JPH1151644A - Distance measuring instrument for vehicle - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a distance measuring instrument for vehicle which measures the distance to an object by electrically detecting the deviations of both images from the optical axis of the instrument by comparing a pair of image signals formed on image sensors through a pair of optical systems and executing operation, based on the principle of the triangulation, by using the deviations and can discriminate the moving direction of the object at an increased computing speed by reducing the amount of computation. SOLUTION: A window setting means 8 sets the windows of whole image signals respectively formed on image sensors 3A and 3B at a plurality of locations scattered in the horizontal direction and an arithmetic means 10 computes the correlation between the image signals and the distance to an object 4 at every window. Then a distance calculating means 13 calculates the distance to the object 4, based on the distances computed at the windows having a correlation between them, and a moving direction discriminating means 15 discriminates the moving direction of the object 4, based on the horizontal distribution of the correlative windows.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a distance measuring apparatus for a vehicle, and more particularly to a method for comparing a pair of image signals formed on an image sensor by a pair of optical systems to determine a shift amount of both images from an optical axis. The present invention relates to a vehicle distance measuring device that electrically detects and executes a calculation based on the principle of triangulation using the deviation amount to measure a distance to an object.

[0002]

2. Description of the Related Art Conventionally, such a distance measuring device is already known, for example, from Japanese Patent Application Laid-Open No. 4-50611. In this device, a window is set for one of the image signals imaged on both image sensors, and the distance to the object is calculated by comparing the two image signals based on the set window and calculating the window. The symmetry in the temporary window determined as the part of the image signal that matches the most is evaluated by comparing the image signals within the window at a certain time and the next time, and the temporary window that is evaluated as having the best symmetry is determined as the window. age,
The moving direction and the moving amount of the object are obtained based on the change amount of the window from the previous time.

[0003]

However, in the above-mentioned conventional device, since it is necessary to evaluate the symmetry of the temporary window, a relatively expensive two-dimensional CCD in which many pixels are arranged without gaps is used as an image sensor. It needs to be used, which increases the amount of arithmetic processing and reduces the processing speed.

The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a vehicle distance measuring device capable of judging the moving direction of an object while reducing the amount of arithmetic processing. I do.

[0005]

According to a first aspect of the present invention, there is provided an optical system comprising: a pair of optical systems for comparing a pair of image signals formed on an image sensor by an optical axis; In the vehicle distance measurement device that electrically detects the amount of deviation from, and performs a calculation based on the principle of triangulation using the amount of deviation to measure the distance to the object, Window setting means for setting the windows of the formed entire image signal at a plurality of locations dispersed in the horizontal direction; and calculating means for performing a correlation calculation and a distance calculation of the image signal for each window set by the window setting means. Distance calculating means for calculating the distance to the object based on the distance of the window whose correlation has been obtained by the calculation means; and a window having a correlation obtained by the calculation means. Characterized in that it comprises a moving direction determination means for determining a moving direction of the object based on the lateral distribution.

According to such a configuration, when the moving direction of the object changes in a direction different from the moving direction of the own vehicle, the correlated window is biased to the left or right, and In the direction determining means, it is possible to determine the moving direction of the object based on the horizontal distribution of the window from which the correlation has been obtained. In addition, in the window setting means, windows are set at a plurality of locations distributed in the horizontal direction, and it is only necessary to execute the calculation in a range narrower than the entire image area of the image sensor, so that the amount of calculation processing is small, Processing speed increases.

According to a second aspect of the present invention, in addition to the configuration of the first aspect, the window setting means sets the current window based on a previous determination result of the moving direction determination means. By setting, in the current distance measurement to the target object, measurement following the movement of the target object can be efficiently performed.

[0008]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The embodiments of the present invention will be described below based on one embodiment of the present invention shown in the accompanying drawings.

