JP2003288691A - Intrusion prediction device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an intrusion prediction device for precisely predicting an intrusion. <P>SOLUTION: A millimeter wave radar 10 capable of measuring vehicle-to-vehicle distance and relative speeds and an image sensor 20 are loaded on a vehicle and connected to a computer 30. Within the computer device 30, a peripheral vehicle measurement part 37 for measuring a physical quantity such as position and speed of a peripheral vehicle, a peripheral vehicle recognition part 38 for recognizing the size and type of the peripheral vehicle, and a prediction part 39 for predicting the intrusion based on the measurement result and recognition result are formed. When the type of the peripheral vehicle is a large vehicle, the vehicle-to-vehicle distance with the preceding vehicle is large, and the relative speed is large, for example, it is determined whether the measurement value of the peripheral vehicle measurement part 37 and the recognition result of the peripheral vehicle recognition part 38 correspond to a specified pattern or not, and an intrusion probability value according to the pattern is outputted. Otherwise, a pattern matched to the measurement value and the recognition result is selected, and the probability value (past data) possessed by the pattern is outputted. Since the type and size of vehicle are taken into consideration for prediction, the precision of prediction is improved more than in the past. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an interrupt predicting device for predicting an interrupt ahead of a surrounding vehicle while a vehicle is running. In particular, the present invention relates to an interrupt prediction device that accurately predicts the possibility of an interrupt by recognizing not only the position and speed of the surrounding vehicle but also the vehicle type of the surrounding vehicle and determining whether or not the result is an interrupt pattern. The present invention can be applied to an active cruise device and an interrupt warning device.

[0002]

2. Description of the Related Art Conventionally, there is a device for predicting an interruption of a side vehicle to the front of the own vehicle while the vehicle is traveling. For example, the vehicle traveling control device disclosed in Japanese Patent Laid-Open No. 5-217099 is one example. This is a device in which each sensor such as a radar device that measures vehicles in front of right, in front of center, and in front of left is mounted on the host vehicle to calculate an inter-vehicle distance and a relative speed with respect to each front vehicle. Then, an interruption is predicted based on the result, the throttle actuator is controlled, and the inter-vehicle distance to the preceding vehicle is increased. That is, in the active cruise control, the interruption is predicted from the situation of the vehicle ahead and the interruption is prepared for, so that the sense of security during traveling is improved.

Another example is the vehicle running control device disclosed in Japanese Patent Laid-Open No. 2001-171389. This is to mount an image processing device on the own vehicle, and image the preceding vehicle traveling on the adjacent traveling road by the device, and change its image, for example, the amount of approach of the preceding vehicle to the own vehicle traveling road, operation of the blinker, It is a device that determines the possibility of changing the lane of an adjacent preceding vehicle by detecting a decrease in the width of the adjacent traveling road. The purpose of the present invention is to optimize the travel control for the interruption of the adjacent vehicle, that is, to correct the drive shaft torque, acceleration / deceleration, inter-vehicle distance, acceleration limit value, inter-vehicle time, and the like.

[0004]

[Patent Document 1] Japanese Unexamined Patent Publication No.
-217099 gazette discloses a vehicle travel control device,
A device for predicting the possibility of interruption of an adjacent front vehicle from the state of the front vehicle, the operation of a side vehicle or a rear vehicle,
It is not a device that predicts the interruptability of a side vehicle and the rear vehicle from the situation. That is, it is effective for predicting the interruption of the vehicle ahead, but not effective for predicting the interruption of other vehicles. In other words, it is not a device that predicts the probability of interruption faster. From the experience, the interruption of the left and right front vehicles may depend on the vehicle type. For example, a large vehicle has a high driver's seat, and the inter-vehicle distance in front of the vehicles adjacent to the left and right can be clearly understood. Therefore, it is said that there is a high probability that a large vehicle will be interrupted. In particular, large vehicles followed by sports vehicles are said to have a high probability of interruption. In addition, in the case of an up lane, interruption of a large vehicle is sufficiently predicted even if there is no vehicle ahead. In the above-mentioned conventional example, the size and vehicle type of the surrounding vehicles are not taken into consideration, so that there is a problem that it is not possible to deal with such cases accurately.

Further, the vehicle running control device disclosed in Japanese Unexamined Patent Publication No. 2001-171389 images a vehicle in front of left or in front of right, and changes in the image (lateral movement), blinker operation, and adjacent traveling. By detecting a decrease in road width, etc., the possibility of changing the lane of an adjacent preceding vehicle is sought. This method of detecting the lateral movement of the device has a problem that it cannot be detected until the movement for interruption is actually started. That is, the response to the interruption is delayed, and sudden braking may be applied in the active cruise (vehicle distance control) mode.

