JP2003281700A - Cutting-in vehicle detecting device and method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a cutting-in vehicle detecting device obtaining a traveling track of other vehicles cutting into an own vehicle traveling/stopping lane or an area on which traveling of the own vehicle is assumed from a time series image obtained from an image inputting device mounted on the vehicle by tracking a vehicular potential area, and detecting the cutting-in vehicle from the traveling track thereof. <P>SOLUTION: This cutting-in vehicle detecting device determining the cutting-in vehicle, is composed of an image inputting portion 1 taking in the time series image obtained by an image input device mounted on the vehicle by a single eye, a processing are setting portion 2 setting a reference border line for detecting cutting-in from the time series image, a potential area setting portion 3 setting the vehicular potential area on the outside to the reference border line, a traveling track assuming portion 4 assuming a traveling track of other vehicles by the vehicular potential area having a moving direction special to a specified cutting-in with reference to the reference border line, and a cutting-in determining portion 5 determining the cutting-in vehicle from the other vehicles when the traveling track of the other vehicle has a specified special feature of cutting-in. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention uses an image obtained from a TV camera attached to a moving body represented by a vehicle such as an automobile to drive a lane in which the vehicle is traveling or a traveling of the vehicle. The present invention relates to an image processing device and a method for detecting a vehicle that is interrupting an area in which is predicted.

[0002]

2. Description of the Related Art "Interruption vehicle" means the traveling of the own vehicle,
Alternatively, it is a vehicle entering a stopped lane or an area where the own vehicle is predicted to travel. Since it is a phenomenon that occurs suddenly at a short distance from the own vehicle, it is extremely dangerous and needs to be detected quickly.

As means for detecting a lane in which the own vehicle is running or a vehicle that is expected to run in the area where the own vehicle is expected to travel, there are roughly proposed a method using a radar and a method using an image. ing.

As a system using a radar, Japanese Patent Laid-Open No. 7-8
In 9367, a method of detecting using a laser radar is proposed. The method using the laser radar has a problem that it is vulnerable to bad weather such as rain. In the case of an interrupting vehicle, the laser radar irradiates it at an angle, so it cannot receive reflected waves, has a strong directivity, and is difficult to use on curved roads. There is the problem of interfering with the lasers used.

On the other hand, in the method using the image, for example, the traveling lane of the own vehicle can be detected.
It is possible to limit the area that is dangerous to the traveling of the host vehicle. In this way, various kinds of information can be acquired from the image. Also, the device can be constructed at low cost. Therefore, instead of using a radar as a sensor, an image is used.

The method of using this image includes a method of using a plurality of TV cameras and a method of using an image taken by a single eye by a single TV camera.

As a system using a plurality of TV cameras, in Japanese Patent Laid-Open No. 11-213295, the distance to a vehicle is stereoscopically measured to obtain a movement vector of another vehicle in a three-dimensional scene, and its extension line and traveling are used. A method of detecting an interrupted vehicle when a lane intersects has been proposed. This stereo vision method is
There is a problem that the detection is delayed because the detection cannot be performed until the two TV cameras enter the common field of view. In addition, since the parallax is obtained from the left and right images by matching each other's points, there is a problem that the amount of calculation becomes large. Therefore, as the camera, it is preferable to use a single camera rather than a plurality of cameras.

The following two methods are known as methods using an image photographed with a single eye by a single TV camera.

A first method is disclosed in Japanese Patent Laid-Open No. 8-2495.
Japanese Unexamined Patent Publication No. 97 proposes a method of detecting an interrupting vehicle by paying attention to a change in the number of edges on the road surface. In this method, it is assumed that an interrupting vehicle exists when the maximum value of the number of edge points inside a certain area set in advance exceeds a certain threshold value. In this method, since the number of edges is detected, the number of edges is increased if there is a pattern on the road surface even if there is no vehicle, and therefore the number of edges may be erroneously detected.

The second method is Japanese Patent No. 307506.
Japanese Unexamined Patent Publication No. 4 (1994) proposes a method of first detecting a vehicle and determining that the detected vehicle is an interrupting vehicle when the detected vehicle sufficiently overlaps a traveling lane area. This method requires accurate vehicle detection, but it is difficult to detect the vehicle accurately.

[0011]

As described above, in the second method using a single TV camera, accurate detection of a vehicle is necessary to detect an interrupting vehicle.

