JP2018005441A - Inter-vehicle distance alarm and collision alarm device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an inter-vehicle distance alarm device and a collision alarm device that shorten a time required for information processing, prevent rear-end collision accidents, and notify a driver of a collision with obstacles by an alarm to allow the collision to be obviated.SOLUTION: An inter-vehicle alarm device and a collision alarm device comprise imaging means 1, image recognition means 2, inter-vehicle distance determination means 3, stationary object determination means 4 and moving object determination means 5. The image recognition means 2 is configured to replace an image of an object of a prescribed width in upper and lower directions from a center point of an image signal with hue, saturation and luminance, and give the object a sequential number in order from the center point to a horizontal direction by the hue and saturation to identify the object. The inter-vehicle distance determination means 3, the stationary object determination means 4 and the moving object determination means 5 are configured to: make a measurement from the image for the identified object; calculate a distance between a self-vehicle 30 and the identified object in order; detect a speed of the self-vehicle 30 to calculate a time until the self-vehicle 30 reaches the object; and when an arrival time falls within a prescribed time, set off an alarm.SELECTED DRAWING: Figure 1

Description

The present invention relates to a collision warning device that prevents a collision between an inter-vehicle distance warning device, a preceding vehicle, and an obstacle ahead when the vehicle is running.

  Many vehicle collision warning devices for vehicles that collide with an object that may collide with the host vehicle when driving an automobile calculate the possibility of collision based on image processing of an in-vehicle camera, and issue a warning ( For example, see Patent Document 1.)

In this case, as shown in FIG. 19, in the vehicle collision warning device 101, the imaging unit 102 captures an image around the vehicle, and the motion amount calculating unit 103 detects the motion amount from the image captured by the imaging unit 102. The information extraction unit 104 extracts an image region including an object with a possibility of collision as a region of interest according to the calculation result of the motion amount calculation unit 103, and the collision time calculation unit 105 The collision time is calculated, and the notification / control unit 106 notifies the fact that there is a possibility of collision according to the collision time calculated by the collision time calculation unit 105, or controls the vehicle so as to avoid the collision.

  As shown in FIG. 20, the motion amount calculation unit 103 extracts vertical and horizontal edge images from the image captured by the imaging unit 102 by the horizontal edge detection unit 111 and the vertical edge detection unit 112, and calculates the horizontal edge speed. The means 113 and the vertical edge speed calculating means 114 calculate the amount of motion of the edge image in the vertical direction and the horizontal direction. In addition, the information extraction unit 104 extracts an image region including an object with a possibility of collision as an attention region by the attention region setting unit 115 according to the calculation result of the motion amount calculation unit 103, and the collision time calculation region setting unit 116. The collision time calculation area is set in step S1 and is transmitted to the collision time calculation means 105.

  At this time, since the vertical and horizontal edge images are extracted from the image captured by the imaging means 102 and the amount of motion of the vertical and horizontal edge images is calculated, it is necessary to process the entire image. In addition, the amount of processing of the image is large, it takes time for processing, and the processing apparatus is increased in size.

JP 2006-88940 A

  Therefore, the present invention is not based on a large device, the information processing time is short, the driver is informed of the shortage of the inter-vehicle distance by an alarm, and the rear-end collision is prevented, and the collision with the obstacle is notified by an alarm. It is an object of the present invention to provide an inter-vehicle distance warning and a collision warning device that can avoid a collision in advance.

In order to solve the above-mentioned problem, the present invention of claim 1 is characterized in that the inter-vehicle distance warning and the collision warning device have imaging means for imaging a vehicle ahead, a stationary object or a moving object,
Image recognition means for recognizing an image of a front vehicle, a stationary object or a moving object that is an object imaged by the imaging means;
Having inter-vehicle distance determining means for determining the inter-vehicle distance with the preceding vehicle recognized by the image recognition means;
A stationary object judging means for judging a stationary object in front of the image taken by the imaging means;
A moving object judging means for judging a moving object ahead imaged by the imaging means;
The image recognizing unit exchanges the image of the object having a predetermined width in the vertical direction from the center point of the image captured by the image capturing unit with hue, saturation, and luminance, and sequentially from the center point to the horizontal direction according to the hue and saturation. Identify and identify the objects with serial numbers,
The inter-vehicle distance determining means, the stationary object determining means, and the moving object determining means measure the identified object from the image, and sequentially calculate the distance from the object identified as the own vehicle,
The speed of the own vehicle is detected from the speed data of the own vehicle, the time for the own vehicle to reach the target is calculated, and if the arrival time is within a predetermined time, an inter-vehicle distance warning and a collision warning are issued. Device.

