JPH0961157A - Distance-measuring device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a distance-measuring device in which the amount of data is small because data on only a distance-measuring point is used, in which a processing operation is performed with a simple configuration and in which the fear of a malfunction can be reduced. SOLUTION: A phase difference is computed on the basis of image signals obtained by one pair of sensors arranged so as to correspond to a plurality of distance-measuring points Ps (m, n) which are set to be a lattice shape by a plurality of lines (m's) arranged in the horizontal direction and by a plurality of windows (n's) arranged in the vertical direction, and measured distances in the distance-measuring points Ps (m, n) are found. The absolute value of the difference in the measured distances between the distance-measuring points Ps (m, n) situated in respective vertexes of detection areas As (m, n) composed of two-dimensional regions partitioned by at least three of the distance measuring points Ps (m, n) is computed. When all of computed values are smaller than a reference value, the detection areas As (m, n) are extracted as a box in which an object 20 exists. By using the measured distances between the distance-measuring points Ps (m, n) constituting the box, a distance up to the object 20 is found.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a device for measuring a distance, in particular, an object located in front of an own vehicle such as a preceding vehicle.

[0002]

2. Description of the Related Art Development of a technique for improving safety when a vehicle is traveling is in progress. For example, in order to avoid a rear-end collision accident, the distance to an object that obstructs traveling is measured to determine a predetermined distance. There is a technique (for example, Japanese Examined Patent Publication No. 3-44005) that causes the driver to be warned when the vehicle is closer than the distance and automatically controls the traveling speed.

[0003]

As a device for measuring the distance mounted on such a device, a pair of left and right cameras equipped with an image sensor such as a charge coupled device (CCD) are provided in front of the vehicle. Of distance by a so-called stereo method based on the principle of triangulation for calculating the displacement of the position of an object on the image of a camera of the same (for example, Japanese Patent Laid-Open No. Hei 7
However, there is a problem that a processing method is complicated because of a large amount of data, and an arithmetic unit for processing this requires a high-performance and expensive one. .

There is also a technique (for example, Japanese Unexamined Patent Publication No. 7-71916) for determining that an object exists when the difference between the measured distances to vertically adjacent objects is smaller than a threshold value. There is a problem that it is easy to make a malfunction such as issuing an alarm when an object having a small width but a long length such as a line is determined as an object. The present invention was conceived focusing on such a problem,
An object of the present invention is to provide a distance measuring device having a small amount of data, a simple processing method, and less fear of malfunction.

[0005]

A distance measuring device for solving the above-mentioned problems includes a plurality of horizontally arranged lines and a plurality of vertically arranged windows arranged in a grid pattern. A two-dimensional area defined by at least three distance measuring points by calculating a phase difference from image signals obtained by a pair of sensors arranged corresponding to the distance measuring points to obtain a measured distance at the distance measuring points. An absolute value of the difference in the measured distances between the distance measuring points located at the respective vertices of the detected area, and when all of the calculated values are smaller than a reference value, an object exists in the detected area. It is extracted as a box, and the distance to the object is obtained by using the measurement distance of the distance measurement points forming the box.

The reference value is an absolute value which is a constant value, and a relative value obtained by multiplying a measured distance of one of the distance measuring points forming the box by a constant ratio. It is a value obtained by adding and.

Further, the detection area includes a two-dimensional area in which a projection shape defined by three of the distance measuring points has a triangular shape, and the absolute difference between the distance measuring points between the distance measuring points located at each apex is absolute. A value is calculated, and when all of the three calculated values obtained thereby are smaller than the reference value, the detected area is extracted as the box.

The detection area includes a two-dimensional area having a quadrangular projected shape defined by the four distance measuring points, and the absolute difference between the distance measuring points located at each apex is absolute. A value is calculated, and when all of the six calculated values obtained thereby are smaller than the reference value, the detected area is extracted as the box.

[0009]

BEST MODE FOR CARRYING OUT THE INVENTION Images obtained by a pair of sensors arranged corresponding to a plurality of distance measuring points set in a grid pattern of a plurality of horizontal lines and a plurality of vertical windows The phase difference is calculated from the signal to obtain the measurement distance at the distance measuring points, and the distance between the distance measuring points located at each apex of the detection area formed by the two-dimensional area defined by at least three of the distance measuring points is calculated. The absolute value of the difference between the measured distances is calculated, and when all of the calculated values are smaller than the reference value, the detection area is extracted as a box in which an object exists, and the measurement of the distance measuring points that constitute this box is performed. By calculating the distance to the object using the distance, the measurement data of only the distance measuring points are used, so the data amount is small, the processing method is simple, and It is possible to reduce the worry of work.