1 to 7 show an embodiment of the present invention. FIG. 1 is a block diagram showing a configuration of a distance measuring device, FIG. 2 is a diagram showing an arrangement of windows with respect to an image area,
FIG. 3 is a diagram for explaining the differential waveform shaping by the calculation means;
FIG. 4 is a diagram for explaining the correlation calculation by the calculation means, and FIG.
FIG. 6 is a diagram for explaining the principle of distance measurement, FIG. 6 is a diagram for explaining the correlation evaluation by the calculating means, and FIG. 7 is a diagram for explaining the interpolation calculation by the calculating means.

First, referring to FIG.
A and 1B are arranged, for example, on the rear side of a windshield in a vehicle cabin of the vehicle, and these imaging means 1A and 1B are provided.
Is composed of optical systems 2A and 2B including lenses, and image sensors 3A and 3B disposed rearward by a focal length f of the optical systems 2A and 2B, respectively.
The two optical systems 2A and 2B are arranged at an interval above and below the base line length BL.

According to such imaging means 1A and 1B,
An object 4 such as a preceding vehicle existing in front of the own vehicle is an optical system 2.
A and 2B form an image on the image sensors 3A and 3B. As shown in FIG. 2, the image sensors 3A and 3B have a plurality of, for example, line CCDs 5, 5. A one-dimensional image sensor is arranged in a horizontal direction, and the viewing angle and the resolution in the horizontal direction can be optimally determined by appropriately setting the intervals between the line CCDs 5, 5,.

Image signals obtained by the two image sensors 3A and 3B are converted into digital signals by individual A / D converters 6A and 6B, and stored in individual image memories 7A and 7B, respectively.

Each of the image memories 7A and 7B has a line CCD.
The image signals obtained by 5, 5,... Are temporarily stored, but only the image signal selected by the window setting means 8 among the image signals is cut out from the image memories 7A, 7B. . In the window setting means 8,
As shown by hatching in FIG. 2, a plurality of windows W, W... Spaced at a horizontal direction are set, and a plurality of image sensors 3A and 3B are arranged at equal intervals in the horizontal direction. Of the line CCDs 5, 5,... Are selected as the image signals of the windows W, W,. Will be.

In addition, the window setting means 8 initially sets the image sensors 3A and 3B as shown in FIG.
Are set so as to be symmetrical left and right with respect to the center in the horizontal direction. After the initial setting, the positions of the windows W, W are determined based on the signal from the window memory 9. .. Are set by the window setting means 8.

The image signals of the windows W, W... Set by the window setting means 8 are input to the calculating means 10, and the calculating means 10 correlates the image signals of the respective combinations of the windows W, W corresponding to each other. The calculation and the distance calculation are executed.

In performing the correlation calculation, the calculation means 10 is used to avoid a decrease in the degree of correlation coincidence due to poor image contrast at dusk or in a shade or the like, and hence a decrease in distance measurement accuracy. Then, as shown in FIG. 3, the luminance of the image signal from the window W is read in to form a luminance waveform, and then the luminance of a plurality of pixels, for example, every nine pixels is subtracted vertically to form a differential waveform.

In the correlation operation by the arithmetic means 10, subtraction is performed while fixing one of the differentiated luminance data of the window W corresponding to each other in the two image sensors 3A and 3B and shifting the other one by one. It is executed as follows.
For example, as shown in FIG. 4A, the luminance data on the upper side, that is, the image sensor 3B side is fixed, and the luminance data on the lower side, that is, the image sensor 3A side is shifted up and down one by one. By subtracting from the data, a shift amount corresponding to the correlation value is obtained as shown in FIG. 4B, and the shift amount when the correlation value is the smallest corresponds to the shift amount of the image signal from the optical axis. It is determined as the corresponding value.

The correlation calculation for obtaining the shift amount corresponding to the correlation value is for obtaining the shift amount from the optical axis of the image formed on both image sensors 3A and 3B. By shifting both luminance data until the correlation between the luminance data obtained from the lower image sensor 3A and the luminance data obtained from the upper image sensor 3B becomes the best, the point at the point where the correlation is the best is obtained. The shift amount n can be obtained. The shift amount n is obtained from the shift amount nA of the image signal obtained from the lower image sensor 3A from the optical axis of the optical system 2A and the shift amount nA from the upper image sensor 3B. The obtained image signal is obtained as the sum with the shift amount nB from the optical axis of the optical system 2B.