The present invention has been made in order to solve the above-mentioned problems, and its purpose is not only to measure the position and speed of a peripheral vehicle but also to recognize the vehicle type thereof, and by using them, the peripheral vehicle makes an interrupt movement. Predict interrupts before entering. Also, judge by the interrupt pattern including the vehicle type,
It is to predict it quickly. Then, this device is mounted in various vehicle-mounted systems, for example, an active cruise system, to further improve running stability and driving safety.

[0007]

According to another aspect of the present invention, there is provided an interrupt predicting device, which measures at least a preceding vehicle and a preceding vehicle measuring means for measuring an inter-vehicle distance to the preceding vehicle, an image sensor for picking up an image around the own vehicle, and an image. Peripheral vehicle measuring means for detecting the peripheral vehicle from the output of the sensor and measuring at least the position and speed of the peripheral vehicle, and peripheral vehicle recognizing means for recognizing the vehicle type of the peripheral vehicle and / or the size of the vehicle from the output of the image sensor. And a prediction means for predicting an interruption of the surrounding vehicle based on the measurement result of the preceding vehicle measuring means, the measurement result of the surrounding vehicle measuring means, and the recognition result of the surrounding vehicle recognizing means.

An interrupt predicting device according to a second aspect of the present invention is the interrupt predicting device according to the first aspect, wherein the predicting means has an interrupt generation pattern, the measurement result of the preceding vehicle measuring means, and the surrounding vehicles. It is determined whether the measurement result of the measuring unit and the recognition result of the surrounding vehicle recognizing unit correspond to the interrupt occurrence pattern, and the interrupt of the surrounding vehicle is output as a probability value using both measurement results and the recognition result as parameters. Characterize.

The interrupt predicting device according to claim 3 is the interrupt predicting device according to claim 2, wherein the items of the pattern are at least the following distance from the preceding vehicle, the vehicle type, the speed, the number and the order of the surrounding vehicles. It is characterized by consisting of.
Further, the interrupt predicting device according to claim 4 is the interrupt predicting device according to claim 2 or 3, wherein the predicting means has a large inter-vehicle distance from the preceding vehicle and / or has a small vehicle speed. The feature is that the larger the probability value is, the larger the certain value is.

An interrupt predicting device according to a fifth aspect of the invention is the interrupt predicting device according to any one of the second to fourth aspects, in which the predicting means determines an inter-vehicle distance between the host vehicle and the preceding vehicle. The probability value is set to be larger as the inter-vehicle distance between the surrounding vehicles is larger and smaller. Further, the interrupt predicting device according to claim 6 is the interrupt predicting device according to any one of claims 2 to 5, wherein the predicting means calculates the probability value when the surrounding vehicle is a large vehicle. It is characterized by being large.

Further, the interrupt predicting device according to claim 7 is the interrupt predicting device according to claim 6, wherein the determination of a large vehicle is performed by detecting the vehicle height or the vehicle width of the surrounding vehicle extracted from the output of the image sensor. It is characterized by determining from. Further, the interruption predicting device according to claim 8 is the interruption predicting device according to claim 6, wherein the determination of a large vehicle traveling in an adjacent lane is performed by a predetermined portion of a peripheral vehicle extracted from the output of the image sensor. It is characterized in that it is determined from the azimuth angle and the distance between the surrounding vehicle and the own vehicle. Further, the interruption predicting device according to claim 9 is the interruption predicting device according to claim 6, wherein the determination of a large vehicle traveling in an adjacent lane is performed by determining the azimuth angle of each predetermined part of the peripheral vehicle obtained from the output of the image sensor. , And the distance between the two in the vehicle width direction.

[0012]

In the interrupt predicting apparatus according to the first aspect of the present invention, the preceding vehicle measuring means measures at least the inter-vehicle distance to the preceding vehicle and the vehicle speed thereof. The preceding vehicle measuring means is, for example, a millimeter wave radar device, a laser radar device, or the like. Further, the image sensor captures an image around the vehicle. This image sensor is provided, for example, on the side of the own vehicle and / or on the rear side. The peripheral vehicle measuring means extracts the peripheral vehicle from the output of the image sensor, and measures at least the position and speed of the peripheral vehicle. These are calculated from, for example, continuous images such as optical flow for every predetermined time. The surrounding vehicle measuring means can measure other information such as blinking of the blinker, the amount of approach to the own vehicle, and the approach speed, if necessary. The peripheral vehicle measuring means includes means for measuring all the factors related to the peripheral vehicle.