The reference value is an absolute value which is a constant value, and a relative value obtained by multiplying a measured distance of one of the distance measuring points forming the box by a constant ratio. By adding a value, the reference value can be set according to the measurement distance to be determined, and compared to the case where the set value is a fixed value, the determination at a long distance measurement distance can be made more accurately. It can be carried out.

The detection area includes a two-dimensional area in which a projection shape defined by three of the distance measuring points has a triangular shape, and the absolute difference of the measuring distances between the distance measuring points located at each apex is absolute. A value is calculated, and when all of the three calculated values obtained by this are smaller than the reference value, the detected area is extracted as the box, or the detected area is four of the distance measuring points. The projection shape defined by is composed of a quadrangular two-dimensional area, and the absolute value of the difference in the measurement distances between the distance measurement points located at each vertex is calculated. By extracting the detection area as the box when it is smaller than the reference value, necessary and sufficient measurement data can be obtained, and the distance to the object can be easily measured. Kill.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below with reference to the embodiments described in the accompanying drawings on the assumption that the present invention is applied as a device for measuring a distance to an object located in front of a vehicle.

FIG. 1 is a block diagram showing the configuration of the embodiment, and 10 is a plurality of distance measuring points P set in a grid pattern.
Distance measurement unit (camera) for obtaining an image signal including an object (vehicle) 20 located in front of the own vehicle corresponding to (m, n)
And a pair of left and right lenses 11L and 11R as sensors
Image sensors 12L and 12R composed of CCDs having a vertical direction that coincides with the base line direction of the image sensors 12L and 12R are arranged.
The signals are amplified by L and 13R, converted into digital converted signals by sampling at a predetermined period by AD converters 14L and 14R, and stored in memories 15L and 15R, respectively. The image signals stored in the memories 15L and 15R are distance measuring units. It is output to the microcomputer 30 connected to the subsequent stage of 10, and calculation for calculating the distance to the vehicle 20 is performed.

FIG. 2 is a distance measuring point P measured by the image sensors 12L and 12R of the distance measuring unit 10 shown in FIG.
It is a figure explaining the arrangement of (m, n), and in this embodiment,
There are 16 lines arranged in the horizontal direction and 7 lines arranged in the vertical direction, for a total of 112 distance measuring points P (1,
1) to P (16, 7) are set, and these distance measuring points P (1, 1) to P (16, 7) are grid-like regularly arranged in lines and windows, and the distance measuring points are set. P (m,
The angle between the lines n) to P (m + 1, n) is 0.56 degrees (“K” in step 309 of FIG. 6), and the distance measuring point P
The angle between windows of (m, n) to P (m, n + 1) is 0.
It is 97 degrees. The distance measuring point P (m, n)
The "m" is the order from the left side of the line, the "n" is the order from the lower side of the window, and the position of m = 8.5 is set to be the center of the vehicle. Image sensor 12
L and 12R are scanned by the drive pulse signal and sequentially read for each distance measuring point P (m, n). For example, the distance measuring points P (1,1) to P (16,1) where n = 1.
Are read one by one from the left end of the line, and distance measuring points P (1,2) to P (16,2) for n = 2 and distance measuring points P (1,3) to P (16 for n = 3. , 3), when the final distance measuring points P (1, 7) to P (16, 7) of n = 7 are completed, this series of scanning is repeated in a predetermined cycle.

FIG. 3 is a flow chart for explaining the processing in the microcomputer 30, and the image sensor 12 of FIG.
The image signals of the distance measuring points P (m, n) obtained from L and 12R are output to the microcomputer 30 and the measuring distances D (m of the distance measuring points P (m, n) based on the principle of triangulation. , N) is measured (step 100 in FIG. 3).

Next, as shown in the enlarged view of FIG. 4, four adjacent distance measuring points P (m, n) to P (m + 1, n +).
Detection area A consisting of the two-dimensional area defined in 1)
Distance measuring points P (m, n) located at the respective vertices of (m, n)
n) to P (m + 1, n + 1) measurement distance D (m, n) to
The absolute value of the difference of D (m + 1, n + 1) is calculated, and when all of the calculated values are smaller than the reference value, the detected area A
(M, n) is extracted as a box B (m, n) in which an object exists (step 200 in FIG. 3).

