JPH10206116A - Object recognizing device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable the correct recognition of an object by recognizing the object when a representative distance for every line on the basis of distance data for each region obtained by dividing images is smaller than a threshold and setting a smaller threshold for a line on the side of nearer distances. SOLUTION: At a distance measuring means 16, the image of a multilevel line-type CCD 11 is plurality divided every line in the direction of windows to measure a distance for each region. This distance data is processed at a low-pass filter and then processed at an 8 adjacent point processing part 25. The low-pass filter is configured so that the degree of smoothing is smaller in liens on the side of nearer distances than in lines on the side of farther distances. Next, at a line distance calculating part 26, on the basis of a distance in every region measured by the distance measuring method 16, a representative distance in every line is calculated. In an object recognizing part 20, when the number of continuos calculations of a representative distance for every line within every predetermined length of time by the line distance calculation part 26 is larger than a threshold value, an object is recognized from the representative distance.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an object recognizing apparatus, and more particularly to a technical field related to an apparatus for recognizing an object based on distance data obtained by a multi-stage line CCD.

[0002]

2. Description of the Related Art Conventionally, as an object recognizing device of this type, for example, as shown in Japanese Patent Application Laid-Open No. 5-52562, a captured image is divided into a plurality of windows, and the distance between each window is divided. An object that recognizes an object based on the object is known. That is, in this device, an object such as a preceding vehicle is imaged by a pair of image sensors arranged in the vertical direction, an image is displayed by one of the image sensors, and the display screen is divided into a plurality of windows. The distance to the object is measured for each window, the window of the target object is recognized based on the distance value, a tracking window is set, and the distance in the tracking window is measured.

[0003]

In recognizing an object based on its distance, a large number of CCDs arranged in a certain direction are arranged in multiple stages in a direction orthogonal to the arrangement direction. A multi-stage line type CCD is provided, and two lines obtained based on the luminance signal from the multi-stage line type CCD are provided.
It is conceivable to recognize a specific object from the dimensional distance data and keep following it. In other words, this multi-stage line type CCD has a configuration in which CCDs are thinned in one direction in comparison with sensors for cameras and the like in which a large number of CCDs are arranged vertically and horizontally. The number is reduced, and the cost can be reduced.

[0004] However, on the other hand, there is a problem that the dispersion of ranging data and the influence of noise are great, and it is difficult to calculate distances with high accuracy and to recognize objects accurately. In particular, CCD lines in a multi-stage line type CCD are spaced at equal intervals,
When one end of the multi-stage line type CCD in the line direction is a short distance line for detecting a short distance, and the other end is a long distance line for detecting a long distance, the recognition object is When approaching relatively, the time that the object crosses the near line in the near area is shorter than the time that the object crosses the far line in the far area. If the data processing takes time to perform, the object recognized on the long distance side will be lost on the short distance side.

[0005] The present invention has been made in view of the above points, and an object thereof is to provide a multi-stage line type CCD.
By devising the processing of the distance data by using the multi-stage line type CCD, high-precision distance calculation can be performed and accurate object recognition can be performed, and the recognition object can be recognized regardless of the distance to the recognition object. The purpose is to ensure that the vehicle can continue to follow.

[0006]

In order to achieve the above object, according to the present invention, an image obtained by a multi-stage line type CCD is divided into a line direction and a window direction, and a distance is measured for each region. Then, the representative distance for each line is calculated based on the distance data for each area, and when the number of times that the representative distance is continuously calculated at a predetermined time is greater than a threshold value, the representative distance is calculated. , And the threshold value is set to be smaller as the distance is closer to the line.

Specifically, according to the first aspect of the present invention, as shown in FIG. 1, CCD lines composed of a large number of CCDs arranged along a window direction are arranged in multiple stages in a line direction orthogonal to the window direction. A multi-stage line type CCD 11 for capturing an image as luminance information, and a multi-line CCD 11 obtained based on a luminance signal from the multi-stage line type CCD 11.
The object is an object recognition device that recognizes a specific object from dimensional distance data.

