JP3494915B2 - Reference position detection device for headlight optical axis adjustment - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a reference position detecting device for adjusting an optical axis of a headlight.

[0002]

2. Description of the Related Art In an automobile assembly line, after assembling a headlight, it is inspected whether or not the optical axis of the low beam of the headlight satisfies a predetermined standard.
If it does not meet the requirements, adjustment work is being performed.
Then, as a reference, the headlight (low beam) is actually turned on to irradiate the front screen, and from the bright / dark boundary line (cut line) of the irradiation pattern (projection pattern) formed on the screen, the optical axis To identify.

That is, although the projection pattern varies depending on the type of headlight used, the portion facing the headlight is the brightest area a as shown in FIG. 1, and the areas b, c, and d are closer to the periphery. , E, it becomes dark. Then, the area f having a certain height or more
Becomes very dark. That is, as a whole, almost no light is emitted above the headlight for a certain distance or more, and the area below the headlight is dark, and the area below the headlight is irradiated with light to a certain extent, so that the degree of darkening is reduced.

The lower limit of the pitch-dark area f becomes the above-mentioned cut line, and a horizontal line L1 extending in the horizontal direction.
And is separated by an inclined line L2 extending in an oblique direction. Then, the optical axis is set based on the position of the intersection of the horizontal line L1 and the inclined line L2.

When the cut line is visually determined, the vicinity of the cut line on the light projection pattern appears to be blurred, so that the variation among the inspectors is large and the detection accuracy is deteriorated. Therefore, there is a method of performing image processing on image data obtained by imaging a light projection pattern with a camera and automatically extracting the cut line, as disclosed in Japanese Patent Laid-Open No. 4-25741.
It has been developed by the invention disclosed in the publication.

The invention disclosed in this publication divides a given image into predetermined pixels, shifts the pixel of interest in the vertical direction, obtains the luminance of a pixel in the vicinity of the pixel of interest, and changes the luminance. A pixel of interest whose amount is a predetermined value is detected as a light / dark boundary point, and such processing is repeated at predetermined pixel intervals in the horizontal direction to extract a large number of bright / dark boundary points P in the horizontal direction as shown in FIG. Then, each light / dark boundary point P
A line L connecting (resembling a straight line) is obtained, and the line L is used as a cut line.

[0007]

However, the above-mentioned conventional method has the following problems. That is, the image data to be image-processed includes fine flicker of the reflector. Therefore, the flicker increases the amount of change in luminance in a portion that is not originally a light / dark boundary point, and as a result, the flicker portion may be detected as a light / dark boundary point. Therefore, the bright and dark points come to the wrong position, which causes erroneous detection of the cut line.

Further, in the conventional method in which the bright and dark boundary points are locally extracted and the line connecting them is used as the cut line, the bright and dark boundary points P adjacent to each other in the horizontal direction are also caused by the fluctuation of light.
2 varies in the vertical direction as shown in FIG. 2, it is not possible to actually draw a line that passes through each light / dark boundary point P, and even if the same light / dark boundary point P is detected, the line drawing There is also a problem that the position of the finally specified cut line varies.

The present invention has been made in view of the above background, and an object of the present invention is to solve the above problems and to cause flickering of a reflection plate and fluctuation of light (brightness and darkness when viewed microscopically). It is possible to specify the cut line easily and reliably without being affected by the vertical movement of the boundary position), and to specify the reference position when adjusting the optical axis of the headlight (low beam). A reference position detecting device for optical axis adjustment is provided.