Then, the peripheral vehicle recognizing means recognizes the vehicle type of the peripheral vehicle and / or the size of the vehicle from the output of the image sensor. These are contour shapes in image processing,
Recognize by measuring the image of vehicle height, vehicle width, etc. Then, the predicting unit predicts the interruption of the surrounding vehicle ahead of the host vehicle based on the measurement result of the preceding vehicle measuring unit, the measurement result of the surrounding vehicle measuring unit, and the recognition result of the surrounding vehicle recognizing unit.

For example, if the inter-vehicle distance to the preceding vehicle is large, the peripheral vehicle is a large vehicle, the speed thereof is high, and the inter-vehicle distance between the peripheral vehicles is small, an interruption occurs. Predict. That is, if the predetermined condition is satisfied, 100
% Interruption is predicted and, for example, control is performed so as to reduce the speed of the host vehicle. Or warn the driver. As a result, before the driver visually recognizes the interruption, it is possible to give advance notice and control the speed. Therefore, if the present invention is applied to, for example, an active cruise system or the like, it is possible to further improve traveling safety.

An interrupt predicting device according to a second aspect of the invention is the interrupt predicting device according to the first aspect, wherein the predicting means has a (past) interrupt generation pattern. This pattern is, for example, a pattern in which the inter-vehicle distance to the preceding vehicle is sufficiently large, and a plurality of peripheral vehicles travel in adjacent lanes at a shorter inter-vehicle distance and faster than the host vehicle. Further, the inter-vehicle distance to the preceding vehicle is sufficiently large, and the following large vehicle is a pattern in which the turn signal is turned on when the inter-vehicle distance to the own vehicle is short. Then, it is determined whether or not the measurement result of the preceding vehicle measuring means, the measurement result of the surrounding vehicle measuring means, and the recognition result of the surrounding vehicle recognizing means correspond to a predetermined pattern. When it corresponds, the interruption of the surrounding vehicle to the own vehicle is output with a probability value using both the measurement results and the recognition result as parameters.

The probability value is, for example, the inter-vehicle distance to the preceding vehicle is D1, the own vehicle speed is V1, and the inter-vehicle distance to the surrounding vehicles is D.
4. Where v is the relative speed of the surrounding vehicles, it is calculated by the equation (1).

## EQU00001 ## P = A.times.D1 + B.times. (100-V1) -C.times.D4 + E.times.v (1) where A, B, C and E are coefficients. This is because the probability value increases as the inter-vehicle distance D1 with respect to the preceding vehicle and the relative speed v of the surrounding vehicle increase, and the inter-vehicle distance D4 of the surrounding vehicle decreases and the speed V1 of the own vehicle increases. The smaller the value, the greater the probability of interruption. According to the present invention, the interruption is predicted by this probability value before the surrounding vehicles start the interruption operation. If the speed of the host vehicle is suppressed based on the probability value, the speed can be gently suppressed. For example, when the probability is 50%, the speed is reduced to 50% of the speed control amount (deceleration amount). That is, the control is more comfortable than ever before.

The interrupt predicting device according to claim 3 is the interrupt predicting device according to claim 2, wherein the items of the pattern are at least the following distance from the preceding vehicle, the vehicle type, the speed, the number and the order of the surrounding vehicles. It is characterized by consisting of. When the surrounding vehicle interrupts, the distance between the own vehicle and the preceding vehicle needs to be a predetermined distance or more. Further, when the peripheral vehicle interrupts, the speed of the peripheral vehicle needs to be higher than the speed of the own vehicle. In addition, since acceleration of a large vehicle is insufficient in the up lane or the like, there is an instruction to change the lane from the overtaking lane to the traveling lane. This is particularly noticeable when there is an ordinary car behind. Therefore, if the inter-vehicle distance from the preceding vehicle, the vehicle type, the speed, the number of vehicles and the order of the surrounding vehicles are used as parameters, the interruption can be predicted accurately.

The interrupt predicting device according to claim 4 is the interrupt predicting device according to claim 2 or 3, wherein the predicting means has a large inter-vehicle distance to the preceding vehicle and / or the own vehicle speed. The smaller is, the larger the probability value is. The larger the inter-vehicle distance between the host vehicle and the preceding vehicle, the easier the surrounding vehicles are to interrupt. Also, the lower the own vehicle speed, the easier the surrounding vehicles are to interrupt. Therefore, the larger the inter-vehicle distance to the preceding vehicle and / or the smaller the vehicle speed, the larger the probability value. By doing so, the interrupt can be predicted more accurately.