One end of the multi-stage line type CCD 11 in the line direction is a short-distance side line for detecting a short-distance side, and the other end is a long-distance line for detecting a long-distance side. A distance measuring unit 16 that divides an image obtained by the line type CCD 11 into a plurality of parts for each CCD line and in a window direction to measure a distance for each region, and a distance for each region measured by the distance measuring unit 16 And the number of times that the representative distance for each line is continuously calculated by the line distance calculating means at predetermined time intervals is greater than a threshold value. An object recognizing means for recognizing an object based on the representative distance, wherein the threshold value is smaller for a short distance side line than for a long distance side line. It is assumed to be Ku set.

With the above configuration, first, in the distance measuring means 16, the image of the multi-stage line type CCD 11 is divided into a plurality of lines for each line and in the window direction, and the distance is measured for each region. Next, a representative distance for each line is calculated by the line distance calculating means 26 based on the distance for each area measured by the distance measuring means 16, and a representative distance for each line is calculated by the line distance calculating means 26 in the object recognition means 20. When the number of times that the distance is continuously calculated every predetermined time (one sampling time) is larger than a threshold value, the object is recognized from the representative distance. In other words, when the validity of the distance data is high, the representative distance is calculated. On the other hand, when the validity of the distance data is low, the representative distance is not calculated. Since the validity of the distance data and the representative distance can be determined to be high, the object is recognized from the representative distance when the number of times is greater than the threshold value.

Therefore, the representative distance for each line is obtained based on the distance data for each area as described above.
When the effectiveness of the representative distance is high, the object is recognized from the representative distance, so that even if there is variation or noise in the distance measurement data, the influence can be reduced as much as possible.
High-precision distance calculation becomes possible, and accurate object recognition can be performed.

Since the threshold value is set smaller for a short distance side line than for a long distance side line, the object recognition can be performed in a short time on the short distance side line. If the CCD lines of the CCD are at equal intervals, when the recognition object relatively approaches, the object recognized on the long distance side will not be lost on the short distance side. In addition, even if the threshold value is set smaller for a line on the short distance side, since the distance detection accuracy on the short distance side is higher than that on the long distance side, object recognition on the short distance side can be performed accurately. Therefore, it is possible to reliably follow the recognition object regardless of the distance.

According to a second aspect of the present invention, in the first aspect of the present invention, there is provided a smoothing processing means for performing a smoothing process on the distance data of each area measured by the distance measuring means 16,
It is assumed that the smoothing processing means is configured to make the degree of smoothing smaller on a short distance side line than on a long distance side line.

According to the present invention, since the distance data for each area measured by the distance measuring means 16 is subjected to smoothing processing by the smoothing processing means, it is more accurate without being affected by sudden noise or the like. Can recognize the object. Since the degree of smoothing is smaller for the near line as compared to the far line, the time required for the processing is shorter for the near line, and even when the distance data is smoothed. In addition, the recognition object is not lost on the short distance side.

According to a third aspect of the present invention, in the first or second aspect of the present invention, the object recognizing means is configured to output a signal such as an alarm based on the result of the object recognition. This makes it easy to know the recognition object,
When the risk of the object approaching relatively is high, the effect can be further exhibited.

[0015]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 2 shows a vehicle C (automobile) equipped with the object recognition device according to the embodiment of the present invention, and the object recognition device is an object O (see FIG. 5, shown in FIGS. 11 and 12).

In FIG. 2, reference numeral 1 denotes a vehicle body of a vehicle C, 2 denotes a vehicle room formed at a substantially central portion in the front and rear direction of the vehicle body 1, 3 denotes an engine room formed at a front end of the vehicle body 1, and 4 denotes a vehicle room. An instrument panel 5 at the front end is a package tray at the rear end of the vehicle compartment 2, and 6 is a rear window glass. Then, as shown in FIG. 3, the object recognition device receives the left and right rear side detection sensors 10 and 10 for measuring the distance to the object O, and the output signals of the respective detection sensors 10. Controller 15 which receives a signal from the controller 15
The display device 31 includes a display device 31 for displaying by RT, liquid crystal, or the like, and an alarm device 32 for alarming the degree of danger of the object O. Then, as shown in FIG.
Are attached and fixed to both left and right end portions of the package tray 5 in a state of facing obliquely rearward, respectively. The controller 15 is provided at the rear end of the engine room 3, and the display device 31 and the alarm device 32 are provided at the instrument panel 4.