In the first aspect of the present invention, the inter-vehicle distance alarm and the collision alarm device include an imaging unit that images a vehicle, a stationary object, or a moving object in front of the object, and the front object that is the object imaged by the imaging unit. Image recognition means for recognizing an image of a vehicle, a stationary object, or a moving object. For this reason, an image of the front is captured by the imaging unit, the vehicle, the stationary object or the moving object in front of the object is recognized, and an image identifying the vehicle and the stationary object or the moving object is sent to the determination unit. Can do.

Since there is an inter-vehicle distance determining means for determining the inter-vehicle distance with the preceding vehicle recognized by the image recognizing means, an alarm can be issued when the vehicle approaches the preceding vehicle.
Since there is a stationary object determination unit that determines a stationary object in front captured by the imaging unit, an alarm can be issued when there is a possibility of collision with a stationary object in front.
Since it has the moving object judgment means which determines the moving object ahead imaged with the imaging means, when there exists a possibility of colliding with a moving object ahead, an alarm can be issued.

The image recognizing unit performs a process for determination on an image of an object having a predetermined width in the vertical direction from the traveling center point of the image captured by the image capturing unit. For this reason, in order to determine an image of a horizontal target object having a predetermined width in the vertical direction from the center point of the image signal, it is possible to reduce the amount of information to be determined, increase the processing speed, and quickly perform the determination process. It is possible to issue an alarm quickly and to make the alarm device small and inexpensive.

The object is exchanged for hue, saturation, and luminance, and the object is identified by sequentially assigning a sequential number in the horizontal direction from the center point according to the hue and saturation. For this reason, in the judgment of an object, the influence from a light source can be reduced and an object can be reliably recognized even in bad weather or at night. Further, it is possible to reliably identify a plurality of objects and perform a determination process.

The inter-vehicle distance determining unit, the stationary object determining unit, and the moving object determining unit measure the identified object from the image, and sequentially calculate the distance from the object identified as the own vehicle. For this reason, it is possible to calculate the distance to the object and issue a warning quickly only by image recognition, and the warning device can be made small and inexpensive.

The speed of the host vehicle is detected from the speed data of the host vehicle, the time for the host vehicle to reach the object is calculated, and if the arrival time is within a predetermined time, an alarm is issued. For this reason, it is possible to accurately issue a warning necessary for collision prevention according to the speed of the own vehicle. When the arrival time exceeds a predetermined time, the vehicle can travel while maintaining the braking and stopping distance of the subject vehicle. The predetermined time is a time during which it is possible to take measures necessary for collision avoidance, and is, for example, about 2.7 seconds.

According to the second aspect of the present invention, the inter-vehicle distance determination means regards the vehicle width of the front vehicle located at the traveling center point of the image as 1.5 m, and measures the viewing angles at both ends of the vehicle width of the front vehicle from the image. Then, the distance to the vehicle ahead is calculated, the time for the own vehicle to reach the vehicle ahead is calculated, and if the arrival time is within a predetermined time, an inter-vehicle distance warning and a collision warning are issued. Device.

In the present invention of claim 2, the inter-vehicle distance determination means regards the vehicle width of the front vehicle located at the traveling center point of the image as 1.5 m, and measures the viewing angles at both ends of the vehicle width of the front vehicle from the image. Then, the distance to the preceding vehicle is calculated, the time for the host vehicle to reach the preceding vehicle is calculated, and if the arrival time is within a predetermined time, an alarm is issued. For this reason, since it is not necessary to measure both ends of the vehicle ahead, it is possible to quickly and easily measure the distance between the vehicle ahead and the determination time can be shortened. Since the vehicle width of the vehicle ahead is regarded as 1.5 m, even in a large vehicle, the calculation of the inter-vehicle distance is not made shorter than actual, and collision avoidance can be sufficiently handled.

According to a third aspect of the present invention, the inter-vehicle distance determining means measures the angle of the horizontal plane between the position where the tire of the vehicle in front located at the center point of the image contacts the road surface and the imaging means, and picks up the image. An inter-vehicle distance warning and a collision warning device for calculating a distance from a vehicle ahead based on the height of the means from the road surface.