However, accurate detection of this vehicle is very difficult.

Therefore, in the present invention, in consideration of the above points, the vehicle is not accurately detected, but is interrupted in the lane in which the vehicle is traveling or in the area where the vehicle is predicted to travel. It is an object of the present invention to provide an image processing apparatus and a method for obtaining a traveling locus of another vehicle coming in by tracing a vehicle candidate area and detecting an interrupting vehicle from the traveling locus.

[0014]

According to a first aspect of the present invention, there is provided image acquisition means for capturing a time-series image obtained by, for example, monocular photography by an image input device mounted in the vehicle, and the vehicle of the vehicle. A reference boundary that sets a reference boundary line of the range of the dangerous area with respect to the time-series image as a dangerous area that is a range in which the vehicle is in danger when another vehicle interrupts The line setting means and the vehicle candidate regions that are all or some of the regions estimated to be other vehicles are set as the vehicle tracking start region that is the region outside the dangerous region with the reference boundary line as a boundary. Candidate region setting means, traveling locus estimation means for estimating a traveling locus of the vehicle candidate region based on the time-series image, and the estimated traveling locus of the vehicle candidate region is a predetermined interrupt condition with respect to the reference boundary line. When holding is an interrupt vehicle detecting apparatus characterized by comprising, an interrupt vehicle determining unit that determines the vehicle candidate region as an interrupt vehicle.

According to a second aspect of the present invention, the interrupting vehicle determining means determines that an interrupting condition that the traveling locus of the vehicle candidate area intersects with the reference boundary line or an extension line of the traveling locus of the vehicle candidate area is the reference. 2. The interrupt according to claim 1, wherein the vehicle candidate region is determined to be an interrupting vehicle when at least one of the interrupting conditions of intersecting the boundary line with a predetermined angle range is satisfied. It is a vehicle detection device.

According to a third aspect of the present invention, the candidate area setting means obtains an edge intensity distribution in a specific direction by arranging an edge detection filter in the vehicle tracking start area, and determines a peak position of the obtained edge intensity distribution. The edge position is set to the specific direction, and when the edge detection filter is arranged in the vehicle tracking start area, the edge detection is performed as the distance from the vanishing point, which is the infinity point on the image, increases. Edge detection means for enlarging the filter application area and the vehicle candidate area is set at the edge position, and the size of the vehicle candidate area is determined according to the distance from the vanishing point or the distance between the reference boundary lines. 3. An interrupting vehicle detection device according to claim 1 or 2, further comprising: a vehicle candidate area setting means for setting a size.

According to a fourth aspect of the present invention, the traveling locus tracking means estimates the traveling locus only in a vehicle candidate region having a motion direction peculiar to a predetermined interruption with respect to the reference boundary line. The interrupting vehicle detection device according to any one of claims 1 to 3.

According to a fifth aspect of the present invention, there is provided an image acquisition step of capturing a time-series image captured by, for example, a single eye by an image input device mounted on the host vehicle, and a traveling range of the host vehicle. And, a reference boundary line setting step of setting a reference boundary line of the range of the dangerous area with respect to the time-series image, with a range in which the own vehicle is dangerous when another vehicle comes in, as a dangerous area, The vehicle candidate area, which is all or part of the area estimated to be the vehicle,
A candidate area setting step of setting a vehicle tracking start area, which is an area outside the dangerous area with the reference boundary as a boundary, and a travel trajectory estimation for estimating a travel trajectory of the vehicle candidate area based on the time-series image. An interrupting vehicle determining step of determining the vehicle candidate area as an interrupting vehicle when the estimated traveling path of the vehicle candidate area has a predetermined interrupting condition with respect to the reference boundary line;
The method for detecting an interrupting vehicle is characterized by comprising:

According to a sixth aspect of the present invention, there is provided an image acquisition function of, for example, a time-series image captured and acquired by a single eye by an image input device mounted on the host vehicle, and a traveling range of the host vehicle. , And a reference boundary line setting function for setting a reference boundary line of the range of the dangerous area with respect to the time-series image, with a range in which the own vehicle is dangerous when another vehicle comes in All or a vehicle candidate area that is estimated to be a vehicle, a candidate area setting function that sets a vehicle tracking start area that is an area outside the dangerous area with the reference boundary as a boundary, A traveling locus estimating function for estimating a traveling locus of the vehicle candidate region based on the time-series image; and a case where the estimated traveling locus of the vehicle candidate region has a predetermined interruption condition with respect to the reference boundary line, A program interrupt vehicle detecting method characterized by the determining interrupt vehicle determination function both candidate regions as an interrupt vehicle, the realized by a computer.