As shown in FIG. 5, each of the detection sensors 10
Is a pair of upper and lower Cs arranged in a vertical direction at a predetermined distance.
The image pickup apparatus includes CD chips 11, 11 and lenses 12, 12 arranged corresponding to the CCD chips 11, 11.
Each of the CCD chips 11 is composed of a multi-stage line type CCD in which CCD lines composed of a large number of CCDs arranged in a vertical window direction are arranged in multiple stages in a line direction (horizontal direction) orthogonal to the window direction. Each of the CCD chips 11 passes through the lens 12 through the rear window glass 6 of the vehicle C through the lens 12 and has a range of an angle θ1 in the vertical direction and a range of an angle θ2 in the horizontal and horizontal directions (FIGS. 10 and 12).
) Is captured as luminance information.

As shown in FIG. 4, each of the detection sensors 10 is connected to a distance measuring circuit 16 (distance measuring means) in a controller 15. Each of the distance measuring circuits 16 is connected to both CCs.
A parallax calculating unit 17 that calculates the parallax (phase difference) of the object image at the D chips 11 and 11 and a distance calculating unit 18 that calculates a distance to the object O based on a signal from the parallax calculating unit 17 are provided. . Then, as shown in FIGS. 6 and 7, each distance measuring circuit 16 converts the image captured by each CCD chip 11 into n lines for each CCD line in the line direction (horizontal direction).
, And each line is divided into m windows in the window direction (vertical direction), so that substantially the entire image is composed of m × n areas E, E,. The parallax between the same regions E, E in the images obtained by the CCD chips 11, 11 is obtained, and the distance to the object O is calculated for each region E from the parallax.

That is, the images captured by both CCD chips 11, 11 are as shown in FIG. 6, but the images of both CCD chips 11, 11 are at the same line position (line i in the illustrated example). As shown in FIG. 8, the two CCD chips 11 and 11 are displaced in the vertical direction to generate parallax, and the distance to the object O is measured using the parallax. This principle will be described with reference to FIG. 9. The triangles P.O1.Q and O1.P1 in FIG.
・ Q1 and triangles P · O2 · Q and triangle O2
Since P2 and Q2 are similar to each other, the distance from the detection sensor 10 (lens 12) to the object O is a, the distance between the centers of the lenses 12, 12 is B (constant),
The focal length of the lens 12 is f (constant), and both CCD chips 1
Assuming that the displacement amounts of the object image from the lens center at 1, 1 and 11 are x1 and x2, respectively, a · x1 / f = Ba−x2 / f, and from this equation, a = B · f / (x1 + x2) Is obtained. That is, the distance a to the object O can be measured by the parallax (phase difference) of the object image between the two CCD chips 11, 11.

G (white dots) in FIGS. 6 and 7 are:
The distance measuring points (ranging points) are arranged in a vertical and horizontal lattice so as to correspond to the CCD of the CCD chip 11, and the distance measuring points G are included in each area E. The window of each CCD line is divided so that a part of the window overlaps with the adjacent window, and the same distance measuring points G, G, G, and G are located in regions E and E adjacent in the vertical direction (window direction). …It is included. O 'is an image of the object.

Further, as shown in FIG.
A plurality of lines formed by dividing the image captured by the chip 11 line by line are arranged such that the line position for measuring a short distance outside the vehicle C has a smaller number, while the center position in the vehicle width direction has a smaller number. The line position for measuring a long distance is set to a large number, and the number is sequentially increased from the outer line toward the center line in the vehicle width direction.

