JPH0642966A - Rear-end collision prevention and alarm apparatus - Google Patents

Abstract

PURPOSE:To obtain a rear-end collision prevention and alarm apparatus wherein a data processing amount is reduced, a high-speed processing operation can be performed in a real-time manner and an alarm is issued in order to prevent a collision against a moving object such as a car or the like while the object is being tracked. CONSTITUTION:Quantized data on the same coordinate point is compared by a comparison means 2, a difference is taken and shading difference data is produced. The shading difference data is accumulated in the row direction at every row by a row accumulation means 3 and in the column direction at every column by a column accumulation means 4. A tracking means 5 tracks a preceding vehicle by the output of the row accumulation means 3, and a distance measuring means 6 measures a distance up to the preceding vehicle by the output of the column accumulation means 4. By means of the distance up to the preceding vehicle and by means of the speed of an own vehicle measured by a speed measuring means 7, an alarm judgment means 8 judges whether an alarm in order to prevent a collision is issued or not, and an alarm means 9 issues the alarm by the result of the judgment.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a rear-end collision prevention alarm device, and more particularly to a rear-end collision prevention alarm device which issues an alarm to prevent a rear-end collision with a moving object such as an automobile while tracking the object. .

[0002]

2. Description of the Related Art Conventionally, as a method of identifying a preceding vehicle, for example, an image recognition method by an optical flow method disclosed in Japanese Patent Laid-Open No. 2-241855, for example, Japanese Patent Laid-Open No. 3-1
A data matching method based on window processing disclosed in Japanese Patent No. 70012 has been proposed.

[0003]

However, in the above-described conventional apparatus, since the processing for sequentially comparing each pixel is required, the data processing amount is enormous, about 1 to 10 Mbytes, and therefore a high-speed image is obtained. It has not been put to practical use for vehicles that require treatment.

Therefore, the present invention solves the above-mentioned conventional problems, reduces the amount of data processing and enables high-speed processing in real time, and allows a moving object such as an automobile to be tracked while colliding with the object. It is an object of the present invention to provide a new rear-end collision prevention warning device that issues a warning for prevention.

[0005]

In order to achieve the above object, a rear-end collision prevention warning device according to the present invention is obtained by photographing a traveling road at regular time intervals as shown in the basic configuration diagram of FIG. Further, a plurality of frame memories 1 for storing the grayscale quantized data of the image for one frame respectively and the quantized data at the same coordinate points of the plurality of frame memories 1 are compared to obtain a difference, and the grayscale difference data is obtained. The comparing means 2 to be created, the column accumulating means 3 for accumulating the grayscale difference data obtained by the comparing means 2 in each column in the column direction, and the grayscale difference data obtained by the comparing means 2 in the row direction for each row. Row accumulating means 4 for accumulating, tracking means 5 for tracking a preceding vehicle by the output of the column accumulating means 3, distance measuring means 6 for measuring the distance to the preceding vehicle by the output of the row accumulating means, and own vehicle Speed to measure speed A measuring means 7, an alarm determining means 8 for determining whether or not to issue an alarm for preventing a rear-end collision based on the distance measured by the distance measuring means 6 and the speed measured by the speed measuring means 7, and the alarm determining means. And an alarm means 9 for issuing an alarm based on the judgment by the above.

[0006]

With the above-mentioned structure, the data stored in the plurality of frame memories 1 in a matrix are obtained at fixed time intervals.
The grayscale quantized data of the image for the frame is stored in a matrix in a plurality of frame memories 1, and the quantized data at the same coordinate points of the image are compared by the comparison means 2 to obtain the grayscale difference data. The obtained gray level difference data is accumulated in the column direction by the column accumulating means 3 and in the row direction by the row accumulating means 4 in the row direction.

The magnitude of the cumulative value obtained by the row accumulating means 3 reflects the presence or absence of a change between a plurality of images, and the change and the magnitude of the cumulative value of the portion corresponding to the preceding vehicle having a relatively small relative speed are: Since it is small, the tracking means 5 can track the preceding vehicle by the output of the row accumulating means 3.