In the present invention of claim 3, the inter-vehicle distance determining means measures the angle of the horizontal plane between the position where the tire of the front vehicle located at the center point of the image is in contact with the road surface and the imaging means from the image. The distance from the vehicle ahead is calculated based on the road surface of the means. For this reason, the vehicle ahead can be reliably recognized even in a narrow vehicle such as a motorcycle, and the determination time can be shortened. Since the calculation is based on the height of the vehicle ahead from the road surface, it is possible to quickly calculate the distance from the vehicle ahead and issue an alarm.

According to a fourth aspect of the present invention, the stationary object judging means measures the change in the viewing angle between the center point of the image and the stationary object, detects the speed of the own vehicle from the speed data of the own vehicle, and The distance to the traveling center line of the own vehicle is calculated from the traveling distance of the own vehicle and the change in the viewing angle between the traveling center point of the image and the stationary object, and the distance is 1 m. If it is within the range, the distance between the stationary object and the own vehicle is calculated, and the speed of the own vehicle is detected from the speed data of the own vehicle to reach the stationary object. When the arrival time is within a predetermined time, the inter-vehicle distance warning and the collision warning device issue a warning.

In the present invention of claim 4, the stationary object judging means measures a change in the viewing angle between the moving center point of the image and the stationary object, detects the speed of the own vehicle from the speed data of the own vehicle, and The travel distance of the vehicle is calculated, and the distance to the travel center line of the host vehicle is calculated from the travel distance of the host vehicle and the change in the viewing angle between the stationary center and the travel center point. For this reason, it is possible to calculate the distance to the stationary object only by measuring the change in the viewing angle, and to quickly calculate the distance to the stationary object ahead and issue an alarm.

If the distance is within 1 m, there is a possibility of a collision. Further, the distance between the stationary object and the own vehicle is calculated, and the speed of the own vehicle is detected from the speed data of the own vehicle. The time to reach a stationary object is calculated, and if the arrival time is within a predetermined time, an alarm is issued. For this reason, it is possible to quickly determine whether or not a stationary object has a collision, and to issue an alarm with a sufficient time for avoiding the collision.

According to the fifth aspect of the present invention, the moving object determining means measures a change in traveling of the viewing angle of the width between the both sides of the moving object and the angle between the center of the moving object and the center point of the image. As the viewing angle of the width between the two sides of the moving object increases, and the angle between the center of the moving object and the traveling center point hardly changes, there is a possibility of a collision, and an inter-vehicle distance warning is issued. And a collision warning device.

In the fifth aspect of the present invention, the moving object determination means measures a change in traveling of the viewing angle of the width between both sides of the moving object and the angle between the center of the moving object and the advancing center point of the image. As a result, the viewing angle of the width between both sides of the moving object becomes large, and when the angle between the center of the moving object and the moving center point of the image hardly changes, there is a possibility of collision and an alarm is issued. For this reason, even for a moving object, the possibility of collision can be quickly determined only by changing the viewing angle, and an alarm can be issued with sufficient time for collision avoidance.

The present invention of claim 6 is the inter-vehicle distance alarm and the collision alarm device that the alarm issued by the inter-vehicle distance alarm and the collision alarm device changes so that the alarm level increases as the estimated collision time approaches.

In the present invention of claim 6, the warning issued by the inter-vehicle distance warning and the collision warning device is changed so that the warning level increases as the estimated collision time approaches, so a different warning is issued depending on the degree of danger, and the driver Attention can be urged and collision avoidance can be urged accurately.

According to the seventh aspect of the present invention, the correction when turning a curve is performed by measuring the viewing angle between the front of the host vehicle and the center of travel of the image with respect to the center of travel of the image. This is an inter-vehicle distance warning and collision warning device that corrects an angle with respect to a stationary object.

According to the seventh aspect of the present invention, the correction when turning a curve is performed by measuring the viewing angle between the front of the vehicle and the traveling front of the image with respect to the traveling center of the image, To correct. For this reason, the correct distance to the stationary object can be calculated by correcting the shift of the center of travel of the image to the left and right when turning a curve with the viewing angle from the center of travel.

The present invention according to claim 8 is an inter-vehicle distance warning and collision warning device that corrects the height of the horizontal line by the boundary where the stationary object moves with respect to the apparent horizontal line of the image.