As shown in FIG. 4, a specific object is provided to the controller 15 based on two-dimensional distance data in the vertical direction and the horizontal direction obtained based on the sensor 10, that is, a signal from each distance measuring circuit 16. An object recognition unit 20 that recognizes O;
An object selecting section 21 for selecting whether or not the object O is a new object based on an output signal of the object recognizing section 20;
And a danger determining unit 22 for determining whether the object O selected in Step 1 is a dangerous object for the vehicle C (own vehicle). The object recognizing unit 20 displays a display signal based on the recognition result of the object O. Is output to the display device 31 and an alarm signal is output to the alarm device 32 through the object selection unit 21.

The controller 15 also includes a range cut unit 24 for excluding an unnecessary range where the object O cannot be originally located in recognizing the object O, a distance data for each measured area, and a surrounding area. An eight-neighbor point processing unit 25 as effective point number giving means for comparing the eight adjacent areas (eight-neighbor point processing) and giving an effective point number, and a line distance calculating unit 26 for calculating the distance for each line , A guardrail determination unit 27 for determining a guardrail, and a grouping unit 28 for grouping distance data for each object O.

FIG. 11 shows a range cut range Z1 in the vertical direction excluded by the range cut section 24, and FIG. 12 shows a range cut range Z2 in the horizontal direction. These range cut ranges Z1 and Z2 are shown. Is detected based on the angle of the line and the distance at that position. FIG.
2, F is the road surface of the vehicle C, M is a white line on the road surface F that sets the vehicle traveling lane on the road, F1 is a roadside zone installed on both sides of the road, and H is an implant.

As shown in FIG. 13, the eight adjacent point processing operation of the eight adjacent point processing unit 25 is performed by dividing the distance data of a certain area E (i, j) into eight adjacent areas R1 to R5. This is for determining the correlation of the distance data of R8, and is specifically performed as shown in FIG. That is, in the first step S1, distance data d (i, j) for each area E (i, j) divided into the number n of lines and the number m of windows is read, and in the next step S2, each area E (i, j) is read. The number of effective points P (i, j) of j) is initialized to P (i, j) = 0. The number of effective points P (i, j) is calculated for each area E (i, j).
j), the larger the value is, the higher the validity of the distance data of the area is, and it is determined that the distance data is reliable and credible. In the next step S3, the number of effective points to the area (lattice point) located at the left and right ends and the upper and lower ends of all the areas is increased, and the effective points P
(I, j) is set so as to be increased by +1 and the number of effective points P (i, j) is increased by +2 in the four corner areas. Thereafter, in step S4, it is determined whether or not to perform the adjacent point processing.
In the case of O, after setting the number of effective points P (i, j) to P (i, j) = 8 in step S11, the process proceeds to step S12, while if the determination is YES, the process proceeds to step S5.

The determination of whether or not to perform the adjacent point processing in step S4 is, for example, to determine whether or not a reference distance value previously determined for each line position is larger than a predetermined value. Is larger than a predetermined value, the adjacent point processing is not performed (proceed to step S11), while if the reference distance value is not larger than the predetermined value, adjacent point processing is performed (proceed to step S5).

In step S5, a distance threshold value d0 is set. This distance threshold value d0 is determined by the number of assigned points p
For example, and is set to a constant, for example.

After step S5, the process proceeds to step S6,
The distance data d (Ri) of the adjacent region Ri is read, and in the next step S7, a distance difference dx = | d (i, j) -d (Ri) | between the region E and the adjacent regions R1 to R8 is calculated. Thereafter, in step S8, it is determined whether or not the distance difference dx is smaller than the distance threshold value d0. When the determination is NO, the process proceeds to step S12, and when the determination is YES, the process proceeds to step S9. The number p of points to be given is set.