Further, since the accumulated value obtained by the line accumulating means 4 greatly changes in a portion corresponding to a preceding vehicle whose relative speed is relatively small and its position on the screen does not change so much, it becomes a memory address on the frame memory. By setting the distance in advance, the distance measuring means 6 can measure the distance from the preceding vehicle by detecting the position where the cumulative value changes suddenly by the output of the row cumulative value means 4.

It is empirically known that it is necessary for safe driving to maintain the distance to the preceding vehicle as the speed of the own vehicle is higher, and the relationship between the speed of the own vehicle and the distance to the preceding vehicle. Since the alarm range can be set in advance by whether the warning determination unit 8 issues a warning for preventing a rear-end collision according to the distance to the preceding vehicle measured by the distance measuring unit 6 and the speed measured by the speed measuring unit 7. It is determined, and the alarm means 9 can issue an alarm according to the result.

Therefore, it is possible to warn of the danger of a rear-end collision by measuring the distance to the preceding vehicle while tracking the preceding vehicle without using a large-capacity memory.

[0011]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 2 is a diagram showing an embodiment of a rear-end collision prevention warning device according to the present invention. In the figure, reference numeral 11 denotes a CCD camera as an image pickup means including an optical system 11a and a CCD image sensor 11b. Reference numeral 12 denotes an A / D conversion unit, which inputs an analog image signal from the CCD camera 11 and A / D converts it to obtain quantized data at its output. 1
Reference numeral 3 denotes a storage device which is composed of two frame memories M1 and M2 having the same configuration and constitutes a plurality of frame memories 1. Each of the frame memories M1 and M2 outputs a grayscale value quantum for one frame output from the A / D converter 12. Alternate data is stored alternately.

When the input image is as shown in the figure, the photographing condition of the CCD camera 11 is, for example, that the route of the vehicle including the center line LO and the side line LS is a on the lowest line of the input image. , B point, and the distance from the host vehicle to points a and b is set to 10 m. Further, the point at infinity FOE is set so as to come to a point ⅔ from the lowest line on the center line of the input image.

The capacity of the A / D converter 12 is arbitrarily set so as to perform quantization in the range of 4 to 8 bits depending on the required resolution. On the other hand, the frame memories M1 and M2 are
In detail, the configuration is as shown in the figure, so that n quantized data Xn in the X-axis direction and k quantized data Yk in the Y-axis direction are stored corresponding to the input image. It is configured. The capacity n × k is 64 to 512 bytes (= n, depending on the required resolution as in the A / D converter 12).
It is arbitrarily set within the range of k). In the example of FIG. 3, both frame memories M1 and M2 have n and k of 25.
It is set to 6 bytes.

Reference numeral 14y denotes a Y-direction comparator, which sequentially quantizes the quantized data from Y _{1 to} Yk (256) for each column of both frame memories M1 and M2 in which the quantized data is recorded. The difference between the quantized data representing the gray value, that is, the gray value difference is calculated and output.

Reference numeral 15y is an averaging memory composed of m memories X _{1 to} Xm. Each of the memories X _{1 to} Xm has a capacity of n × k, like the frame memories M1 and M2.
The gray value difference data between the same sample points of the two frames output from the direction comparator 14y is stored.

[0016] 16y is a X-axis counter of n counters, each counter Y (column shading value difference data for each sequence stored in the m memory X ₁ through XM averaging memory 15y ) Direction, count all, and hold the count value. As a result, in the X-axis counter 16y, n averaged accumulated values of the grayscale value differences in the Y direction are obtained. Therefore, the average change value of the grayscale value difference accumulated value in the X-axis direction can be known from the contents of the X-axis counter 16y,
The contents of the X-axis counter 16y are serially output.

Reference numeral 14x is an X-direction comparator, which sequentially quantizes the quantized data from Y _{1 to} Yn (256) for each column of both frame memories M1 and M2 in which the quantized data is recorded. The difference between the quantized data representing the gray value, that is, the gray value difference is calculated and output.

Reference numeral 15x is an averaging memory composed of m memories Y _{1 to} Ym. Each of the memories Y _{1 to} Ym has a capacity of n × k like the frame memories M1 and M2.
The gray value difference data between the same sample points of the two frames output from the direction comparator 14x is stored.