[0010]

In order to achieve the above object, a reference position detecting device for adjusting an optical axis of a headlight according to the present invention performs an expansion process on grayscale image data of a given light projection pattern. Expansion processing means (corresponding to the “image expansion section 3” in the embodiment) and contraction processing means for performing contraction processing on the grayscale image data (corresponding to the “image contraction section 4” in the embodiment). A difference processing unit (corresponding to the “difference processing unit 5” in the embodiment) for obtaining a difference image between the expansion image and the contraction image generated by the expansion processing unit and the contraction processing unit respectively; Means for extracting a horizontal component area including a horizontal line forming a light-dark boundary line (cut line) in the light projection pattern (corresponding to the “image dividing unit 7” in the embodiment), and the extracted horizontal Pair with the component domain The horizontal line determining means (in the embodiment, the “horizontal line determining unit 8”) that specifies the horizontal line (corresponding to the symbol L1 in the embodiment) based on the density value integration information obtained by projecting in the vertical axis direction. ")
And a detection unit (corresponding to the “optical axis position determination unit 10” in the embodiment) for obtaining the optical axis adjustment reference position based on the horizontal line determined by the horizontal line determination unit. The means and the contraction processing means are configured to perform the expansion processing or the contraction processing using a relatively large mask that is not microscopically affected by flicker and the like (claim 1).

The projection pattern obtained by turning on the headlamp and irradiating it on the screen is divided into a bright part where the light hits and a dark part where the light does not hit. A light-dark boundary (cut line) that is the boundary of the part occurs. The cut line has a horizontal line extending in the horizontal direction and an oblique line extending in the oblique direction, and the horizontal line and the oblique line are continuous. In the present invention, the main subject is mainly to specify the horizontal line position and obtain the optical axis adjustment reference position based on the specified horizontal line position.

First, by obtaining the difference image, the density of the boundary portion between the light-exposed bright portion and the non-exposed dark portion of the projection pattern is increased. Since such a boundary exists around the bright part, it appears on the upper side and the lower side, and the cut line to be extracted is the upper side.
Due to the characteristics of the light projection pattern emitted from the headlamp, the cut line side to be detected has a large density change (abruptly dark) and the opposite side (lower side) has a small density change (gradually dark). Therefore, when the difference image is taken, of the boundaries around the bright portion, the cut line side becomes particularly bright.

Then, since the horizontal component area including the horizontal line forming the light-dark boundary line (cut line) is extracted based on the light-dark density value integration information obtained from the difference image, it is partially (microscopically). Even if there are variations in light and shade, the pixels on the horizontal line are bright, so the density value integration is particularly large in the horizontal line portion. Therefore, the horizontal line can be accurately specified based on the integration information.

Further, a rotation image is generated by rotating the difference image by a predetermined angle, and the reference line (embodiment) is obtained based on density value integration information obtained by projecting the rotation image in the vertical axis direction. Then, the diagonal line determination means (implemented) for determining the diagonal line (corresponding to the symbol L2 in the embodiment) by obtaining the reference line and performing the inverse transformation of the generation of the rotated image on the reference line. (Corresponding to the "diagonal line determination unit 9"), the detection means obtains an intersection of the horizontal line and the diagonal line, and sets it as the optical axis adjustment reference position. Good (Claim 2). This is realized by the second embodiment.

The lines that make up the cut line are:
In addition to the horizontal lines mentioned above, there are diagonal lines. By rotating the difference image by a predetermined angle, the pixels forming the diagonal line are arranged in a line parallel to the horizontal axis. Therefore, the position of the line (reference line) obtained by rotating the diagonal line can be specified based on the density value integration information obtained by projecting on the vertical axis similarly to the detection of the horizontal line described above. Therefore, the diagonal line can be specified by inversely converting the detected reference line (rotating it in the opposite direction by a predetermined angle).

A means (in the embodiment, corresponding to the "image dividing section 7") for extracting a vertical component area including an oblique line forming a bright / dark boundary line in the light projection pattern from the difference image. , For the extracted vertical component region,
A rotation image generated by rotating the rotation image by a predetermined angle is generated, the reference line is obtained based on density value integration information obtained by projecting the rotation image in the vertical axis direction, and the reference line is inversely transformed to the generation of the rotation image. By further comprising a diagonal line determining means for specifying the diagonal line by performing the, the detecting means is configured to obtain the intersection of the horizontal line and the diagonal line, and set it as the optical axis adjustment reference position. Good (Claim 3). This is realized by the first embodiment.

The calculation of the diagonal line is the same as in claim 2. When the diagonal line is obtained, the horizontal line is formed by performing it based on the vertical component area including the diagonal line, but the influence can be suppressed as much as possible.