After such step S9, step S
At 10, the number of valid points P (i, j) up to that point
Is added to the number of assigned points to set a new number of valid points P (i, j) = P (i, j) + p, and the process proceeds to step S12. In this step S12, step S
It is determined whether or not the processes of 6 to S10 have been completed for each of the eight adjacent regions R1 to R8. If the determination is NO, the process returns to step S6, and the same process is performed for the other remaining adjacent regions. On the other hand, if the determination is YES, the process advances to step S13 to set the number of effective points P (i, j) for all the areas E, E,.
10) is determined. When this determination is NO, the process returns to step S4, and the setting of the number of effective points P (i, j) is repeated for another area E. On the other hand, if the determination is YES, the process proceeds to the next distance calculation processing for each line (see FIG. 16).

FIG. 16 shows the processing operation in the line distance calculating section 26, and the number of effective points P set by the eight adjacent point processing section 25 (effective point number giving means).
The distance for each line is calculated based on (i, j).

First, in step T1, the number of lines n
And the distance data d (i, j) for each area E divided into the number m of windows, and the number of effective points P (i, j) for each area E given by the above-described eight adjacent point processing. In step T2, the line representative effective point number PI (i) is initialized to PI (i) = 0. The line representative effective point number PI (i) is set for a line at the time of distance calculation for each line, and the larger this value is, the higher the validity of the distance data of the line is, and the more reliable and reliable it is. Is determined.

In the next step T3, a line representative threshold value P0 corresponding to the line representative effective point number PI (i) is set. The line representative threshold value P0 in step T3 is set to, for example, a constant value.

After step T3, the process proceeds to step T4,
It is determined whether or not the number of effective points P (i, j) for each area is larger than the line representative threshold value P0. When this determination is NO, the process directly proceeds to step T6, but when the determination is YES, in step T5, the representative distance l (i) for each line is updated for averaging, and the line representative effective point number PI is updated. After updating the line representative effective point number PI (i) by adding the effective point number P (i, j) for each area to (i), step T6.
Proceed to. That is, in the line distance calculating unit 26, the number of effective points P set by the eight adjacent point processing unit 25 is set.
Distance calculation is performed for each line in an area where (i, j) is larger than the line representative threshold value P0.

The updating of the representative distance l (i) for each line is performed by the following equation. l (i) = [l (i) × PI (i) + d (i, j) ×
{P (i, j) -PO + 1}] {PI (i) + P
(I, j) -PO + 1}

In step T6, it is determined whether or not the processing has been completed for all window numbers (areas E) of the line, and steps T3 to T5 are repeated for each area E of the line until this determination becomes YES. If the determination in step T6 is YES, the process proceeds to step T7, where it is determined whether or not the processing has been completed for all the line numbers, and steps T2 to T6 are repeated until this determination becomes YES. If the determination in step T7 is YES, the process proceeds to the next object recognition process (see FIG. 17).

FIG. 17 shows the processing operation of the object recognizing section 20 in the controller 15. The object recognizing section 20 calculates the object based on the representative distance l (i) for each line calculated by the line distance calculating section 26. Recognize O. That is, in step W1, the object number k is set, and in step W2, the object detection distance L (k), the number of effective object points PK (k), and the number of data in the object N (k) are all set to 0, and the primary Reset the archive data set.

In the next step W3, it is determined whether or not valid unregistered line data is registered. If this determination is NO, the flow proceeds to step W8. If the determination in step W3 is YES, in step W4, the front / rear position XD (i) and the horizontal position YD (i) of the line data are set. Thereafter, in step W5, it is determined whether or not the object detection distance L (k) has already been defined.
If this determination is NO, the process proceeds to step W6, where the object detection distance L (k) is set to L (k) = XD (i), and the number of object valid points PK (k) is set to PK (k) = PI
In (i), the number of data in the object N (k) is further expressed as N (k)
= 1 and the primary storage data set is set, and then the process proceeds to step W8.

On the other hand, the determination in step W5 is YES
, The process proceeds to step W7, where the object detection distance L
Let (k) be L (k) = ｛PK (k) × L (i) + P (i)
× XD (i)} / {PK (k) + P (i)} and the number of object effective points PK (k) as PK (k) = PK (k)
+ PI (i), and the number of data in the object N (k) by N
After setting (k) = N (k) +1 and updating the primary storage data set, the process proceeds to step W8.