Reference numeral 16x is a Y-axis counter composed of n counters, and each counter Y (row) for each row is the gray value difference data stored in the m memories Y _{1 to} Ym of the averaging memory 15x. Count all in the direction and hold the count value. As a result, in the Y-axis counter 16x, k averaged cumulative values of the grayscale value differences in the X direction are obtained. Therefore, the average change value in the Y-axis direction of the grayscale value difference cumulative value can be known from the contents of the Y-axis counter 16x,
The contents of the Y-axis counter 16x are serially output.

The Y-direction comparator 14y and the X-direction comparator 14x constitute the comparison means 2, and the X-axis counter 1
The 6y and Y-axis counter 16x constitute the column accumulating means 3 and the row accumulating means 4, respectively.

Reference numeral 17 denotes a processing circuit, which is an X-axis counter 16
The average change values in the Y and X directions output from the y and Y axis counters 16x are input. Then, the inputted average change value in the X direction is processed to obtain the vehicle width D of the preceding vehicle according to the portion where the change in the density difference is the minimum, that is, the number of X-axis pixels in the frequency range, as shown in FIG. Is detected and the preceding vehicle is tracked.

Further, the processing circuit 17 processes the input average change value in the Y direction, detects the position where the frequency of the density change drastically changes, as shown in FIG. It measures the distance and outputs the preceding vehicle distance signal. Processing circuit 1
Reference numeral 7 serves as a tracking means 5 and a distance measuring means 6.

Reference numeral 18 denotes a ROM as a correction value storage device that stores a table of correction values used in the processing by the processing circuit 18 described above, and this constitutes a table storage means 61.

Reference numeral 19 denotes a speedometer as speed measuring means 7 for measuring the speed of the host vehicle, for example, processing a distance pulse generated by a rotation sensor (not shown) driven by the rotation according to the rotation of the wheels branched from the transmission. The speed is measured and the speed signal is output.

Reference numeral 20 denotes an alarm processing circuit as the alarm determination means 8, which inputs the preceding vehicle distance signal output by the processing circuit 17 and the speed signal output by the speedometer 19, and uses these signals to prevent a rear-end collision. Determine whether to give an alarm,
When it is determined that an alarm should be issued, an alarm signal is output. As will be described in detail later, the alarm processing circuit 20 performs an alarm process to output two types of alarm signals, a caution alarm and a danger alarm.

Reference numeral 21 is an alarm device such as a buzzer as the alarm means 9, which operates in response to an alarm signal output from the alarm processing circuit 20 to generate an alarm sound, but is generated in response to two kinds of alarm signals. The warning sound is changed so that caution warning and danger warning can be distinguished.

The operation of the apparatus having the above configuration will be described below. First, the CCD camera 11 captures the foreground from the vehicle running and captures it as a video signal. This C
The analog image signal output from the CD camera 11 is A / D converted and quantized by the A / D converter 12. The one frame of quantized data quantized by the A / D conversion unit 12 is recorded in one frame memory M1 of the storage device 13.

Although it can be arbitrarily set, the analog image signal of the foreground photographed by the CCD camera 11 is input after a fixed time Δt, which is generally about 30 milliseconds, and the quantized data for one frame is stored in the other side of the storage device 13. Frame memory M2
To record. As described above, since the quantized data for one frame is alternately recorded in the frame memories M1 and M2, it is controlled by the system controller (not shown).

When the grayscale value quantized data for one frame has been recorded in each of the frame memories M1 and M2, the columns and rows of the frame memories M1 and M2 are individually compared. The quantized data at the same sample point in both frame memories M1 and M2 are read and input to the Y-direction comparator 14y and the x-direction comparator 14x.

The case of comparing the quantized data in the first column of the frame memories M1 and M2 will be specifically described with reference to FIG. First, 1 of M1 and 2 of M2 are compared, and the difference 1 is written in the memory X ₁ of the averaging memory 15y. Next, 1 of M1 and 2 of M2 are compared, and the difference 1 is compared to the same memory X.
Writing to _1, hereinafter 2 and 3, ..., the difference 1 and finally sequentially compares 8 and 5, ..., sequentially sequentially written in the memory X ₁ to 3. This operation is similarly performed for columns 2 to 256. Repeat the same process for lines 1 to 256
The processing of the frame ends. The processing for one frame is completed within the time of 30 milliseconds described above.

