel, and detecting a position where an amount of change in the composite luminance value becomes large when an area near the cutoff line is scanned in an up-down direction from the extracted image;
a differentiation operation for performing a second differentiation operation on the line recognized as an edge by the edge calculation operation;
a cutoff line setting operation for obtaining an intersection point between the first horizontal cutoff line and the oblique cutoff line and an intersection point between the second horizontal cutoff line and the oblique cutoff line as the elbow point, by performing straight-line approximation processing based on the maximum points and minimum points in the curve obtained by the differentiation operation, thereby obtaining a first horizontal cutoff line extending horizontally to one side of a position corresponding to the maximum point and a second horizontal cutoff line extending horizontally to the other side of the position corresponding to the minimum point, and also obtaining an oblique cutoff line connecting the maximum point and the minimum point, and

Images