Further, the detecting means obtains an intersection of the horizontal line and a vertical line (corresponding to the reference symbol L3 in the embodiment) where the optical axis adjusting reference position exists, and the intersection is the optical axis adjusting. It may be configured so as to be the reference position for use (claim 4).

That is, what is necessary for adjusting the optical axis is that it can be directly determined that the position of the optical axis can be detected. According to this idea, as shown in claims 2 and 3, the horizontal line and the oblique line are obtained, and the intersection is set as the optical axis adjusting reference position. However, even if the position of the optical axis is not necessarily known, the reference position associated with the optical axis may be known. That is,
This is because if it is found that the reference position always deviates from the optical axis by a certain amount, the deviation can be taken into consideration for adjustment.

Therefore, in the invention of claim 4, the intersection of the vertical line and the horizontal line is set as the reference position. The vertical line may be a fixed value for the sake of simplicity.
It may be specified based on the actually obtained grayscale image data, as defined in.

That is, the vertical line determining means for obtaining the vertical line (in the embodiment, the "vertical line determining section 1
3)), and the vertical line determining means is configured to obtain a line vertical to the horizontal axis including the geometrical center of gravity or the center of gravity of the light and shade in the given grayscale image data. Yes (claim 6).
This is realized by the third embodiment. Further, the line specified based on the density value integration information can be, for example, a coordinate value for obtaining the maximum frequency of the integrated density values (claim 5).

[0022]

FIG. 3 shows a preferred embodiment of a reference position detecting device for adjusting an optical axis of a headlight according to the present invention. As shown in the figure, the image data (grayscale image) to be processed, which is imaged by the image pickup device 1 such as ITV, is temporarily stored in the frame memory 2. The image pickup device may be a device that continuously captures images, such as a video camera, or may be a device that captures images one by one, such as a digital still camera. Any light source can be used as long as the formed light projection pattern can be obtained as grayscale image data.

The image data stored in the frame memory 2 (gray image for one frame: raw data)
Are subjected to expansion processing and contraction processing in the image expansion unit 3 and the image contraction unit 4, respectively, and the obtained expansion image and contraction image are respectively given to the difference processing unit 5, where the same grayscale image data is processed. A difference image between the expanded image and the contracted image is generated and the edge portion is extracted. Then, the difference image generated by the difference processing unit 5 is temporarily stored in the frame memory 6.

The difference image stored in the frame memory 6 is divided into a horizontal line portion and an oblique line portion by the image dividing unit 7, and the horizontal line portion is sent to the horizontal line determining unit 8 where it is sent. Find the horizontal lines that make up the cut line. Further, the diagonal line portion is sent to the diagonal line determination unit 9 and the diagonal line forming the cut line is obtained there.

Further, the outputs of the horizontal line determining section 8 and the oblique line determining section 9 are given to the optical axis position determining section 10, and the optical axis position (intersection of both lines is determined based on the obtained horizontal line and oblique line). Position).

Next, detailed functions of the respective parts will be described with specific examples. The image expansion unit 3 uses a 32 × 32 mask as shown in FIG. 4, and uses the center of the lower side as the reference pixel as shown in the drawing, and the periphery thereof is a mountain-shaped (substantially semi-circular) effective pixel (painted black). The upper pixel is used as a template with invalid pixels (white), and all pixels are processed by a raster method or the like for a given image. Then, specifically, a process of replacing the highest (bright) density value of the reference pixel and the effective pixel with the density value of the reference pixel is performed.

Further, the image contracting section 4 uses the same template (mask) as described above to replace the lowest (dark) density value of the reference pixel and the effective pixel with the density value of the reference pixel. I have to.

That is, for convenience of illustration, a description will be given below using a 5 × 5 mask. That is, as shown in FIG. 5, a triangle (approximately a semicircle) is provided above the reference pixel (*).
It is assumed that various processes are performed on the original image as shown in FIG. 6 by using the mask having the effective pixel (1). As is clear from FIG. 6, the brightest region (8) exists in the central part and the darkest region (0) exists in the upper part. The boundary between the two regions has a horizontal portion and an inclined portion that extends outward and upward from the left side of the horizontal portion.
Further, the decrease from the brightest region (8) to dark becomes relatively sharply dark on the upper side and gently dark on the lower side.