When the processing for one frame is completed, the next one frame of quantized data is recorded in place of the previous quantized data in the frame memory M1 in which the quantized data is previously recorded. After the recording of (3) is completed, the process of obtaining the difference between the same sample points by the above-mentioned comparison of the quantized data is performed.

Now, the averaging memories 15y and 15x are
Three memories X each₁~ X₃And Y ₁~ Y₃Consisting of
X-axis counter 16y and Y-axis counter 1
The cumulative value of each column and each row of each memory is calculated by 6x.
After obtaining the value, simply add the cumulative values of the three memories and
The value is transferred to the X-axis counter 16y and the Y-axis counter 16x.
Hold each. The averaging memories 15y and 15x are
To prevent the effects of disturbance such as direct light
Is.

[0033] In the embodiment of FIG. 2, each memory X ₁ to X ₃ and Y ₁ to Y ₃ in the averaging memory has been described as having the same capacity and the frame memory, adds the outputs of the comparators 14y and 14x And store it in memory X _{1 to} X
_As shown in FIG. 5, ₃ can be replaced with one having a capacity of n × 1, and similarly, the memories Y ₁ to Y ₃ can also be replaced with one having a capacity of 1 × k. Then, it may be stored in the Y-axis counter 16y to indicate simple addition average value of the accumulated value of the memory X ₁ to X ₃ in FIG.

The averaging memories 15y and 15x are also provided.
By detecting the value of, it is possible to detect a sudden signal value change such as vehicle vibration and cancel the data of the frame.

The output of the X-axis counter 16y corresponds to the input image, as can be seen from FIG. Thus, the traveling direction of the preceding vehicle, that is, the tracking can be determined from the vehicle width and the number of pixels (memory address) of the preceding vehicle.

Therefore, although it cannot be determined in the prior art because it takes a long processing time when there is a large change in the preceding vehicle, such as the curve determination, according to the present invention, the column memory facilitates the frequency analysis. This is because the width of the window and the lateral position (row position) change depending on the distance of the preceding vehicle, the curve, etc., so that it can be automatically tracked by the value of the X-axis counter.

The preceding vehicle, which is traveling at almost the same speed as the host vehicle, has a relative speed close to that of the host vehicle, and therefore the X-axis counter 16y.
It can be determined that the minimum level of the average change value is the preceding vehicle because the output of the vehicle has little change. In the tracking of the preceding vehicle, when the distance to the preceding vehicle is large or the preceding vehicle escapes to the left and right edges of the screen due to a curve or the like, noise components other than the preceding vehicle are significantly reduced and the output signal of the Y-axis counter is reduced. Is stable.

The output of the Y-axis counter 16x can be used to measure the distance to the preceding vehicle, as shown in FIGS. That is, in the figure, the address of the Y-axis counter 16x and the optical distance are set in advance,
Since the point where the rapid change of the change value data occurs is the position of the preceding vehicle, the distance of the preceding vehicle can be measured.
In this embodiment, address 256 is set to 10 m for explanation.

Due to the above-mentioned operation, the processing circuit 17 performs logical processing for calculating and judging the position and distance of the preceding vehicle based on the output of the X-axis counter and the output of the Y-axis counter, and is stored in the ROM 18 as the judgment reference. Use the correction value table.

The alarm processing circuit 20 inputs the preceding vehicle distance signal output by the processing circuit 17 and the speed signal output by the speedometer 19, and determines whether to issue an alarm for preventing a rear-end collision based on these signals. To do. In order to make this determination, the alarm processing circuit 20 previously has determination data indicating a caution warning range or a danger warning range depending on the relationship between the speed of the host vehicle and the distance to the preceding vehicle, as shown in FIG. According to this judgment, a caution warning signal is output when it is judged that it is in the caution warning range, and a danger warning signal is outputted when it is judged that it is in the danger warning range.
To enter.

The alarm device 21 generates an alarm sound capable of distinguishing the two types of alarm signals in response to the input of the caution alarm signal or the danger alarm signal, and alerts the driver or warns of danger.