The present invention sets a vehicle area candidate from a time-series image acquired by a single eye by an image input device mounted on the vehicle, and tracks the vehicle area candidate by tracking the vehicle area candidate. It is calculated to determine whether the vehicle is an interrupting vehicle.

By using the interruption condition peculiar to the interruption vehicle at the time of tracking, stable tracking can be realized and the processing time can be shortened.

Further, an interrupting vehicle can be detected by tracking the vehicle using a characteristic peculiar to the interrupting vehicle without recognizing the vehicle.

[0023]

DESCRIPTION OF THE PREFERRED EMBODIMENTS In the present embodiment, attention is paid to the fact that another vehicle has a movement to enter a region where the own vehicle is traveling or a region where the own vehicle is predicted to be traveling. In other words, the vehicle that is interrupting the vehicle travels on the image with respect to the boundary line of the area where the vehicle is traveling, the area where the vehicle is stopped, or the area where the vehicle is scheduled to travel. If the trajectories are intersecting or if there is a movement that intersects, the presence of the interrupting vehicle and the position on the image are detected. The traveling locus of the vehicle on the image is calculated by tracking the vehicle candidate area.

[0024]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment for specifically realizing the embodiment of the present invention will be described below with reference to FIGS.

FIG. 1 shows a basic configuration example of an interrupting vehicle detection device of this embodiment.

As shown in FIG. 1, the interrupting vehicle detection device is composed of an image input section 1, a processing area setting section 2, a candidate area setting section 3, a running locus estimating section 4 and an interrupt determining section 5, and each of these sections. The functions 1 to 5 are realized by the program stored in the computer.

A time-series image captured by a single eye is acquired by the image input unit 1 which is a device (for example, one TV camera) capable of acquiring an image around the vehicle attached to the vehicle.

From the time-series image captured by this monocular, in order to determine whether or not another vehicle is interrupting, a region (hereinafter referred to as a dangerous region) that is dangerous for the own vehicle to travel is set in the image as a processing region. The area is set by the unit 2.

Further, a tracking start area is set in the image in order to detect a candidate for an interrupting vehicle.

The "interrupting vehicle" is a vehicle that enters the area where the host vehicle is traveling, the area where the host vehicle is stopped, or the area where the host vehicle is predicted to travel. Say.

(A) Processing Area Setting Unit 2 As shown in FIG. 2, the processing area setting unit 2 comprises a traveling lane detection unit 21 and a reference boundary line setting unit 22.

(A-1) Traveling lane detection unit 21 In order to determine that an interrupting vehicle is an interrupt when the vehicle crosses the boundary line of the traveling lane, the lane in which the vehicle is currently traveling is detected. That is, since the host vehicle is traveling on the road, the white line is used as the boundary line of the traveling lane using the result of the white line detection. The host vehicle may be stopped.

The intersection of the two detected white lines is taken as the vanishing point.

The "vanishing point" is the point at infinity on the image. Japanese Patent Laid-Open No. 7-89 discloses white line detection.
Various methods such as the method according to 367 have been proposed, and any method may be used.

(A-2) Reference boundary line setting section 22 The interrupting vehicle comes in from the outside of the traveling lane at a short distance in front of the host vehicle.

Therefore, as shown in FIG. 6, the area sandwiched between the traveling lanes up to a certain distance from the vehicle is set as the dangerous area.

An area outside the dangerous area is set as a "vehicle tracking start area".

Further, the left and right boundary lines of the dangerous area are set as "reference boundary lines".

The width of the vehicle tracking start area at this time is set so that the width becomes smaller as it approaches the vanishing point.

(B) Candidate Area Setting Section 3 As shown in FIG. 3, the candidate area setting section 3 comprises an edge detecting section 31 and a vehicle candidate area setting section 32.

By the way, a vehicle has many vertical and horizontal line segments. Therefore, there is a high possibility that the region with many edges having vertical and horizontal directions is the vehicle region. Therefore, the edge in the vertical direction or the horizontal direction is noticed, the edge is detected, and the area having a high density is set as the “vehicle candidate area”.