When expansion processing is performed on such an image, FIG.
The darkest area (0) is reduced and the central brightest area (8) is enlarged, as shown in FIG. When the contraction process is performed, the darkest area (0) is enlarged as shown in FIG.
The brightest area in the center (8) is deflated. Also, in the enlarged / contracted images, the darkest direction from the brightest state to the surroundings is the same as the original image even though the decrease start position is different, and the upper side becomes relatively suddenly darker,
The lower side is fading gradually.

Further, the difference processing unit 5 is provided with the expansion / contraction 2
The two image data are compared with each other to obtain the difference in gradation (gradation) of corresponding pixels. When each enlarged / contracted image shown in FIGS. 7 and 8 is given, a difference image as shown in FIG. 9 is obtained. To be In other words, the density of corresponding pixels in the dilated image and the dilated image is almost the same in the originally facing part of the bright headlight (center part) and conversely in the dark region where the light of the headlight is not irradiated (upper part), and the difference between them Is small, it becomes dark in the difference image.

Pixels existing at the boundary between the bright portion and the dark portion are bright in the expanded image and dark in the contracted image, so that the difference in density is large and the difference image is bright. Then, as shown in the original image of FIG. 6, the density change at the upper boundary of the bright part (the cut line position to be originally extracted) is abrupt, so that the density difference between the pixels subjected to the expansion / contraction processing is It becomes larger and brighter in the difference image. On the other hand, since the lower side of the bright portion gradually becomes darker, the density difference between the pixels subjected to the expansion / contraction processing becomes smaller, and the difference image becomes darker. In other words, by expanding and contracting using the same mask and taking the difference between them, the density of the pixels near the cut line becomes brighter and the other parts become darker, so it is possible to extract the pixels near the cut line. It will be possible.

Furthermore, in this embodiment, 32 × is actually used.
A relatively large mask (template) such as 32
Since it is used, it is less likely to be affected by microscopic effects such as flicker.

The image division unit 7 extracts a horizontal component image and a vertical component image from the difference image stored in the frame memory 6, and its specific function is to execute the flowchart shown in FIG. ing.

That is, first, the grayscale image data is read from the frame memory 2 in which the original image is stored, and the division reference position is calculated (ST1). The algorithm for calculating the division reference position may be, for example, the one shown in the flowchart shown in FIG. 11 or the one shown in the flowchart shown in FIG. Further, instead of calculating in this way, it may be a fixed value determined in advance (because the approximate position is determined for each vehicle).

As a concrete example, first, a density histogram of the original image obtained as shown in FIG. 11 is obtained (ST1a). As a result, the number of pixels having each density value is known, and the maximum density value and the minimum density value are also known.

Next, a density value of a specific% (for example, 70%) between the maximum density value and the minimum density value is obtained, and the original image is binarized using that value as a threshold (ST1b). That is,
Based on the histogram obtained in step 1a, the number of pixels corresponding to each density value is sequentially added from the lowest density value,
The density value when the total number reaches a specific% (for example, 70%) of the total number of pixels is set as the threshold value. As a result, for example, if the specific% is 70%, pixels having a density value of 30% remain from the bright side. That is, the originally bright portion is extracted by binarization.

It should be noted that, instead of using the histogram in this way, the maximum density value and the minimum density value are simply acquired and specified based on the two density values (without considering the number of pixels of each density value). The density value of% may be obtained (when the minimum density value is 0% and the maximum density value is 100%, the specific% is simply calculated from the density value ratio). Furthermore, various methods can be adopted, such as binarization processing based on a predetermined specific threshold value.

Then, the geometrical barycentric position of the binarized image (the portion where the light from the headlamp is projected becomes bright) thus obtained is obtained, and the obtained position is set as the division reference position. (ST1c). As a result, FIG.
The process of step 1 shown in FIG.