In step W8, it is determined whether or not the processing has been completed for all the line numbers, and steps W3 to W7 are repeated until the determination becomes YES. Step W8
Is YES, it is determined whether or not the registration of the object O is possible in steps W9 to W11. First, step W
In 9, it is determined whether the number N (k) of data in the object is equal to or greater than a predetermined value. Note that the predetermined value can be variably set so as to decrease as the distance increases. If the determination in step W9 is NO, it is considered that the distance data is due to noise or the like, and the object O is not registered in step W10, the object data of the object number k is initialized, and the process ends. I do. On the other hand, when the determination in step W9 is YES, in step W11 the object O
After registering, the process ends. That is, the object recognition unit 2
A value of 0 registers an object corresponding to the distance of each line as a new object only when the number N (k) of distance data for each line calculated by the line distance calculation unit 26 is equal to or greater than a predetermined value.

After the processing operation in the object recognition unit 20,
A display process for displaying an object on the display device 31 and a warning process for a warning on the warning device 32 are performed.

The above-described series of processing operations are repeatedly performed at predetermined time intervals (one sampling time). As long as the object O once recognized is within a range that can be captured by each CCD chip 11, the object selection unit At 21, capture is continued while performing selection with a new object.

Here, in practice, the object recognition section 20 performs a pre-recognition processing operation prior to the object recognition processing operation. In this pre-recognition processing operation,
It is determined whether or not the number of times that the representative distance for each line is continuously calculated for each line by the line distance calculating unit 26 is larger than a threshold value. In the object recognition process, preparations are made so that an object can be recognized based only on the representative distance. That is, the number of effective points given by the eight adjacent point processing unit 25 is equal to the line representative threshold value P0.
If the distance data is higher and the validity of the distance data is higher, the representative distance is calculated. On the other hand, if the validity of the distance data is lower, the representative distance is not calculated. As the number increases, the representative distance has higher reliability and can be used for object recognition, so that the object can be recognized based only on the representative distance.

FIG. 18 shows a pre-recognition processing operation in the object recognition section 20. First, in step U1, for each line,
After reading the number of times LD_num (i) in which the representative distance is continuously calculated for each sampling time, step U2 is performed.
And the threshold LD_L of the effective number LD_num (i)
Set the limit. This threshold LD_Limit
Is a threshold LD_th (i) set for each line, as shown in FIG. 19, and, for example, as shown in FIG. The threshold value LD_Limit increases proportionally from the side line toward the inner line, that is, toward the far distance side line, or the threshold increases as the line position increases, as shown in FIG. Value LD_L
The limit is set so as to increase step by step.

In the next step U3, the number of valid times LD_num
It is determined whether or not (i) is greater than the threshold LD_Limit. If the determination is YES, the process proceeds to step U4, where the flag Line_Ob_f (i) is set to 1 as data usable for object recognition, and the process proceeds to step U5. move on. On the other hand, if the judgment in the step U3 is NO, the process directly proceeds to the step U5.

In the step U5, it is determined whether or not the processing has been completed for all the lines, and the steps U3 and U4 are repeated until the determination becomes YES. Step U5
Is YES, the process proceeds to the object recognition process. Therefore, in the object recognition process, the flag Li
The object O is recognized based only on the representative distance of the line for which ne_Ob_f (i) is set to 1 (which is valid line data in step W3 in the object recognition processing in FIG. 17).

The distance data for each area E measured by each distance measuring circuit 16 is subjected to smoothing processing by a low-pass filter as a smoothing processing means. Point processing is performed. _{That, d f (i, j)} = a × d f (i,
j) + (1-a) × d (it, after the processing operation of j), the 8 neighbors treated with _d f (i, j) instead of d (i, j) is performed Will be.