(B-1) Edge detection section 31 Since the horizontal and vertical edge components are required to set the vehicle candidate area, the edge detection section 31 performs edge detection.

In the environment in which the vehicle is running, environmental changes such as rain, shadows, and lighting fluctuations may occur drastically.
Edge detection using local information of an image like an obel filter is easily affected by noise, and a desired result is often not obtained.

Therefore, "Edge Extraction"
n Method Based on Separab
ility of Image Features, IE
ICE TRANS. INF. & SYS. , VO
L. E78-D, NO. 12DECEMBER 199
The edge detection using the separability filter proposed in 5 "is used.

This edge detection method is not easily affected by noise and can be obtained without depending on the brightness difference of the feature points. In FIG. 7, the degree of separation η is an amount that represents the difference between internal information (for example, brightness) in the regions 1 and 2 by discriminant analysis, and the edge detection using this degree of separation uses the following formula. To process.

[0046]

[Equation 1]

[Equation 2]

[Equation 3] Where η₁Is the number of images in region 1, η_TwoIs the territory
Number of images in area 2, σ_TIs the variance of the entire region, P₁Is
Area 1 average brightness value, P_TwoIs the average brightness value of region 2, P
_mIs the average brightness value of the entire region, P_iIs the position in the image
The brightness value of the position i. The degree of separation η takes a value in the range of 0 to 1.

This degree of separation η is taken as the edge strength of the boundary line between regions 1 and 2. Further, the areas 1 and 2 are combined to be an application area of the separability filter.

Various techniques have been proposed for edge detection, and the edge detection technique is not limited to the edge detection technique using the separability filter, and any edge detection technique can be used.

Edge strength for obtaining a separation degree by arranging a separation degree filter as shown in FIG. 8A for the vehicle tracking start area set by the processing area setting unit 2 and for obtaining the center position of the filter. And

The size of the separability filter is set so that the width becomes smaller as it approaches the vanishing point.

In the case of horizontal edges, separability filters of the type shown in FIG. 9 are densely arranged toward the vanishing point to find horizontal edges. Furthermore, the peak of the edge intensity was detected toward the vanishing point, and the peak position was set as the edge position.
For example, as shown in FIG. 8B, the horizontal axis represents the horizontal edge strength,
When there is a graph in which the vertical axis represents a position, a position having a peak value equal to or greater than a threshold value in the position direction is detected and the position is set as a position where a horizontal edge exists.

In the case of a vertical edge, the same process as in the case of a horizontal edge is performed using a separability filter of the type shown in FIG.

(B-2) Vehicle candidate area setting unit 32 Since the vehicle includes many vertical and horizontal line segments, the vehicle candidate area is set around the edge position detected by the edge detecting unit 31.

As shown in FIG. 11, it is assumed that an edge is detected at the position of the thick line in the vehicle tracking start area. That is, in the case of a horizontal edge, an area above the edge position on the image is set, and in the case of a vertical edge, the area with the edge position as the lateral center is set as the vehicle candidate area.

At this time, the number of vehicle candidate areas is not limited to one vehicle, but a plurality of vehicle candidate areas can be set.

Here, the size of the vehicle candidate area is set so that the width becomes smaller as it approaches the vanishing point.

(C) Traveling locus estimating unit 4 As shown in FIG. 4, the traveling locus estimating unit 4 is composed of a candidate area tracking unit 41 and a traveling locus calculating unit 42.

(C-1) Candidate Area Tracking Unit 41 As shown in FIG. 5, the candidate area tracking unit 41 comprises a matching unit 411 and an interrupting motion tracking unit 412.

(C-1-1) Matching Unit 411 The matching unit 411 performs template matching on a vehicle candidate region with an image between frames. Template matching is a process based on the idea that when two images are given, it is possible to determine whether or not they are the same by superimposing the two and seeing the difference.

At this time, the difference between the two images can be represented by various scales. Commonly used scales are sums of residual absolute values of luminance values (SAD), normalized correlation coefficients (CC) between images, and “Object Search U”.
sing Orientation Code Mat
ching, IAPR Workshop on Ma
chine Vision Application
(MVA2000), NOV. 28-30, 2000 "
There is a direction coding method (OCM) proposed in.

The matching method is not limited to this method, and any matching method may be used depending on the apparatus and the required accuracy.

The matching section 411 outputs where the vehicle candidate area has moved, as coordinate values on the image.