As another example of the algorithm for calculating the division reference position, a density histogram of the image is acquired based on the acquired original image as shown in FIG. 12 (ST
1a '). Then, an appropriate minimum density threshold value is set (S
T1b '). The minimum density threshold value may be a fixed value that is set in advance, or may be determined based on the density histogram. Then, it is possible to calculate the gray scale barycentric position in the gray scale image data consisting of pixels having the minimum density threshold or more in the original image, and set it as the division reference position (ST1c ').

That is, for example, it is assumed that the density histogram is as shown in FIG. 13 (since there is a dark portion where no light is applied, the frequency is large even in a low density value area). In that case, pixels having a density value less than the minimum density threshold value X are excluded, and only pixels having a large density value (greater than or equal to the minimum density threshold value X) irradiated by the light of the headlamp (area indicated by hatching in FIG. 13). Based on the above, it is possible to accurately specify the vicinity of the center of the portion where the light of the headlamp is illuminated and becomes brighter by obtaining the density center of gravity.

Next, the process proceeds to step 2 shown in FIG. In this processing step, the difference image stored in the frame memory 6 is acquired, and a position offset from the division reference position obtained in step 1 by a predetermined distance (in the case of FIG. 9,
Offset to the right). By offsetting in this way, the extracted image can mainly extract horizontal components including horizontal lines. Then, by appropriately setting the offset amount, it is possible to extract an image that includes horizontal lines and does not include diagonal lines. Therefore, when determining a line position to be described later, normal processing is performed without being affected by diagonal lines. can do. Then, the horizontal component image thus extracted is given to the horizontal line determining unit 8 in the next stage.

Next, in step 3, the vertical component mainly including the diagonal line is extracted. That is, basically the same processing as the above-described extraction of the horizontal component image is performed. However, the direction of offset is the opposite side. That is, the vertical component image obtained by extracting the image outside the position (offset to the left in the case of FIG. 9) offset from the division reference position obtained in step 1 by the diagonal line determination unit in the next stage. Give to 9. Then, by appropriately setting the offset amount, the vertical component image includes diagonal lines and does not include horizontal lines.

The offset amounts from the division reference position when extracting the horizontal component and the vertical component may be the same (absolute value) or may be different. Further, one or both may be cut out in the respective directions from the offset amount = 0, that is, the division reference position, and further, ±
Any value can be taken. That is, the horizontal and vertical components may overlap in some areas.

The horizontal line determining section 8 determines the horizontal line position which is the horizontal portion of the cut line based on the horizontal component image received from the image dividing section 7. Specifically, the flow chart shown in FIG. 14 is realized. It has a function to do.

That is, first, a given image (a grayscale image of the horizontal component in the difference image) is integrated and projected with the density value on the Y axis (vertical axis) (ST11). That is, the total sum of the densities of the pixels located in the same column is obtained. As an example, as a result of the image division unit 7 performing division processing on the difference image shown in FIG.
Assuming that a region surrounded by a broken line in FIG. 15 is extracted as a horizontal component image, the integrated values shown on the respective sides are obtained as a result of the density value integrated projection.

Next, the projection histogram mode value is detected (ST12). That is, from the result of cumulative projection,
For example, a histogram as shown in FIG. 16 is obtained, and its peak is obtained. For simplicity, the maximum integrated value (frequency) in each column is found and that column (Y coordinate value) is acquired. Then, the line parallel to the X axis passing through the Y coordinate value of the peak thus obtained is determined as the horizontal line L1. In this way, using the data projected by integrating the density value,
It is possible to uniquely obtain the peak value from the entire histogram, and it is possible to execute it by a relatively simple process.

On the other hand, the diagonal line determination unit 9 determines the diagonal line position of the cut lines based on the vertical component image received from the image division unit 7. Specifically, the flowchart shown in FIG. 17 is realized. It has a function to do.

That is, first, the given image (a grayscale image of the vertical component in the difference image) is rotated by a predetermined angle θ (ST21). This angle corresponds to the tilt angle of the diagonal line. Since the tilt angle θ is almost constant, it can be set in advance as a fixed value, and is set to 15 degrees in this embodiment.