An interrupt predicting device according to a fifth aspect of the invention is the interrupt predicting device according to any one of the second to fourth aspects, in which the predicting means determines the inter-vehicle distance between the host vehicle and the preceding vehicle. The probability value is larger as the inter-vehicle distance between the surrounding vehicles is larger. Even in such a case, the surrounding vehicle easily interrupts ahead of the own vehicle. Therefore, also in this case, if the interrupt probability is increased, the interrupt can be predicted more accurately.

An interrupt predicting device according to claim 6 is the interrupt predicting device according to any one of claims 2 to 5, wherein the predicting means interrupts when the surrounding vehicle is a large vehicle. The probability value of is large. Large vehicles are required to drive in the driving lane for safety.
In addition, since acceleration of a large vehicle is insufficient in the up lane or the like, there is an instruction to change the lane from the overtaking lane to the traveling lane.
In addition, since the driver's seat of a large vehicle is located at a high position, it is easy to grasp a vacant area in front and it is easy to interrupt. Therefore,
Even if the surrounding vehicle is a large vehicle, if the probability value of interruption is increased, the interruption can be predicted more accurately and more reliably.

Further, the interrupt predicting device according to claim 7 is the interrupt predicting device according to claim 6, wherein the determination of a large vehicle is performed by determining the vehicle height or vehicle width of the surrounding vehicle extracted from the output of the image sensor. It is judged from. For example, the inter-vehicle distance between the host vehicle and its surrounding vehicles is calculated from the intersection coordinates of the line extending in the horizontal direction on the image from the wheels of the extracted peripheral image and the lane boundary lane. Then, by multiplying the vehicle height on the image obtained by the image sensor by (inter-vehicle distance) / (focal length), almost the actual vehicle height is calculated. Therefore, a large vehicle can be determined from this vehicle height value. Alternatively, the vehicle width is calculated in the same manner. With this configuration, it can be easily determined whether the vehicle is a large vehicle.

Further, the interrupt predicting device according to claim 8 is the interrupt predicting device according to claim 6, wherein the determination of a large vehicle traveling in an adjacent vehicle is performed by determining a predetermined value of a peripheral vehicle extracted from the output of the image sensor. It is determined from the azimuth angle to each part and the distance between the surrounding vehicle and the host vehicle. At this time, the azimuth angle of the adjacent lane to the peripheral vehicle is the azimuth angle to the corner of the peripheral vehicle, and the separation distance is the shortest distance between the host vehicle and the peripheral vehicle. This shortest distance is roughly calculated, for example, by measuring the vertical length of the license plate in front of the vehicle. This is because the license plate has a specified length, and the length detected on the image has the relationship of Expression (2).

[Equation 2]       Vertical length ≒ length on image x shortest distance / focal length (2)

Therefore, when the shortest distance is D2, the azimuth angle to the front corner is θ ₀ , and the azimuth angle to the rear corner is θ ₁ as shown in FIGS. 8 and 9, the vehicle length L is expressed by Calculated in (3).

[Equation 3] L≈L ₁ −L ₀ = D 2 · (cos θ ₀ tan θ ₁ −sin θ ₁ ) ... (3) Then, the calculated vehicle length L is a predetermined value or more, for example, 5
If it is m or more, it is determined to be a large vehicle. With this determination, it is possible to accurately determine a large vehicle simply by capturing an image of the surrounding vehicle and extracting the license plate and the vehicle corner.

The interrupt predicting device according to claim 9 is the interrupt predicting device according to claim 6, wherein the determination of a large vehicle traveling in an adjacent vehicle is determined by a predetermined portion of a peripheral vehicle obtained from the output of the image sensor. It is determined from the azimuth angle to and the distance between the two in the vehicle width direction. If the side and / or the rear is imaged by the image sensor, the lane (white line) indicating the boundary between the surrounding vehicle and the traveling lane is imaged. Therefore, by measuring the distance in the vehicle width direction between the wheel of the peripheral vehicle and this lane extracted by the image sensor and the distance between the own vehicle and the lane, the distance between the two in the vehicle width direction is specified. In addition, the azimuth angles of the front corner portion and the rear corner portion of the surrounding vehicle are respectively specified. The position of the front corner portion is specified by the distance in the vehicle width direction and the azimuth angle of the front corner portion. Similarly, the position of the rear corner portion is specified by the separation distance and the azimuth angle of the rear corner portion. That is, the positions of the front corner and the rear corner are specified. The vehicle length is calculated by subtracting the positions of the front corner and the rear corner. Then, a vehicle having a predetermined value or more, for example, 5 m or more is determined to be a large vehicle. Even with such a calculation, it is possible to accurately determine a large vehicle.