By the way, the shift amount corresponding to the correlation value is obtained by the above-described correlation operation. In order to evaluate the reliability of the correlation value, the arithmetic means 10 executes the following correlation evaluation. . That is, as shown in FIG. 6, a threshold value larger than the smallest correlation value by a predetermined value is set, the minimum value of the correlation value is one, and the threshold value is shifted by a predetermined shift amount ΔS from the shift amount of the minimum correlation value. When the correlation value is larger than the threshold value at the position, that is, when the state is as shown in FIG. 6A, the reliability is evaluated to be high. On the other hand, when the correlation value is smaller than the threshold value at a position shifted by the predetermined shift amount ΔS from the shift amount of the minimum correlation value even if the correlation value has one minimum value, that is, as shown in FIG. In the state, the reliability is evaluated to be low. When there are a plurality of minimum values of the correlation value, that is, in the state shown in FIG. 6C, the reliability is evaluated to be low. FIG. 6 (B)
The state indicated by is caused by the fact that the slope of the correlation value becomes not sharp near the minimum correlation value based on the luminance shift or noise of the image signals of the two image sensors 3A and 3B, and the degree of correlation coincidence decreases. Since the distance measurement accuracy is reduced, the correlation evaluation is reduced. The state shown in FIG. 6C is caused by a pattern image typified by a mixed perspective or a striped pattern. Since there is a possibility that other distances may be measured, the correlation evaluation value is reduced.

Further, in the arithmetic means 10, since the pixels of the CCDs 5, 5,... In each of the image sensors 3A, 3B have a certain size, interpolation between the pixels is performed, and the image sensors 3A, 3B perform the interpolation. An interpolation operation is performed to improve the resolution capability of. In this interpolation calculation, as shown in FIG. 7, the first point P1 which is the minimum correlation value
And the second and third points P which are correlation values close to the minimum correlation value.
2 and P3, a straight line L1 passing through the first point P1 and one of the second and third points P2 and P3, and a second and third point P2 having the same inclination as the straight line L1. , P3 at the intersection with the straight line L2 passing through the other P2 is obtained as an interpolation value, and this interpolation value is used as the shift amount n '. In addition, as shown by a chain line, the first point P1 and the second point P1
Although an intersection of a straight line passing through No. 2 and a straight line passing through the third point P3 is also obtained, this intersection is not adopted as an interpolation value because it has a correlation value larger than the minimum correlation value.

After the interpolated shift amount n 'is thus obtained, the calculating means 10 performs a distance calculation based on the principle of triangulation. That is, the distance D to the object 4 is obtained as D = (BL × f) / n ′.

Each window W, W obtained by the arithmetic means 10
Is stored in the distance memory 11, and the correlation evaluation value for each combination of the windows W, W is stored in the correlation evaluation memory 12.

Distance memory 11 and correlation evaluation memory 12
The distance to the object 4 is calculated by the distance calculation unit 13 based on the information stored in the storage unit, and the calculated distance is output from the distance calculation unit 13.

In the distance calculating means 13, the combination of the windows W, W having a low correlation evaluation is excluded, and among the combinations of the remaining windows W, W,
Combinations that have obtained a distance at least a certain distance from the previously obtained distance are excluded, and the last remaining combination of windows W, W is adopted, and the distances obtained by the adopted combinations, respectively. Is calculated as the distance to the object 4.

The windows W and the combination of W adopted in the distance calculation by the distance calculating means 13 are stored in the adopted window memory 14. The combination of the windows W and W stored in the adopted window memory 14 represents the horizontal distribution of the window W obtained by the calculation by the calculation means 10 and based on the distribution, the object 4 Is determined by the moving direction determining means 15. That is, when the moving direction of the target object 4 changes in a direction different from the moving direction of the own vehicle, the correlation evaluation value of the window W looking at the background other than the target object 4 becomes lower, or the distance obtained last time. By obtaining a distance far from the window W, the correlated window W is biased to the left or right, and the movement of the object 4 is determined based on the horizontal distribution of the correlated window W. The direction can be determined by the moving direction determination unit 15.