The value of a in the smoothing process by the low-pass filter is a constant of 0 to 1, but is set to be smaller (as the line position becomes smaller) for the short distance line as compared to the long distance line. I have. In other words, this low-pass filter is configured so that the degree of smoothing is smaller in the near line as compared to the far line. At this time, when the line position is smaller than the predetermined value, a = a is set so that the smoothing process is not performed at all.
It may be set to 0.

Therefore, in the above embodiment, when the image is captured as luminance information by the left and right rear side detection sensors 10, 10, first, each distance measuring circuit 1 of the controller 15
At 6, the image of each detection sensor 10 is divided in the line direction and the window direction, and the distance d (i, j) is measured for each region E. Next, the distance d (i, j) for each of the measured areas E is calculated by the eight adjacent point processing unit 25.
_{(D f (i, j)} ) and the effective number of points P (i, j) of the distance data for each area E based on the difference dx of the distance between adjacent regions R1~R8 is given, the line distance calculator 26 The representative distance l (i) for each line is calculated based on the number of effective points P (i, j), and the representative distance l for each line is calculated by the line distance calculation unit 26 in the object recognition unit 20.
The number of times (i) is continuously calculated is the threshold LD_Li
When it is larger than mit, the object O is recognized from the representative distance l (i). As described above, the validity of the distance data for each region E is determined as the number of valid points P (i, j) from the relationship with the adjacent regions R1 to R8, and based on the number of valid points P (i, j). Representative distance l for each line
(I) is obtained, and when the effectiveness of the representative distance l (i) is high, the object O is recognized from the representative distance l (i). The influence can be reduced as much as possible, and a highly accurate distance calculation can be performed, so that accurate object recognition can be performed.

The threshold LD_Limit is:
Since the shorter distance line is set smaller than the longer distance line, the object recognition can be performed in a shorter time in the shorter distance line, and when the CCD lines in the CCD chip 11 are equally spaced, When the recognition object O relatively approaches, the object O recognized on the long distance side is not lost on the short distance side. Moreover, the threshold value LD_Limi
Even if t is set to be smaller for the near line, the object O can be accurately recognized on the near side because the distance detection accuracy on the near side is higher than that on the far side. Therefore,
The recognition object O can be reliably followed irrespective of the distance.

Further, the distance data for each area E is subjected to a smoothing process by a low-pass filter, and the degree of smoothing is set to be smaller for a short distance side line than for a long distance side line. The object O can be recognized more accurately without being affected by noise and the like, and the time required for the processing is shorter for the short distance side line, and the recognition object O is prevented from being lost at the short distance side. can do.

Since the object recognizing section 20 is configured to output a signal such as an alarm based on the recognition result of the object O, it is possible to easily notify the occupant of the vehicle C of the recognized object O. Even when there is a high degree of risk that the object O is relatively approaching, the object O can be continuously followed, so that an appropriate warning can be given to the occupant.

[0052]

As described above, according to the first aspect of the present invention, the image of the multi-stage line CCD is divided into a plurality of lines for each line and in the window direction, and the distance is measured for each region. Calculates the representative distance for each line based on the distance data, and recognizes the object based on the representative distance when the number of times that the representative distance is continuously calculated at a predetermined time is greater than a threshold value. In addition, by setting the threshold value to be smaller for the line on the short distance side, even if there is a variation or noise in the distance measurement data, the influence thereof is reduced as much as possible, and accurate object recognition is performed. Can be performed, and the recognition object can be reliably captured irrespective of the distance.

According to the second aspect of the present invention, the smoothing processing means for smoothing the distance data of each area measured by the distance measuring means is provided, and the smoothing degree of the smoothing processing means is set to the long distance side line. By making the line smaller on the short distance side as compared with, the object can be more accurately recognized without losing the recognition object on the short distance side.

According to the third aspect of the present invention, a signal such as a warning is output based on the recognition result of the object, so that the recognized object can be easily known and the object relatively approaches. If the risk is high, the effect can be further enhanced.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of the present invention.