Thus, for example, when a given image has a bright region R extending in an oblique direction as shown in FIG. 18, by rotating by an angle θ, a bright region R'as shown in FIG. Is transformed to extend horizontally. It should be noted that the center of rotation at the time of this rotation processing can take an arbitrary position, and can be, for example, the center of the image.

When the image-converted area R'extends in the horizontal direction in this way, the line position L based on the area R'is obtained by the same processing as the horizontal line determination section.
You can know 2 '. And the line position L
The diagonal line L2 to be obtained can be acquired by rotating 2'in the opposite direction.

That is, the density value is integrated and projected onto the Y axis (vertical axis) (ST22). That is, the total sum of the densities of the pixels located in the same column is obtained. Next, the projection histogram mode value is detected (ST23). That is, the vertical axis coordinate position of the maximum frequency position is obtained. Then, the diagonal line position L2 is obtained based on the vertical axis coordinate position.

That is, L2 'is defined by Y = b. Here, b is the vertical axis coordinate position obtained in step 23. Then, the coordinates (X1, Y1) are converted to (XC, YC)
As a result of rotating by θ degrees around the center, the coordinate value after rotation is (X
2, Y2), the following formula is established.

[0053]

[Equation 1]

Therefore, the straight line L1 defined by Y = b is
When rotated by θ degrees with reference to (XC, YC), the straight line L2 after rotation is as follows.

[0055]

[Equation 2]

As shown in FIG. 20, since the straight lines L1 and L2 respectively obtained as described above constitute a part of the cut line, the actual cut line is the straight lines L1 and L2.
2 is extended and connected. Then, the intersection point P becomes the reference position for adjusting the optical axis. The actual optical axis is a line connecting the intersection point P and the headlamp, but the intersection point position P is necessary for adjustment.

Therefore, the intersection (optical axis position) is calculated by the optical axis position determining unit 10. Specifically, the Y coordinate value of the intersection is the Y coordinate value (a) of the horizontal line L1, and the X coordinate value is obtained by substituting the Y coordinate value (a) into Equation (1) above.
Will be solved. In other words, it becomes like the following formula. Then, the optical axis position determination unit 10 performs the calculation processing based on the data respectively supplied from the horizontal line determination unit 8 and the oblique line determination unit 9.

[0058]

[Equation 3]

Then, by performing the above respective processes on the grayscale image as shown in FIG. 1, the adjustment reference position P is obtained as shown in FIG. In the figure, reference characters L1 and L
2 are horizontal lines and diagonal lines, respectively, and L3
Is a perpendicular line to the X axis including the division reference position.

Although not shown in the drawing, this device is equipped with a monitor device so that at least the reference point determined by the optical axis position determination unit can be displayed. And as an example of the display form, it is as shown in FIG. Of course, the monitor is not limited to the display means, and a printing device such as a printout may be provided. Furthermore, various output means can be connected, such as determining whether the obtained reference position matches the regular position and suggesting the adjustment amount.

FIG. 22 shows a second embodiment of the present invention. This embodiment differs from the above-described first embodiment in that an oblique line is obtained. That is, the first
In the embodiment, if the image dividing unit 7 obtains the division reference position, the horizontal component image and the vertical component image are obtained based on the obtained reference positions. However, in the present embodiment, only the horizontal component image is obtained and the vertical component image is obtained. Images are not extracted.

The diagonal line determination unit 9 determines the diagonal line based on the entire difference image stored in the frame memory 6. That is, as shown in FIG. 23, when the difference image is taken, a region R1 near the diagonal line is obtained.
Becomes relatively bright, but the region R2 near the vertical line becomes dark.

Therefore, when pixels on the diagonal line side are included in the determination of the horizontal line, it is erroneously recognized as a position moved above the true horizontal line due to the influence of the diagonal line side with bright density. To identify diagonal lines,
The presence of pixels on the horizontal line side does not have much effect. Therefore, in the present embodiment, the processing of FIG. 17 is performed for diagonal lines based on the entire difference image. Since the other configurations and effects are similar to those of the above-described first embodiment, detailed description thereof will be omitted.