[0025]

DESCRIPTION OF THE PREFERRED EMBODIMENTS (Embodiment) FIG. 1 shows an interrupt predicting apparatus of the present invention. The figure is a system configuration diagram. The interruption predicting apparatus of the present embodiment is installed at a millimeter wave radar 10 which is a preceding vehicle measuring means capable of detecting a relative speed with respect to a preceding vehicle and an inter-vehicle distance, a side of the own vehicle, and / or a rear side, and images the surroundings of the own vehicle. The image sensor 20 and the computer device 30. This computer device 30
A peripheral vehicle measuring unit 37 that is a peripheral vehicle measuring unit that measures a peripheral vehicle based on the output results of the millimeter wave radar 10 and the image sensor 20, and a peripheral vehicle recognizing unit 3 that recognizes the peripheral vehicle.
8, and a prediction unit 39 that is a prediction unit that predicts an interruption of a peripheral vehicle based on the results of the millimeter wave radar 10, the peripheral vehicle measurement unit 37, and the peripheral vehicle recognition unit 38. The computer device 30 is, for example, a one-chip type microprocessor, as shown in FIG. 2, a ROM 31 in which an arithmetic processing program is written, a CPU 32 for executing the program, and a RAM as a work area memory at the time of execution.
33 and an I / O interface 34 for inputting / outputting with the outside. The millimeter wave radar 10 is mounted on the front part of the own vehicle shown in FIG. 3, and the image sensor 20 is installed on the side and the rear part of the own vehicle as shown in the figure.

Next, the operation of the interrupt predicting device of this embodiment will be described with reference to the flowchart of FIG. This embodiment is an example of predicting an interruption of a peripheral vehicle during follow-up control and outputting the interruption to the driver and the vehicle speed control device (FIG. 1). The interrupt predicting apparatus of this embodiment has an engine key O (not shown).
Starting with step s100 at N. First, in step s100, the presence of a preceding vehicle is confirmed by the millimeter wave radar 10. If it does not exist, that is, if it is no, it waits in step s100 until it is detected. If so, step s
The process proceeds to 102, and the inter-vehicle distance D1 with the preceding vehicle is detected. Next, the process proceeds to step s104, and it is determined whether the follow-up control is currently being performed. If no, the process returns to step s100. If yes, the process moves to step s106, and the image sensor 20 captures the right rear image. This image sensor 20
Then, as shown in FIG. 5, for example, a large vehicle or an ordinary vehicle is imaged on the rear right side.

Then, the process proceeds to step s108. In step s108, the number of surrounding vehicles is obtained from the output image. The number of units is calculated by, for example, labeling processing in image processing. Then, it is determined whether the number is two or more.
If it is no, that is, if it is 0 or 1, the probability of interrupting is low, so the process returns to step s100. If yes, the pattern has a high probability of interruption (FIG. 7), and the process proceeds to step s110. This pattern is a pattern in which the inter-vehicle distance D1 with the preceding vehicle is large and the inter-vehicle distance D4 between the surrounding vehicles is small.

Therefore, in step s110, the inter-vehicle distance D4 between the two is calculated. This is because the smaller the inter-vehicle distance D4, the greater the probability of interruption. The inter-vehicle distance D4 is calculated by calculating the frames W and W'which specify the peripheral vehicles from the output of the peripheral image (FIG. 6), and the frame W,
It can be calculated from the difference between the y coordinates of the origins (x ₀ , y ₀ ) and (x ′ ₀ , y ′ ₀ ) of W ′. This is because the y coordinate of each point on the lane S in FIG. 6 has a correlation with the distance from the host vehicle. Then, the process proceeds to step s112.

At step s112, the relative speed v of the surrounding vehicle is obtained from the image output. This is because the greater the relative speed v of the surrounding vehicles, the greater the probability of interruption. The relative speed v is calculated, for example, by extracting the surrounding vehicle in the frame W ₁ in FIG. 6 and moving the origin (x ₀ , y ₀ ) in the image after a predetermined time. As a result, the relative speed v of the surrounding vehicles is calculated. Also, the speed V1 of the host vehicle is obtained from a vehicle speed sensor (not shown). This is because the lower the speed of the host vehicle, the greater the probability that the surrounding vehicles will interrupt. Then, the process proceeds to step s114.