FIG. 2 is a perspective view showing positions of respective components of the object recognition device according to the embodiment of the present invention in a vehicle.

FIG. 3 is a block diagram illustrating a schematic configuration of an object recognition device.

FIG. 4 is a block diagram illustrating a detailed configuration of the object recognition device.

FIG. 5 is a side view showing a concept of measuring a distance to an object by a detection sensor.

FIG. 6 is a diagram showing an image captured by a CCD chip.

FIG. 7 is a diagram showing a concept of dividing a line in an image captured by a CCD chip in a window direction to divide an area.

FIG. 8 is an explanatory diagram showing a state in which images obtained by upper and lower CCD chips are shifted on the same line and parallax occurs.

FIG. 9 is a diagram showing a principle of measuring a distance to an object by upper and lower CCD chips.

FIG. 10 shows C in an image obtained by a CCD chip.
It is a top view which shows the ranging direction of a CD line.

FIG. 11 is a side view showing a range cut area in a vertical direction.

FIG. 12 is a plan view showing a horizontal range cut area.

FIG. 13 is a diagram showing an arrangement of eight adjacent regions adjacent to the region.

FIG. 14 is a diagram showing a specific example from the processing of eight adjacent points to the calculation of the distance for each line.

FIG. 15 is a flowchart showing an eight adjacent point processing operation.

FIG. 16 is a flowchart illustrating a distance calculation processing operation for each line.

FIG. 17 is a flowchart illustrating an object recognition processing operation.

FIG. 18 is a flowchart illustrating a pre-recognition processing operation performed prior to an object recognition processing operation.

FIG. 19 is a flowchart illustrating setting of a threshold.

FIG. 20 is a diagram illustrating a setting example of a threshold.

[Explanation of symbols]

 C vehicle 10 rear side detection sensor 11 CCD chip 15 controller 16 distance measurement circuit (distance measurement means) 20 object recognition section (object recognition means) 25 8 adjacent point processing section 26 line distance calculation section (line distance calculation means) 31 Display device 32 Alarm device E, E (i, j) area R1 to R8 Neighboring area d (i, j) Measurement distance dx Distance difference from adjacent area P (i, j) Number of effective points l (i) Line representative Distance LD_Limit Threshold O Object O 'Object Image

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁶ Identification code FI H04N 7/18 G06F 15/62 380 (72) Inventor Toru Yoshioka 3-1, Fuchu-cho, Shinchi, Aki-gun, Hiroshima Prefecture Mazda Corporation ( 72) Inventor Hiroki Uemura 3-1, Shinchi, Fuchu-cho, Aki-gun, Hiroshima Prefecture Mazda Co., Ltd.

Claims (3)

[Claims] 1. A multi-stage line type CCD comprising a plurality of CCD lines arranged along a window direction and arranged in multiple stages in a line direction orthogonal to the window direction to capture an image as luminance information, An object recognition device for recognizing a specific object from two-dimensional distance data obtained based on a luminance signal from the multi-line CCD, wherein one end of the multi-line CCD in a line direction is near. An image obtained by the multi-stage line type CCD is used as the short distance side line for detecting the distance side, and the other end side is set as the long distance side line for detecting the long distance side.
Distance measuring means for measuring the distance for each area by dividing the area into a plurality of parts for each D line and in the window direction; and calculating the representative distance for each line based on the distance for each area measured by the distance measuring means. An object for recognizing an object based on the representative distance when the number of times the representative distance for each line is continuously calculated for each line by the line distance calculating means is greater than a threshold value; An object recognizing device comprising: a recognition unit; wherein the threshold value is set to be smaller for a short distance side line than for the long distance side line. 2. The object recognition apparatus according to claim 1, further comprising: a smoothing processing unit for performing a smoothing process on the distance data of each area measured by the distance measuring unit; The object recognition device characterized in that the distance is set to be smaller on the short distance side line than on the long distance side line. 3. The object recognition device according to claim 1, wherein the object recognition means is configured to output a signal such as a warning based on a recognition result of the object. .