Next, at least the distances of the extracted plurality of boxes B (m, n) are generally obtained, and the box B (m, n) having the smallest distance is selected from the calculated distances. B (m, n), that is, at least the distance of the object 20 in front is obtained (step 300 in FIG. 3), but in this embodiment, the direction (angle of the center of the object 20 with respect to the center of the own vehicle) is also obtained.

Then, it is judged whether or not to continue or end these measurements and calculations (step 400 in FIG. 3).

FIG. 5 is a flow chart for explaining the concrete processing contents of the box extraction processing shown in step 200 of FIG. 3, in which "m = 1" is selected to select the first line (step 201). A flag indicating the presence / absence of a box is set to "BOX = 0" of "none" (step 20).
2). Next, select "n =" to select the first window.
1 ”(step 203), and at each apex of the detection area A (m, n) formed of a two-dimensional area defined by four adjacent distance measuring points P (m, n) to P (m + 1, n + 1). Distance measuring points P (m, n) to P (m + 1, n)
+1) measurement distance D (m, n) to D (m + 1, n + 1)
The absolute value of the difference is calculated and compared with the reference value S (step 2
04-209), when all of the calculated values are smaller than the reference value S, the detected area A (m, n) is extracted as a box B (m, n) in which an object exists (step 21).
Along with 0), a flag indicating the presence / absence of a box is set to “BOX = 1” of “present” (step 211). On the other hand, the distance measuring points P (m, n) to P (m + 1, n + 1)
When the absolute value of the difference between the measured distances D (m, n) to D (m + 1, n + 1) is calculated and compared with the reference value S and all the calculated values exceed the reference value S, the detected area A (m , N)
The flag is set to "BOX = 0" because there is no object in (step 212).

Then, n is incremented by 1 until "n = 6" (steps 213 and 214), and the above process is repeated. When "n = 6" is reached, m is incremented by 1 this time. By repeating the above processing until m = 15 ”(steps 215 and 216), the detection area A is detected.
The presence or absence of a box is detected for all (m, n).

The circled numbers 1 to 6 added to steps 204 to 209 in FIG. 5 correspond to the processing at the portions shown by the same numbers in FIG. Further, in FIG. 5, the reference value S is set to 1 meter, but this value is arbitrarily set in consideration of the usage environment and the like by experiments and the like, and a fixed value is set as in this embodiment. Not limited to what you use,
It may be a combination of an absolute value and a relative value described later.

FIG. 6 is a flow chart for explaining the specific processing contents of the distance and direction calculation processing shown in step 300 of FIG. 3. First, the counter value N = 0 is set (step 301). Then, when there is no box for all of the detected areas A (m, n) in the box extraction processing of step 200, the value of the distance DIST and the direction ANGL is not displayed because there is no object 20 in front (steps 302 to 304). ).

On the other hand, when the box B (m, n) is present in any of the detection areas A (m, n), "m = 1" and "n = 1" are set (steps 305, 306), and the detection area is detected. Check if A (1,1) is a box (step 3
07), and if it is the box B (1,1), the average distance BD (1,1) and the assumed direction BA (1,1) in the box B (1,1) are calculated (step 308,
309). The assumed direction BA (m, n) represents the amount of left / right deviation from the center of the host vehicle, and m = 8.5 in this embodiment.
Since the position of is the center of the vehicle, the assumed direction BA
When (m, n) is negative, it is on the left side of the vehicle, and when it is positive, it is on the right side of the vehicle 20 (see FIG. 1).
Indicates that is located.

Next, after the judgment of the count value N = 0 (step 310), the count value N = 1 is set (step 311). Then, the average distance BD (1,1) obtained previously
And the assumed direction BA (1,1) are stored as the distance DIST to the object 20 (see FIG. 1) and the direction ANGL (steps 312, 313). The above process is repeated until n = 6 (steps 314 to 317), but after the second time,
The distance DI stored in the new average distance BD (m, n)
The new average distance BD only when smaller than ST
Distance DIST in (m, n) and hypothesized direction BA (m, n)
And update the direction ANGL (step 318).