In the above-described embodiment, it is sufficient to prepare a memory of about 100 kbytes, which is 1/10 or less of the memory capacity required for the conventional image processing, and about 30 mm for measuring the distance from the preceding vehicle. Real-time processing of seconds is possible.

[0043]

As described above, according to the present invention, the grayscale quantized data of an image for one frame obtained at fixed time intervals are stored in a matrix, and the quantized data of the same coordinate point is quantized. The data is compared and the difference is calculated to create density difference data, and this density difference data is accumulated in the column direction for each column and in the row direction for each row. Since the distance to the preceding vehicle is measured by the value, the preceding vehicle can be tracked and the distance to the preceding vehicle can be measured without using a large capacity memory.

Therefore, the amount of data processing is reduced to enable high-speed processing in real time, and when the vehicle is approaching rapidly due to the distance to the preceding vehicle and the speed of the host vehicle, a warning is issued to prevent a rear-end collision and the risk of a rear-end collision is increased. Can warn.

[Brief description of drawings]

FIG. 1 is a block diagram showing a basic configuration of a target tracking type distance measuring device according to the present invention.

FIG. 2 is a diagram showing an embodiment of a target tracking type distance measuring device according to the present invention.

3 is a diagram showing a specific configuration of a frame memory in FIG.

FIG. 4 is an explanatory diagram for explaining the operation of the comparator in FIG.

5 is an explanatory diagram for explaining the operation of the averaging memory in FIG. 2. FIG.

FIG. 6 is an explanatory diagram for explaining the operation of the counter in FIG.

FIG. 7 is a diagram for explaining the principle of vehicle width detection by the processing circuit in FIG.

FIG. 8 is a diagram for explaining the principle of measuring the distance to the preceding vehicle by the processing circuit in FIG.

FIG. 9 is a diagram used for explaining the principle of FIG. 8;

10 is a diagram for explaining the principle of making an alarm determination in the alarm processing circuit in FIG.

[Explanation of symbols]

 1 Frame Memory 2 Comparison Means 3 Column Accumulation Means 4 Row Accumulation Means 5 Tracking Means 6 Distance Measuring Means 7 Speed Measuring Means 8 Alarm Judging Means 9 Alarm Means

─────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] December 11, 1992

[Procedure Amendment 1]

[Document name to be corrected] Drawing

[Name of item to be corrected] Figure 3

[Correction method] Change

[Correction content]

[Figure 3]

[Procedure Amendment 2]

[Document name to be corrected] Drawing

[Name of item to be corrected] Figure 4

[Correction method] Change

[Correction content]

[Figure 4]

[Procedure 3]

[Document name to be corrected] Drawing

[Name of item to be corrected] Figure 5

[Correction method] Change

[Correction content]

[Figure 5]

[Procedure amendment 4]

[Document name to be corrected] Drawing

[Name of item to be corrected] Figure 6

[Correction method] Change

[Correction content]

[Figure 6]

[Procedure Amendment 5]

[Document name to be corrected] Drawing

[Name of item to be corrected] Figure 7

[Correction method] Change

[Correction content]

[Figure 7]

[Procedure correction 6]

[Document name to be corrected] Drawing

[Correction target item name] Figure 8

[Correction method] Change

[Correction content]

[Figure 8]

Claims (1)

[Claims] 1. A plurality of frame memories for respectively storing, in a matrix form, grayscale quantized data of one frame of an image obtained by photographing a traveling road at regular time intervals, and the same coordinate points of the plurality of frame memories. Comparing means for comparing the quantized data of ## EQU1 ## to obtain a difference, and creating grayscale difference data; a column accumulating means for accumulating the grayscale difference data obtained by the comparing means in each column in the column direction; Row accumulating means for accumulating the light and shade difference data for each row in the row direction, tracking means for tracking the preceding vehicle by the output of the column accumulating means, and distance for measuring the distance to the preceding vehicle by the output of the row accumulating means Whether or not an alarm for preventing a rear-end collision is issued based on the measuring means, the speed measuring means for measuring the speed of the own vehicle, and the distance measured by the distance measuring means and the speed measured by the speed measuring means. And determining alarm determination means, collision prevention alarm device, characterized in that it comprises an alarm means for generating an alarm by determining by said alarm judging means.