In the present invention of claim 8, the correction in the vertical direction change is performed by correcting the height of the horizontal line by the boundary where the stationary object moves with respect to the apparent horizontal line of the image. For this reason, when determining the inter-vehicle distance from the angle in the vertical direction, the center point and the horizontal direction can be matched, and correction in the change in the traveling direction in the vertical direction can be performed.

1, showing an embodiment of the present invention, is a flowchart showing steps of overall determination of an inter-vehicle distance warning and a collision warning device. FIG. 9 is a flowchart illustrating an interrupt process for calculating a vehicle speed used in the determination step of the inter-vehicle distance alarm and the collision alarm device according to the embodiment of the present invention. FIG. 9 is a flowchart illustrating an interrupt process for performing an alarm process used in a warning step of an inter-vehicle distance alarm and a collision alarm device according to an embodiment of the present invention. FIG. 9 is a flowchart illustrating the detailed steps of the image recognition unit, the inter-vehicle distance determination unit, and the stationary object determination unit in the determination step of the inter-vehicle distance alarm and the collision alarm device according to the embodiment of the present invention. FIG. 4 is a diagram illustrating an embodiment of the present invention and an image in front of the host vehicle taken by a front monitoring camera that is an imaging unit. FIG. 4 is an image diagram illustrating an embodiment of the present invention, in which an image captured by a camera as an imaging unit is decomposed into hue, saturation, and luminance, and the hue and saturation are grouped. FIG. 4 is a diagram illustrating an embodiment of the present invention, and is a diagram illustrating horizontal viewing angles of a front monitoring camera that is an imaging unit, a preceding vehicle, and a stationary object. FIG. 5 is a diagram illustrating an embodiment of the present invention and is a diagram illustrating a vertical viewing angle between a front monitoring camera serving as an imaging unit and a preceding vehicle. FIG. 4 is a diagram illustrating an embodiment of the present invention, in which a viewing angle of a preceding vehicle is measured, a distance from the preceding vehicle is determined, and an alarm level is determined. FIG. 5 is a diagram illustrating an embodiment of the present invention, in which a visual angle of a moving object is measured, a possibility of collision with the moving object is determined, and an alarm level is determined. FIG. 4 is a diagram illustrating an embodiment of the present invention and a method for calculating a collision possibility from a viewing angle of a moving object. FIG. 5 is a diagram illustrating an embodiment of the present invention, in which a visual angle of a stationary object is measured to determine a distance from the stationary object, and an alarm level is determined. FIG. 4 is a diagram illustrating a method for calculating a distance from a viewing angle of a stationary object according to an embodiment of the present invention. The embodiment of the present invention is a diagram illustrating a change in the viewing angle of a stationary object in a straight line. FIG. 5 is a diagram illustrating an embodiment of the present invention and a change in a viewing angle of a stationary object on a left curve. FIG. 4 is a diagram illustrating an embodiment of the present invention, and is a diagram for correcting a distance of a stationary object from a change in a viewing angle of the stationary object on a left curve. FIG. 4 is a diagram illustrating an embodiment of the present invention, and is a diagram illustrating a position on a horizontal line of an object at a time when there is no vibration of a front monitoring camera in correction of vertical vibration. FIG. 5 is a diagram illustrating an embodiment of the present invention, and a direction that changes with respect to a horizontal line of an object at the time of vibration of a front monitoring camera in correction of vertical vibration. It is a flowchart which shows the step of the whole determination of the conventional inter-vehicle distance warning and a collision warning apparatus. It is a flowchart which shows the detailed step of the movement amount calculation means and information extraction means of the conventional inter-vehicle distance warning and collision warning device.

An embodiment of the present invention will be described with reference to FIGS.
As shown in FIG. 1, the inter-vehicle distance alarm and collision alarm device of the present invention activates the inter-vehicle distance alarm and collision alarm device whose initial setting has been completed, and is a front monitoring camera that is an imaging means 1 attached to the front of the vehicle. 10 to image the front.

The front monitoring camera 10 uses, for example, a CCD camera capable of shooting color images.
The front monitoring camera 10 preferably has a horizontal shootable angle of about 120 degrees. In the vertical direction, the imageable angle is preferably about 90 degrees. The image size is preferably 1280 × 720 pixels or more. The frame rate is preferably about 60 fps. Thereby, a predetermined calculation speed and accuracy can be obtained.