In the present embodiment, as shown in step 308 of FIG. 6, the average distance BD (m, n) is the box B
Four distance measuring points P (m, n) forming (m, n)
~ P (m + 1, n + 1) measurement distance D (m, n) ~ D
Although the method of obtaining the average of (m + 1, n + 1) is used, a method of taking the minimum value of the measurement distances D (m, n) to D (m + 1, n + 1) may be used.

Further, one from a plurality of boxes B (m, n)
Select one (minimum value box) and select the average distance BD (m,
n) and the assumed direction BA (m, n), the average distance BD obtained by the calculation in step 308 of FIG.
The smallest one (minimum value distance) of (m, n) was selected, but the box B obtained in steps 204 to 209 of FIG.
Distance measuring points P (m, n) to P constituting (m, n)
Measurement distances D (m, n) to D (m between (m + 1, n + 1)
The sum of the absolute values of the differences of (+1, n + 1) may be selected to be the smallest. The calculation formula in this case is the following formula 1.

[0027]

[Equation 1]

Further, since the error in the measured distance D (m, n) increases in proportion to the square of the distance as the distance increases, if a fixed value is used as the reference value S as in the above embodiment, the distance is The distance measurement distance D (m, n) is the short distance measurement distance D.
It is more difficult to make a correct determination than (m, n).
Therefore, it is desirable to set the reference value S according to the measurement distance D (m, n). Reference value S in this case
Is preferably the following expression 2, the preceding term of the addition symbol is a constant absolute value, and the following term is a relative value determined based on the measurement distance D (m, n) to be determined, for example, a = 1 m, b = 1
Set as 0.

[0029]

[Equation 2]

FIGS. 7 to 9 are views for explaining the experimental example of this embodiment and the results thereof. It was possible to satisfactorily capture the object 20 which is the preceding vehicle and measure its distance DIST and direction ANGL. . That is, by the box extraction processing of FIG.
Six detection areas A (6,5), A (6,6), A
(7,5), A (10,5), A (10,6), A (1
1, 5) are boxes B (6, 5) and B (6, 6), respectively.
6), B (7,5), B (10,5), B (10,
6) and B (11,5) (see the diagonally shaded portion in FIG. 8 and FIG. 9 in which the portion is enlarged and displayed), and the average distance BD (m, n) has the smallest minimum value distance. The minimum value box to have is box B (6,6), and the average distance BD (6,6) = 17.
The distance DIST is 80 meters, and the left and right deviation amounts from the center of the host vehicle are located on the left side by two lines, and thus can be measured as the direction ANGL = -1.12 degrees. It should be noted that in FIG. 8, the distance-measuring point P (m, n) without a numeral is a portion where the measuring distance D (m, n) is 50 meters or more.

Further, the marking symbols 21 such as white lines, letters and numbers drawn on the road surface are also obtained as distance data by the difference between the lightness and darkness of the road surface, but the detection areas including them are all larger than the reference value S. Since the object is not judged as a box and is ignored, it can be judged that the object is not an obstacle to the traveling of the own vehicle, and it is possible to prevent a problem such as issuing a false alarm. Similarly, it is possible to remove a long center line with a small width, such as a continuous center line,
It is possible to prevent problems such as issuing false alarms.

The distance DIST to the object 20 measured and obtained as described above is sent to the ECU (engine control unit) or the meter side of the vehicle as shown in FIG.
Similar to the conventional technique, it can be used as information for giving attention to the driver and automatically controlling the traveling speed.

The number of distance measuring points P (m, n) set in a grid can be arbitrary. In addition, a distorted state in which lines and windows do not extend in the vertical and horizontal directions, for example, four distance measuring points P (m,
The projection shape defined by n) may be set so that a two-dimensional region having a rhombus as a detection area may be set, but a vehicle occupying a large part of the object 20 has edges in horizontal and vertical directions (distance switching portions). Therefore, as a configuration for efficiently detecting these edges, the distance measuring point P (m,
It is desirable to set n) in a grid shape along the vertical and horizontal directions.

Further, instead of the image sensors 11L and 11R of the distance measuring unit 10, a sensor such as a two-dimensionally arranged optical sensor such as a photodiode may be used, but the distance measuring unit 10 should be small. The CCD is preferable in that it can be used. In addition to the so-called passive type, a so-called active type in which an obstacle is irradiated with a light beam and reflected light is received to measure the distance can be used, but the passive type is used because of its small size and low power consumption. Is desirable.