In step s114, step s116
The parameters for determining the vehicle type are calculated. The parameter is, for example, as shown in FIG. 8, the shortest distance D2 which is the distance from the host vehicle when the surrounding vehicle is imaged in the right rear direction.
Is. Further, as shown in FIG. 9, the azimuth angles θ ₀ and θ ₁ to the corners G and H of the surrounding vehicle are shown. The shortest distance D2 is
For example, it is roughly calculated by measuring the vertical length of the license plate in front of the surrounding vehicle. Since the license plate has a specified length, there is a relation of Expression (2) with the length detected on the image. Therefore, from equation (2), the shortest distance D
2 is easily backcalculated. Next, from the image output, as shown in FIG. 9, the corners G (x ₀ , y ₀ ) and H of the own vehicle to the surrounding vehicles are obtained.
Azimuth angles θ ₀ and θ ₁ to (x ′ ₀ , y ′ ₀ ) are calculated.
In FIG. 9, if the shortest distance D2 and the azimuth angles θ ₀ and θ ₁ are known, the vehicle length L of the surrounding vehicle is calculated according to the equation (3). Note that the symbol D in FIGS. 8 and 9 indicates the distance between the two vehicles in the vehicle width direction, and is used in a modification described later.

Then, the process proceeds to step s116, and if the vehicle length L is a predetermined value or more, for example, 5M or more, the vehicle type is recognized as a large vehicle. Similarly, the following vehicle is also recognized. Then, in step s118, it is determined whether the preceding vehicle type is a large vehicle and the following vehicle is a normal vehicle. This pattern is most likely to cause an interrupt. If no, the process returns to step s100. In the case of yes, it transfers to step s120. In step s120, the probability value is calculated by, for example, Expression (1) as described above. Then, this probability value P is shown in FIG.
To the vehicle speed control device. As a result, the speed is moderately suppressed by the interrupt prediction device before the peripheral vehicle starts the interrupt operation. That is, the sense of security of traveling is further improved. Therefore, if the interrupt prediction device of the present embodiment is mounted on an active cruise system, an obstacle warning device, etc., it is possible to further improve the sense of security during traveling.

(Second Embodiment) The first embodiment is an example of an interrupt predicting device for predicting an interrupt when a peripheral vehicle exists on the right side. The second embodiment is an example of predicting the interruption when there is a following vehicle immediately behind the own vehicle. The system configuration of the interrupt predicting device is the same as that of the first embodiment, and the difference is that the image sensor 20 at the rear of the vehicle is operated and the interrupt of the rear vehicle is predicted based on the image output. The operation of the interrupt predicting device of the second embodiment will be described with reference to the flowchart of FIG.

First, in step s200, the presence of a preceding vehicle is confirmed by the millimeter wave radar 10. If it does not exist, that is, if it is no, it waits in step s200 until it is detected. If it exists, the process proceeds to step s202 and the inter-vehicle distance D1 with the preceding vehicle is detected. Next, step s2
The process proceeds to 04, and it is determined whether the follow-up control is currently being performed. no
If so, the process returns to step s200. On the other hand, if yes, the process proceeds to step s206, and the image sensor 20 images the rear of the vehicle.

Then, in step s208, it is determined from the output image whether or not there is a rear vehicle. If it does not exist, that is, if it is no, the possibility of an interrupt from the rear is 0%, so the process returns to step s200. ye
If s, the process proceeds to step s210. in this case,
As shown in FIG. 11, the image sensor 20 at the rear of the vehicle captures an image of the vehicle, for example, behind. That is, as shown in FIG. 12, the preceding vehicle is present at an inter-vehicle distance D1 in the front, and the vehicle is present at an inter-vehicle distance D3 (<D1) in the rear. This is one pattern in which a rear vehicle may change lanes and make an interruption. Especially, if the rear vehicle is a large vehicle, the possibility is great. Therefore, it is determined in the next steps s210 and s212 whether the rear vehicle is a large vehicle. This determination is performed as follows, for example.

In general, the image sensor 20 installed in the rear part
When the image of the rear is taken, the rear vehicle is obtained in the shape shown in FIG. Here (in step s210), the inter-vehicle distance D3 to the rear vehicle is calculated from this image output. The inter-vehicle distance D3 is roughly calculated, for example, by measuring the length or width of the license plate N at the front of the rear vehicle. Since the license plate N has a specified length, the length detected on the image has the relationship of Expression (2). Therefore,
The inter-vehicle distance D3 is easily calculated from the equation (2).

Next, in step s212, as shown in FIG.
As shown in, the frame W surrounding the rear vehicle is set by image processing, and the coordinates of the origin G (x ₀ , y ₀ ) and the corner I (x ₁ , y ₁ ) on the diagonal are calculated. As a result, the vehicle width and vehicle height of the rear vehicle are calculated. The vehicle width and vehicle height are calculated by the equation (4).