The result of determination by the moving direction determining means 15 is stored in the window memory 9, and the window setting means 8 stores the result of the moving direction determination stored in the window memory 9, .. Are set to be shifted in the moving direction of the object 4 based on the moving direction determination result.

Next, the operation of this embodiment will be described. When the moving direction of the object 4 changes to a direction different from the moving direction of the own vehicle, the moving direction of the object 4 is judged by the moving direction judging means 15. The moving direction judgment by the moving direction judging means 15 determines whether the left or right of the windows W, W for which the correlation is obtained among the windows W, W. The image sensor 3A, 3B
It is sufficient to execute the calculation in a range narrower than the entire image area, and the amount of calculation processing can be reduced. Moreover, since the windows W are set at intervals in the horizontal direction, the image sensors 3A, 3B are relatively composed of line CCDs 5, 5... Arranged at equal intervals in the horizontal direction. An inexpensive one-dimensional image sensor can be used, and cost can be reduced.

The window setting means 8 sets the current window W based on the result of the previous determination by the moving direction determining means 15, so that the current distance measurement to the object 4 can be performed.
The movement can be performed efficiently following the movement of the object 4.

In the above embodiment, the image sensors 3A, 3A
A one-dimensional image sensor composed of a plurality of line CCDs 5, 5,. It is also possible.

Although the embodiments of the present invention have been described in detail, the present invention is not limited to the above embodiments, and various design changes can be made without departing from the present invention described in the appended claims. It is possible to do.

[0031]

As described above, according to the first aspect of the present invention, when the moving direction of the object changes to a direction different from the moving direction of the own vehicle, the window from which the correlation is obtained is either left or right. It is possible to determine the moving direction of the target object based on the bias toward, the window is set at a plurality of locations dispersed in the horizontal direction, the window is narrower than the entire image area of the image sensor In this case, the amount of calculation processing can be reduced, and the calculation processing speed can be increased.

According to the second aspect of the present invention, the distance measurement following the movement of the object can be efficiently performed.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of a distance measuring device.

FIG. 2 is a diagram showing an arrangement of windows with respect to an image area.

FIG. 3 is a diagram for explaining differential waveform shaping by a calculation unit;

FIG. 4 is a diagram for explaining a correlation calculation by a calculation unit.

FIG. 5 is a diagram for explaining the principle of distance measurement.

FIG. 6 is a diagram for explaining the correlation evaluation by the calculation means.

FIG. 7 is a diagram for explaining an interpolation calculation by a calculation unit.

[Explanation of symbols]

 2A, 2B ... optical system 3A, 3B ... image sensor 4 ... object 8 ... window setting means 10 ... calculating means 13 ... distance calculating means 15 ... moving direction determining means W ・ ・ ・ Wind

 ──────────────────────────────────────────────────の Continued on the front page (72) Inventor Kazuto Chisaka 1-4-1 Chuo, Wako-shi, Saitama

Claims (2)

[Claims] 1. A pair of optical systems (2A, 2B) compares a pair of image signals formed on an image sensor (3A, 3B) and electrically detects a shift amount of both images from an optical axis. A distance measuring apparatus for measuring the distance to the object (4) by executing a calculation based on the principle of triangulation using the deviation amount, the two image sensors (3A, 3A);
B) Window setting means (8) for setting the window (W) of the entire image signal formed on each of the plurality of locations dispersed in the horizontal direction, and each window (W) set by the window setting means (8). ) For performing a correlation operation and a distance operation of the image signal for each image signal;
Distance calculation means (13) for calculating the distance to the object (4) based on the distance of the window (W) for which the correlation has been obtained by the calculation in (0), and the correlation by calculation in the calculation means (10). A moving direction determining unit (15) for determining a moving direction of the object (4) based on the distribution of the window (W) obtained in the horizontal direction. 2. The vehicle according to claim 1, wherein said window setting means (8) sets a current window (W) based on a previous judgment result of said moving direction judgment means (15). Distance measuring device.