Next, as shown in FIG. 5, the preceding vehicle 20 and the stationary object 21 (the first stationary object 22, the second stationary object 23, and the third stationary object in FIG. 5) that are the objects captured by the imaging unit 1. An image recognition means 2 for recognizing an image of the vehicle from the lateral direction as the object 24) or the moving object 26; Details of the image recognition means 2 will be described later. For this reason, a front image is taken by the front monitoring camera 10 of the imaging means 1 and the preceding vehicle 20 and the stationary object 21 (in FIG. 5, the first stationary object 22, the second stationary object 23, and the third stationary object 24). ) Or the moving object 26 can be recognized and the identified image can be sent to the judging means.

Next, the data of the preceding vehicle 20 ahead recognized by the image recognition means 2 is sent to the inter-vehicle distance determination means 3. The inter-vehicle distance determination means 3 determines the inter-vehicle distance between the own vehicle 30 and the preceding vehicle 20. When the own vehicle 30 approaches the preceding vehicle 20, an alarm can be issued. Details of the determination of the inter-vehicle distance and the warning will be described later.

Furthermore, the stationary object determination means 4 can determine the distance to the stationary object 21 and the possibility of collision, and can issue an alarm.
Further, the moving object determination means 5 can determine the distance to the moving object 26 and the possibility of collision, and can issue an alarm.

Next, the image recognition means 2 will be described.
A screen ahead of the host vehicle imaged by the front monitoring camera 10 which is the imaging means 1 is as shown in FIG. In the embodiment shown in FIG. 5, the preceding vehicle 20 is displayed at the traveling center point on the front screen, and the first stationary object 22, the second stationary object 23, and the second stationary object 21 are displayed on the right side. Three still objects 24 are displayed in order from the closest one. On the left side, another vehicle that is a moving object 26 approaching from the left is displayed.

About these images, the process for judgment is performed about the image of the object of the horizontal direction of the predetermined width of the up-down direction from the advancing center point of the image of the preceding vehicle 20. FIG. The predetermined width is, for example, a range of one pixel up and down across an apparent horizontal line. The horizontal direction analyzes the image of the entire range. As shown in FIG. 2, it is possible to determine all objects by analyzing all ranges in the horizontal direction. When the stationary object 21 is lost on the way, the range of 11 pixels above and below is monitored, and when it is rediscovered, the apparent horizontal line is corrected in the reverse direction.

As described above, since an image of a horizontal object having a predetermined vertical width (usually 1 pixel) from the advancing center point of the image of the preceding vehicle 20 is determined, the amount of information to be determined is small, and for image recognition. This makes it possible to speed up the processing speed and perform a quick judgment process. For this reason, an alarm can be issued quickly, and the alarm device can be made small and inexpensive.

As shown in FIG. 4, the image recognition unit 2 includes an HSV conversion unit 2a, a center correction unit 2b, an object identification unit 2c, and a viewing angle calculation unit 2d.
In the image recognition means 2, in order to reduce the influence from the light source in judging a plurality of objects, the HSV conversion means 2a converts the values of the video data of the objects into hue, saturation and luminance, respectively.

Next, the center correction unit 2b corrects the center point of the image data so as to be positioned in front of the host vehicle 30. The center correction means 2b is correction at the time of curve, which will be described later, and correction by change in the vertical direction.
Further, as the object identification means 2c, the objects are identified by assigning sequential numbers from the center point according to hue and saturation. In this case, as shown in FIG. 6, the hue and saturation are grouped in each range to identify the object.

To identify the object by grouping the hue and saturation in each range, the hue and saturation are digitized from the signal of the image data, and the colors in the range that can be regarded as the same color are processed as one color. The hue is divided into each color from red to purple shown in FIG. 6, and is roughly divided into 24 stages. The saturation is roughly divided into four stages, as shown by the four circles in FIG. A portion with low saturation is a portion indicated by white at the center. For example, greens with different saturations are processed as different colors such as dark green, green, and light green.

For this reason, in the judgment of an object, the influence from a light source can be reduced and an object can be reliably recognized even in bad weather or at night. Further, it is possible to reliably identify a plurality of objects and perform a determination process.
Further, the viewing angle calculation means 2d is performed. As the viewing angle of the object, the distance from the center point of the image to the object and the width of the object are converted from the number of pixels to the viewing angle. In the present embodiment, the traveling center point and the center point of the image substantially coincide. In this way, determination is performed on all identified objects.