(4) Vehicle width = (x ₁ −x ₀ ) × D3 / focal length Vehicle height = (y ₁ −y ₀ ) × D3 / focal length (4) Therefore, the vehicle height is equal to or more than a predetermined value, for example, If it is 2 m or more, it is determined to be a large vehicle. If no, the possibility is judged to be small and the process returns to step s200. If yes, that is, if the vehicle is a large vehicle, the possibility is high, so step s2
Move to 14.

In step s214, the turn-on of the winker 40 of the rear vehicle is detected from the image, and step s216
In particular, it is determined whether or not the turn signal 40 is turned on to the right for changing lanes. If the blinker 40 is not lit or the left blinker 40 is lit, the possibility of an immediate interruption is small, so the process temporarily returns to step s200. If yes, the speed V1 of the host vehicle is detected in step s218 in preparation for the interruption, and finally the process proceeds to step s220.

In step s220, the probability value is output according to equation (5).

P = A × D1 + B × (100−V1) (5) where A and B are coefficients. This is a conversion formula in which the greater the inter-vehicle distance D1 from the preceding vehicle and the smaller the speed V1 of the host vehicle, the greater the probability of interrupting the rear vehicle. Then, this probability value P is output to the vehicle speed control device of FIG. As a result, the rear vehicle causes the blinker 40 to blink, and at the same time, the speed is suppressed by the interrupt prediction device. Immediately responding to an interruption from the rear will further improve the sense of security of driving. Therefore, if the interrupt prediction device of the present embodiment is mounted on an active cruise system, an obstacle warning device, etc., it is possible to further improve the sense of security of traveling.

(Modification) Although one embodiment representing the present invention has been described above, various other modifications are possible. For example, in the first embodiment, the azimuth angle to each part of the peripheral vehicle and the shortest distance D2 are used for determining the large vehicle, but the distance D in the vehicle width direction between the host vehicle and the peripheral vehicle may be used. Good. For example, in FIG. 8, if the image sensor 20 images the side,
The lane S (white line) indicating the surrounding vehicles and the traveling lane is imaged. Therefore, by measuring the distance between the extracted peripheral vehicle, for example, the wheel and the lane S, and the distance between the host vehicle and the lane S, the distance D between the two in the vehicle width direction is calculated. Therefore, the vehicle length L can be roughly calculated by using the separation distance D and the azimuth angles θ ₀ and θ ₁ of the front corner portion and the rear corner portion of the surrounding vehicle in FIG. That is, the separation distance D may be used instead of the shortest distance D2. Equivalent results are obtained.

Although the probability value is calculated in the first and second embodiments, the calculation is not limited to this. For example, past patterns, that is, probability values corresponding to past interrupt parameters may be stored in advance in a table format as shown in FIG. The left side of the table is the items, and the right side is the range that each item can take. For example, the inter-vehicle distance from the preceding vehicle is D1 _{j to} D1 _j ′. Then, it is possible to select the table T _{j in} which all the measured values and the recognized values obtained at the time of prediction fall within the set range and output the probability value P _j . That is, it may be determined whether or not the obtained measurement value and recognition value correspond to a predetermined pattern as in the above-described example, or conversely, select the measurement value and recognition pattern that are obtained. , The probability value of the pattern may be output. Either is fine. In short, any device may be used as long as it is a device that predicts an interruption by using a pattern including a vehicle type. The method of using the pattern does not matter.

In the above embodiment, the computer device 30 is used.
Although an example in which a vehicle speed control device is provided at the output stage to assist the follow-up control is shown, an output device may be provided with an alarm device for giving a voice warning instead. If constituted in this way, it can be set as an interruption warning device.

[Brief description of drawings]

FIG. 1 is a system configuration diagram of an interrupt prediction device according to a first embodiment.

FIG. 2 is a block diagram of a computer device that constitutes the interrupt prediction device according to the first embodiment.

FIG. 3 is a layout diagram of a millimeter wave radar and an image sensor of the interrupt prediction device according to the first embodiment.

FIG. 4 is a flowchart of an example showing an operation of the interrupt prediction device according to the first embodiment.

FIG. 5 is a diagram of a plurality of surrounding vehicles captured by the image sensor according to the first embodiment.

FIG. 6 is a positional relationship diagram of a plurality of peripheral vehicles captured by the image sensor according to the first embodiment.

FIG. 7 is a pattern diagram of an example in which an interrupt is likely to occur according to the first embodiment.

FIG. 8 is a positional relationship diagram of the own vehicle and surrounding vehicles when the interruption predicting device according to the first embodiment operates.

FIG. 9 is an explanatory diagram for calculating a vehicle length of a peripheral vehicle according to the first embodiment.