The detection area A (m, n) may be a two-dimensional area defined by at least three distance measuring points P (m, n), and the adjacent distance measuring points P may be used.
Instead of (m, n), one line or one window may be skipped and defined by the next distance measurement point P (m, n). However, the use of a large number of distance measuring points P (m, n) complicates the processing, and experiments by the present inventors show that it is desirable to use 3,4 points.

[0036]

The present invention can be obtained by a pair of sensors arranged corresponding to a plurality of distance measuring points set in a grid pattern of a plurality of lines arranged in the horizontal direction and a plurality of windows arranged in the vertical direction. Between the distance measuring points located at the respective vertices of the detection area formed of a two-dimensional area defined by at least three of the distance measuring points by calculating the phase difference from the image signal obtained. The absolute value of the difference between the measured distances is calculated, and when all of the calculated values are smaller than a reference value, the detected area is extracted as a box in which an object exists, and the distance measuring points constituting the box are By obtaining the distance to the object using the measurement distance, the measurement data of only the distance measurement points is used, so the data amount is small and the processing method is simple. It is possible to reduce the worry of malfunction.

The reference value is an absolute value which is a constant value, and a relative value obtained by multiplying a measured distance of one of the distance measuring points constituting the box by a constant ratio. By adding a value, the reference value can be set according to the measurement distance to be determined, and compared to the case where the set value is a fixed value, the determination at a long distance measurement distance can be made more accurately. It can be carried out.

The detection area includes a two-dimensional area in which a projection shape defined by three of the distance measuring points has a triangular shape, and the absolute difference between the distance measuring points between the distance measuring points located at each apex is absolute. A value is calculated, and when all of the three calculated values obtained by this are smaller than the reference value, the detected area is extracted as the box, or the detected area is four of the distance measuring points. The projection shape defined by is composed of a quadrangular two-dimensional area, and the absolute value of the difference in the measurement distances between the distance measurement points located at each vertex is calculated. By extracting the detection area as the box when it is smaller than the reference value, necessary and sufficient measurement data can be obtained, and the distance to the object can be easily measured. Kill.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of an exemplary embodiment of the present invention.

FIG. 2 is a diagram for explaining an array of distance measuring points in the above-described embodiment.

FIG. 3 is a flowchart illustrating processing in the above embodiment.

FIG. 4 is an enlarged view in which an unspecified part of FIG. 2 is extracted.

FIG. 5 is a flowchart illustrating the box extraction processing of FIG.

6 is a flowchart illustrating the distance and direction calculation processing of FIG.

FIG. 7 is a view for explaining an experimental example in the same example.

FIG. 8 is a diagram for explaining the experimental results of FIG.

9 is an enlarged view of a main part of FIG.

[Explanation of symbols]

 10 range-finding unit 20 object (vehicle) 30 microcomputer

Claims (4)

[Claims] 1. From an image signal obtained by a pair of sensors arranged corresponding to a plurality of distance measuring points set in a grid pattern of a plurality of lines arranged in a horizontal direction and a plurality of windows arranged in a vertical direction. The phase difference is calculated to obtain the measurement distance at the distance measurement points, and the measurement distance between the distance measurement points located at each apex of the detection area formed by a two-dimensional area defined by at least three of the distance measurement points. Calculate the absolute value of the difference of, when all of the calculated value is smaller than the reference value, the detection area is extracted as a box in which an object is present, and the measured distance of the distance measuring points constituting this box is calculated. A distance measuring device, characterized in that the distance to the object is obtained using the distance measuring device. 2. The reference value is an absolute value that is a constant value, and one of the distance measuring points that form the box.
2. A value obtained by adding a relative value obtained by multiplying the measured distances of the individual distance measuring points by a fixed ratio, and
The distance measuring device according to claim 1. 3. The detection area comprises a two-dimensional area in which a projection shape defined by three of the distance measuring points has a triangular shape, and an absolute difference between the distance measuring points between the distance measuring points located at each vertex is absolute. The distance according to claim 1 or 2, wherein a value is calculated, and when all of the three calculated values obtained by the calculation are smaller than the reference value, the detected area is extracted as the box. measuring device. 4. The detection area is composed of a two-dimensional area having a quadrangular projected shape defined by four of the distance measuring points, and the absolute difference of the measuring distances between the distance measuring points located at respective vertices. The distance is calculated according to claim 1 or 2, wherein a value is calculated, and when all of the six calculated values obtained by the calculation are smaller than the reference value, the detected area is extracted as the box. measuring device.