(C-1-2) Interruptive Motion Detection Unit 412 The interrupted motion detection unit 412 tracks only vehicle candidate regions having an interrupted motion, and in other cases, that is, the following (i) matching is performed. If it fails, (i
i) and (iii) If the vehicle does not have a direction specific to a predetermined interruption, the tracking is stopped and the vehicle candidate area is discarded. In this way, the unnecessary vehicle candidate areas are discarded because the number of vehicle candidate areas to be tracked is reduced and the calculation load is lightened.

First, it is assumed that another vehicle cuts in from the right side.

(I) When Matching Fails The matching of vehicle candidate areas in the matching unit 411 is not always correct. Therefore, the reliability of matching is evaluated using the value of the residual sum of absolute values in the vehicle candidate area, which is obtained from the processing of the matching unit 411.

That is, if the value of the residual sum of absolute values is smaller than a certain value, the matching is successful, and if it is larger, the matching is unsuccessful.

If the matching is unsuccessful, the tracking is terminated and the vehicle candidate area is discarded.

(Ii) In the case of not having a predetermined interruption-specific direction between short frames When the matching of the vehicle candidate areas in the matching unit 411 has succeeded, the motion direction of the vehicle candidate areas on the image is obtained. .

As shown in FIG. 12, assuming that the vertical direction of the image is 0 degrees, when the motion direction θ of the vehicle candidate region between short frames (for example, one frame) satisfies the following equation, an interrupt occurs. Suppose it has a specific direction and tracking is successful. Otherwise, it is assumed that the tracking has failed.

Θ ₁ = <θ = <θ ₂ where θ ₁ is the vertical upward angle in the image, and θ ₂ is the angle parallel to the reference boundary line.

If the tracking has failed, the tracking is terminated and the vehicle candidate area is discarded. By limiting the tracking direction, a more stable tracking result can be obtained.

If such processing is performed for each frame, the movement history of the vehicle candidate area can be known.

(Iii) When there is no direction specific to a predetermined interrupt between long frames The tracking direction is limited not only between short frames but also between long frames.

When the moving distance of the vehicle candidate area is short, the resolution in the tracking direction is small because of pixel quantization, but when the moving distance is long after tracking for a long frame, the resolution becomes large. Therefore, when information of frames of a certain length is stored for a certain vehicle candidate area, a tracking direction vector is calculated from the coordinates of the point where tracking is started and the current position coordinates, and as shown in FIG. Assuming that the vertical upward direction of the image is 0 degree, when the angle θ of the tracking direction vector satisfies the following expression, it has a direction peculiar to interrupt, and tracking is successful. Otherwise, it is assumed that the tracking has failed. When the tracking has failed, the tracking is terminated and the vehicle candidate area is discarded. By limiting the tracking direction between long frames, a more stable tracking result can be obtained.

Θ ₁ = <θ = <θ ₂ However, assuming that the vertical upward angle in the image is 0 °, θ ₁ is an angle parallel to the reference boundary line, and is the direction toward the vanishing point. Is. θ ₂ is an angle that is parallel to the reference boundary line and is on the opposite side of θ ₁ . The range of this angle changes depending on the accuracy of the tracking direction vector, and if it is correct, the range of angles may be further limited. The same process is performed when another vehicle comes in from the left side.

By performing the processing as described above, a tracking direction vector having a motion direction peculiar to a predetermined interrupting vehicle can be stably obtained during a long frame. Further, the vehicle candidate area output from the candidate area setting unit 3 for each frame increases, but the increase is suppressed by discarding the vehicle candidate area for which tracking has failed,
The amount of processing can be reduced.

(C-2) Traveling locus calculation unit 42 The traveling locus calculation unit 42 calculates a traveling locus from the history of the tracking direction vector of the vehicle candidate area obtained by the interruption motion detection unit 412.

Specifically, as shown in FIG. 14, the running locus is calculated from two points, that is, the position coordinates at which the tracking is started on the image of the first frame and the position coordinates of the vehicle candidate area on the image of the Nth frame. calculate.

Further, the running locus may be calculated by the least square method or the like using many points from all the histories between frames.

(D) Interruption determination unit 5 As described above, the vehicle candidate regions for which the travel locus is calculated by the travel locus calculation unit 42 still have a high probability of an interruption vehicle. Therefore, the interruption determination unit 5 uses the traveling locus of the vehicle candidate region calculated by the traveling locus calculation unit 42 and the reference boundary line set by the reference boundary line setting unit 22 to more surely determine whether or not it is an interruption. judge.
Two kinds of judgment conditions are used.