FIG. 24 shows a third embodiment of the present invention. In the present embodiment, unlike the above-described respective embodiments, a vertical line determination unit 13 is provided instead of the diagonal line determination unit. In other words, what is needed to adjust the optical axis is not a specific diagonal line or horizontal line, but
It is a reference position for adjusting the optical axis. Since the adjustment reference position is the intersection of the horizontal line and the diagonal line, it always exists on the horizontal line (including the extension line).

Therefore, in the present embodiment, instead of the diagonal line, a vertical line for specifying the X coordinate at which the adjustment reference position exists is obtained. The function of the vertical line determination unit 13 that obtains the vertical line is the same as the function that obtains the division reference position in the image division unit 7. That is, FIG.
1 and the processing shown in FIG. 12 are executed to obtain the vertical line reference position (equivalent to the division reference position) based on the grayscale image stored in the frame memory 2. Then, a vertical line that includes the vertical line reference position and is perpendicular to the X axis is obtained, and it functions as a vertical line. As an example, the straight line indicated by reference numeral L3 in FIG. 21 is a vertical line.

The optical axis position determining unit 10 determines the intersection of the vertical line and the horizontal line (including the extension line) determined by the horizontal line determining unit 8 and sets it as the optical axis adjusting reference position. It has become. Then, as is apparent from FIG. 21, the optical axis adjusting reference position P ′ obtained in the present embodiment is closer to the X axis direction than the adjusting reference position P obtained in the first embodiment. It is shifted and not in the same position. However, when the adjustment of the optical axis and has proven how much shift in advance, there is no problem because it can be adjusted accordingly.

Actually, in the processing for obtaining the above-mentioned intersection, Y = a for the horizontal line and X = c for the vertical line.
Therefore, the coordinate values (a and c) are output from the line determining units 8 and 13, respectively, and the optical axis position determining unit 10 calculates the coordinate value (c, a) of the adjustment reference position from the coordinate values. Will be decided.

Further, although not shown, the vertical line is not based on the original image stored in the frame memory 2, but the X coordinate value is set in advance as a fixed value, and the obtained horizontal line is used. The reference position for adjustment may be obtained.

[0069]

As described above, in the optical position adjusting reference position detecting device for a headlight according to the present invention, a horizontal line or an oblique line is specified based on the density value integration information obtained by projecting on the vertical axis. Therefore, it is possible to easily and reliably specify the cut line without being affected by the flickering of the reflection plate and the fluctuation of light (vertical movement of the light-dark boundary position when viewed microscopically). As a result, the reference position for adjusting the optical axis of the headlight (low beam) can be specified.

[Brief description of drawings]

FIG. 1 is a diagram showing an example of a light projection pattern.

FIG. 2 is a diagram illustrating a conventional example.

FIG. 3 is a diagram showing a first embodiment of a reference position detection device for adjusting an optical axis of a headlight according to the present invention.

FIG. 4 is a diagram showing a mask pattern used when performing expansion / contraction processing.

FIG. 5 is a diagram showing a simple mask example for explaining expansion / contraction / difference processing.

FIG. 6 is a diagram showing an example of a grayscale image to be processed.

FIG. 7 is a diagram showing a result of expanding the image of FIG.

FIG. 8 is a diagram showing a result of contraction processing of the image of FIG.

9 is a diagram showing a result of difference processing of the images of FIGS. 7 and 8. FIG.

FIG. 10 is a flowchart illustrating a function of an image division unit.

FIG. 11 is a flowchart showing an example of a division reference position calculation algorithm.

FIG. 12 is a flowchart showing another example of the division reference position calculation algorithm.

FIG. 13 is a diagram for explaining the operation of another example shown in FIG.

FIG. 14 is a flowchart illustrating a function of a horizontal line determination unit.

FIG. 15 is a diagram illustrating an operation of a horizontal line determination unit.

FIG. 16 is a diagram illustrating an operation of a horizontal line determination unit.