FIG. 10 is an exemplary flowchart showing an operation of the interrupt prediction device according to the second embodiment.

FIG. 11 is a rear vehicle view imaged by the image sensor according to the second embodiment.

FIG. 12 is a pattern diagram of an example in which the possibility of interrupt occurrence is high according to the second embodiment.

FIG. 13 is an explanatory diagram of a rear vehicle size determination of the interrupt prediction device according to the second embodiment.

FIG. 14 is a table having interrupt parameters for outputting a comparative probability value according to a modification.

[Explanation of symbols] 10 ... Millimeter wave radar 20 ... Image sensor 30 ... Computer device 37 ... Peripheral vehicle measurement unit 38 ... Peripheral vehicle recognition unit 39 ... Predictor 40 ... blinker S ... lane W, W '... frame D1 ... Distance between preceding vehicle D2 ... Shortest distance from surrounding vehicles D3: Distance between vehicles behind D4 ... Distance between vehicles

─────────────────────────────────────────────────── ─── Continuation of front page (51) Int.Cl. ⁷ Identification code FI theme code (reference) B60R 21/00 627 B60R 21/00 627 628 628E F02D 29/02 301 F02D 29/02 301D (72) Inventor Yuji Uchiyama 1 in 41, Yokosui, Nagakute-cho, Aichi-gun, Aichi Prefecture Toyota Central Research Institute Co., Ltd. (72) Inventor Shinsei Shiraki 1 in 41, Nagakute-cho, Nagakute, Aichi-gun, Aichi Prefecture Toyota Central Research Co., Ltd. In-house F-term (reference) 3D044 AA25 AB01 AC26 AC31 AC59 AD01 AE01 AE04 3G093 BA23 CB10 DB05 DB16 EA01 FA02 FA11 FB01 FB02 5H180 AA01 CC04 CC14 LL04

Claims (9)

[Claims] 1. An interrupt predicting device for predicting an interrupt of a surrounding vehicle to the front of the own vehicle when the vehicle is traveling, and a preceding vehicle measuring means for measuring at least an inter-vehicle distance to the preceding vehicle and a vehicle speed of the preceding vehicle. An image sensor that images the surroundings of the own vehicle; a peripheral vehicle measuring unit that detects the peripheral vehicle from the output of the image sensor and measures at least the position and speed of the peripheral vehicle; and the peripheral vehicle from the output of the image sensor. Vehicle type, and /
Alternatively, based on the peripheral vehicle recognition means for recognizing the size of the vehicle, the measurement result of the preceding vehicle measurement means, the measurement result of the peripheral vehicle measurement means, and the recognition result of the peripheral vehicle recognition means, An interrupt predicting apparatus comprising: a predicting unit that predicts an interrupt ahead of the host vehicle. 2. The predicting means has an interruption occurrence pattern, and the measurement result of the preceding vehicle measuring means, the measurement result of the surrounding vehicle measuring means, and the recognition result of the surrounding vehicle recognizing means are the interruption generating pattern. The interrupt predicting device according to claim 1, wherein the interrupt predicting device outputs the interrupt occurrence as a probability value using the both measurement results and the recognition result as parameters. 3. The items of the pattern include at least the inter-vehicle distance from the preceding vehicle, the vehicle type and speed of the surrounding vehicles,
The interrupt prediction device according to claim 2, wherein the interrupt prediction device comprises the number and the order. 4. The predicting means sets the probability value larger as the inter-vehicle distance from the preceding vehicle is larger and / or the own vehicle speed is smaller. The interrupt prediction device according to Item 3. 5. The prediction means sets the probability value to be larger as the inter-vehicle distance between the own vehicle and the preceding vehicle is larger and the inter-vehicle distance between the surrounding vehicles is smaller. The interrupt prediction device according to any one of claims 2 to 4, which is characterized. 6. The interrupt predicting device according to claim 2, wherein the predicting means increases the probability value when the surrounding vehicle is a large vehicle. . 7. The interrupt prediction device according to claim 6, wherein the determination of the large vehicle is performed based on a vehicle height or a vehicle width of the surrounding vehicle extracted from the output of the image sensor. 8. The determination of the large vehicle traveling in an adjacent lane is made based on the azimuth angle of the peripheral vehicle to each predetermined portion extracted from the output of the image sensor and the distance from the host vehicle. The interrupt predicting device according to claim 6, wherein the interrupt predicting device is a device. 9. The determination of the large vehicle traveling in the adjacent lane is made based on the azimuth angle to the predetermined parts of the surrounding vehicle extracted from the output of the image sensor and the distance between the two in the vehicle width direction. The interrupt predicting device according to claim 6.