(I) When the running trajectory of the vehicle intersects with the reference boundary line on the input time series image (ii) The extension line of the traveling trajectory of the vehicle becomes the reference boundary line on the input time series image On the other hand, the interrupting vehicle can be detected by using one of the conditions when the vehicle intersects with a predetermined angle range.

First, assume that another vehicle is cutting in from the right side.

The predetermined angle range of (ii) is, as shown in FIG. 15, the angle θ _C between the reference boundary line and the extension line of the traveling locus, which is judged to be an interruption, and the condition of the following equation. Is satisfied.

Θ> = θ _{c Since} this θ _C changes depending on the installation position of the camera, etc., it must be determined appropriately.

The same process is performed when another vehicle comes in from the left side.

In the case of (1), since the vehicle is in an intersecting state, it is determined that the vehicle is interrupted when the vehicle starts entering the reference area.

In the case of (ii), the determination can be made at an early stage before entering the dangerous area of the vehicle.

The angle which is judged to be an interrupt is
Since it changes depending on the installation position of the camera, etc., it needs to be determined appropriately. If either of these two determination conditions is satisfied, it is determined that the vehicle is an interruption vehicle. Further, the position of the interrupting vehicle on the image is the intersection of the traveling locus or extension line of another vehicle and the reference boundary line.

<Modification 1> The present invention is not limited to the contents described in the above embodiments.

For example, in the traveling lane detector 21,
A method other than the method of detecting the boundary line of the lane in which the vehicle is traveling may be used.

For example, the same process may be performed by obtaining the boundary line of the region where the traveling of the own vehicle is predicted from the velocity of the own vehicle and the steering angle of the steering wheel, and using it as the traveling lane.

<Modification 2> In the interrupt determination section 5,
The condition for determining whether or not it is an interrupt is not limited to the method on the input time-series image, and another method may be used.

For example, after the image is converted into a back projection image which is an image of a viewpoint seen from directly above the road, (i) the reference locus of the vehicle running trajectory on the back projection image is set on the back projection image. If it intersects with the line (ii), the extension condition of the running trajectory of the vehicle on the back projection image intersects with the reference boundary line set on the back projection image with a predetermined angle range as a determination condition. good.

As described above, the present invention can be variously expanded by replacing partial element processing.

[0095]

According to the present invention, it is possible to stably detect an interrupting vehicle from a time series image acquired by an image input device mounted on the vehicle, and to realize a technique such as device control based on image information. Make a big contribution.

[Brief description of drawings]

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

FIG. 2 is a block diagram showing an embodiment of a processing area setting unit in FIG.

FIG. 3 is a block diagram showing an embodiment of a candidate area setting unit in FIG.

FIG. 4 is a block diagram showing an embodiment of the traveling locus estimation unit in FIG.

5 is a block diagram showing an embodiment of the candidate region tracking unit of FIG.

FIG. 6 is an explanatory diagram of a method of setting a processing area.

FIG. 7 is an explanatory diagram of a separability filter.

FIG. 8 is an explanatory diagram of a method of performing edge detection using a separability filter.

FIG. 9 is an explanatory diagram of a filter in edge detection.

FIG. 10 is an explanatory diagram of a filter in edge detection.

FIG. 11 is an explanatory diagram of a method of setting a candidate area.

FIG. 12 is an explanatory diagram of a method of limiting a tracking direction.

FIG. 13 is an explanatory diagram of a method of limiting a tracking direction.

FIG. 14 is an explanatory diagram of a method of calculating a traveling locus.

FIG. 15 is an explanatory diagram of interrupt determination conditions.