FIG. 17 is a flowchart illustrating the function of a diagonal line determination unit.

FIG. 18 is a diagram illustrating the operation of the diagonal line determination unit.

FIG. 19 is a diagram illustrating the operation of the diagonal line determination unit.

FIG. 20 is a diagram illustrating the function of an optical axis position determination unit.

FIG. 21 is a diagram showing a processing result.

FIG. 22 is a diagram showing a second embodiment of a reference position detection device for adjusting an optical axis of a headlight according to the present invention.

FIG. 23 is a diagram illustrating a function of a diagonal line determination unit.

FIG. 24 is a diagram showing a third embodiment of an optical axis adjustment reference position detection device for a headlight according to the present invention.

[Explanation of symbols]

1 ITV (imaging device) 2 frame memory 3 Image expansion part 4 Image contraction section 5 Difference processing unit 6 frame memory 7 Image division 8 Horizontal line determination section 9 Diagonal line determination unit 10 Optical axis position determination unit 13 Vertical line determination unit

─────────────────────────────────────────────────── --- Continuation of the front page (56) References JP 10-300633 (JP, A) JP 3-103743 (JP, A) JP 1-131429 (JP, A) JP 8- 101092 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) G01M 11/00-11/08 G01N 21/84-21/958

Claims (6)

(57) [Claims] 1. An expansion processing unit that expands the grayscale image data of a given projection pattern, a shrinkage processing unit that shrinks the grayscale image data, the expansion processing unit and the shrinkage. Difference processing means for obtaining a difference image between the expanded image and the contracted image respectively generated by the processing section; and means for extracting a horizontal component area including a horizontal line forming a light-dark boundary line in the light projection pattern from the difference image. A horizontal line determining unit that specifies the horizontal line based on density value integration information obtained by projecting in the vertical axis direction with respect to the extracted horizontal component area; and a horizontal line determined by the horizontal line determining unit. A detection means for obtaining a reference position for adjusting the optical axis based on a line, wherein the expansion processing means and the contraction processing means are subject to microscopic effects such as flicker. A reference position detecting device for adjusting an optical axis of a headlight, characterized in that the expansion process or the contraction process is performed using a relatively large mask having no light. 2. A rotation image obtained by rotating the difference image by a predetermined angle is generated, and the reference line is obtained based on density value integration information obtained by projecting the rotation image in the vertical axis direction. Further comprising a diagonal line determining means for identifying the diagonal line by performing an inverse transformation of the generation of the rotation image to the reference line, the detecting means, obtain the intersection of the horizontal line and the diagonal line, the light The optical axis adjusting reference position detecting device for a headlight according to claim 1, wherein the optical axis adjusting reference position is used as an axis adjusting reference position. 3. A means for extracting a vertical component area including an oblique line forming a light-dark boundary line in the light projection pattern in the difference image, and rotating the extracted vertical component area by a predetermined angle. By generating the rotated image, the reference line is obtained based on the density value integration information obtained by projecting the rotated image in the vertical axis direction, and the reference line is subjected to inverse conversion of the generation of the rotated image. The oblique line determining means for specifying the oblique line is further provided, and the detecting means obtains an intersection of the horizontal line and the oblique line, and sets the intersection as the optical axis adjusting reference position. Position detection device for adjusting the optical axis of the headlight. 4. The detecting means obtains an intersection of the horizontal line and a vertical line where the optical axis adjusting reference position exists, and sets the intersection as the optical axis adjusting reference position. The reference position detection device for adjusting the optical axis of the headlight according to. 5. The headlight according to claim 1, wherein the line specified based on the density value integration information is a coordinate value that obtains the maximum frequency of the integrated density value. Reference position detection device for optical axis adjustment. 6. A vertical line determining means for determining the vertical line is further provided, wherein the vertical line determining means is perpendicular to a horizontal axis including a geometrical center of gravity or a grayscale center of gravity of a bright portion in the given grayscale image data. 5. The reference position detecting device for adjusting the optical axis of a headlight according to claim 4, which is for obtaining a different line.