[Explanation of symbols]

1 Image input section 2 Processing area setting section 3 Candidate area setting section 4 Running locus estimation unit 5 Interrupt judgment section 21 Running lane detector 22 Standard boundary setting section 31 Edge detector 32 Vehicle candidate area setting unit 41 Candidate area tracking unit 42 Traveling locus calculator 411 Matching unit 412 Interrupt motion detector

   ─────────────────────────────────────────────────── ─── Continued front page    F term (reference) 5B057 AA16 AA19 BA02 BA24 DA07                       DA15 DB02 DC16 DC19 DC36                 5H180 AA01 CC04 CC24 LL01 LL02                       LL04                 5L096 BA02 BA04 CA04 CA24 DA03                       FA06 FA69 HA04

Claims (6)

[Claims] 1. An image acquisition means for capturing a time-series image captured and acquired by an image input device mounted on the host vehicle; and a range within which the host vehicle travels and another vehicle interrupts the vehicle. And a reference boundary line setting means for setting a reference boundary line of the range of the dangerous area with respect to the time-series image, with the area where the own vehicle is dangerous as a dangerous area, all estimated to be other vehicles, or A candidate area setting unit that sets a vehicle candidate area that is a partial area to a vehicle tracking start area that is an area outside the dangerous area with the reference boundary line as a boundary; A traveling locus estimating means for estimating based on a time-series image; and when the estimated traveling locus of the vehicle candidate region has a predetermined interruption condition with respect to the reference boundary line, the vehicle candidate region is regarded as an interruption vehicle. An interrupting vehicle detection device comprising: an interrupting vehicle determining unit. 2. The interrupting vehicle determining means is an interrupting condition that a running locus of the vehicle candidate area intersects with the reference boundary line, or an extension line of the running locus of the vehicle candidate area is a predetermined boundary line with the reference boundary line. 2. The interrupting vehicle detection device according to claim 1, wherein the vehicle candidate area is determined to be an interrupting vehicle when at least one of the interrupting conditions of intersecting with an angular range is satisfied. 3. The candidate area setting means obtains an edge intensity distribution in a specific direction by arranging an edge detection filter in the vehicle tracking start area, and determines a peak position of the obtained edge intensity distribution as an edge in the specific direction. When the edge detection filter is arranged in the vehicle tracking start area, the area to which the edge detection filter is applied increases as the distance from the vanishing point, which is the point at infinity on the image, increases. An edge detecting means for enlarging, and setting the vehicle candidate area at the edge position,
3. A vehicle candidate area setting means for setting the size of the vehicle candidate area to a size according to the distance from the vanishing point or the distance between the reference boundary lines. The interrupted vehicle detection device described. 4. The running locus tracking means estimates the running locus only in a vehicle candidate region having a motion direction peculiar to a predetermined interruption with respect to the reference boundary line. The interrupted vehicle detection device according to any one of the above. 5. An interrupted vehicle detection method for detecting an interrupted vehicle entering an area in which the own vehicle is traveling, an area in which the own vehicle is stopped, or an area in which the own vehicle is predicted to travel. In the image acquisition step of capturing a time-series image captured and acquired by a single eye by the image input device mounted on the own vehicle, the traveling range of the own vehicle, and another vehicle interrupts And a reference boundary line setting step of setting a reference boundary line indicating the range of the dangerous area on the time-series image, with the area where the own vehicle is dangerous as a dangerous area, all estimated to be other vehicles, or A candidate area setting step of setting a vehicle candidate area, which is a partial area, as a vehicle tracking start area that is an area outside the dangerous area with the reference boundary line as a boundary; A traveling locus estimation step of estimating a traveling locus of a region based on the time-series image; and, when the estimated traveling locus of the vehicle candidate region has a predetermined interruption condition with respect to the reference boundary line, An interrupted vehicle detection method comprising: an interrupted vehicle determination step of determining an interrupted vehicle. 6. An interrupted vehicle detection method for detecting an interrupted vehicle entering an area in which the own vehicle is traveling, an area in which the own vehicle is stopped, or an area in which the own vehicle is predicted to travel. In a program realized by a computer, an image acquisition function for capturing a time-series image captured and acquired by a single eye by the image input device mounted on the own vehicle, and a traveling range of the own vehicle, and A reference boundary line setting function that sets a reference boundary line indicating the range of the dangerous area on the time-series image as a dangerous area when the own vehicle is in danger when another vehicle comes in A candidate for setting a vehicle candidate area, which is all or a part of the specified area, as a vehicle tracking start area which is an area outside the dangerous area with the reference boundary as a boundary. Area setting function, a traveling locus estimation function for estimating a traveling locus of the vehicle candidate area based on the time-series image, and the estimated traveling locus of the vehicle candidate area has a predetermined interruption condition with respect to the reference boundary line. A program for a method for detecting an interrupted vehicle, which has an interrupting vehicle determination function for determining the vehicle candidate area as an interrupted vehicle when the vehicle has.