Next, the inter-vehicle distance determination means 3 will be described. As shown in FIG. 4, the inter-vehicle distance determination unit 3 first determines whether the preceding vehicle 20 is in front of the host vehicle 30 by the image recognition unit 2. As shown in FIG. 5, when there is a preceding vehicle 20, the inter-vehicle distance calculation means 3a is performed. Based on the inter-vehicle distance calculated based on the inter-vehicle distance calculation means 3a, the expected collision time is calculated by the time calculation means 3b that predicts the collision time.

As shown in FIG. 9, when the vehicle width of the preceding vehicle 20 includes the front of the own vehicle 30, the preceding vehicle 20 travels in the same lane as the own vehicle 30 as shown in FIG. The distance between the preceding vehicle 20 and the host vehicle 30 is calculated.
The inter-vehicle distance between the preceding vehicle 20 and the host vehicle 30 is a viewing angle between the both ends of the rear portion of the preceding vehicle 20 measured by the front monitoring camera 10 attached to the center of the front surface of the host vehicle 30 as shown in FIG. Calculate by A.

The viewing angle A is calculated by the number of pixels between the preceding vehicle 20 of the image captured by the front monitoring camera 10 and both ends of the rear part.
For example, when the image size is 1280 pixels wide and the horizontal angle of view is 120 degrees, one pixel is 0.09375 degrees, so the viewing angle A can be calculated by the number of pixels × 0.09375 degrees.

In the present embodiment, the vehicle width of the preceding vehicle 20 located near the traveling center is regarded as 1.5 m, the viewing angles at both ends of the vehicle width of the preceding vehicle 20 are measured from the image, and the distance from the preceding vehicle 20 is determined. calculate.
That is, the distance between the vehicles can be calculated by (1.5 m / 2) ÷ tan (viewing angle A / 2) −3 m. As shown in FIG. 8, 3 m represents the distance from the front monitoring camera 10 to the front end of the host vehicle 30 (D in FIG. 8) and the distance from the rear end of the preceding vehicle 20 to the ground contact portion of the rear wheel (E in FIG. 8). This is because the distance of 3) is regarded as 3 m and is subtracted.

For this reason, since it is not necessary to measure the dimension of the both ends of the preceding vehicle 20, the image recognition of the preceding vehicle 20 can be made quick and easy, and determination time can be shortened. Since the vehicle width of the preceding vehicle 20 is regarded as 1.5 m, even in a large vehicle, the inter-vehicle distance is not calculated longer than the actual distance, and the collision time is calculated longer than the actual distance. It is possible to prevent the time from being shortened.

Next, the time calculation means 3b first performs an interruption process for calculating the vehicle speed of the host vehicle 30 as shown in FIG. The time calculation means 3b calculates the travel distance and the vehicle speed from the vehicle speed signal interval. As for the vehicle speed, for example, an interrupt process is generated when a vehicle speed signal transmitted from the rotation of the axle or the like is input, and the vehicle speed is calculated from the travel distance and the signal time interval. For example, in the case of a 4-pulse signal for one rotation of the tire, the travel distance is one-fourth of the distance of the tire outer periphery at the signal interval, and the speed can be calculated by travel distance ÷ time. In the interrupt processing for calculating the vehicle speed, a signal is input by rotation of the tire regardless of the image. Only at that time, the vehicle exits from the main loop and calculates the travel distance and speed. Then, return to the main loop.

Next, the inter-vehicle distance calculated above is divided by the speed of the own vehicle 30 to calculate the time for the own vehicle 30 to reach the preceding vehicle 20, and when the arrival time is within a predetermined number of seconds, an alarm is issued.
In FIG. 4, when the arrival time is within 2.7 seconds, an alarm is issued at alarm level 1. Further, when the arrival time exceeds 2.7 seconds, the inter-vehicle distance in the vertical direction is determined based on FIG. For this reason, it is possible to accurately issue a warning necessary for collision prevention according to the speed of the own vehicle. When the arrival time exceeds a predetermined number of seconds, the vehicle can travel while maintaining a safe inter-vehicle distance of the host vehicle 30 with respect to the preceding vehicle 20.

The generation of alarms and alarm levels will be described.
As shown in FIG. 9, when the vehicle width of the preceding vehicle 20 includes the traveling center, when the preceding vehicle 20 is included, it is determined that the preceding vehicle 20 is traveling in the same lane as the own vehicle 30. Then, as described above, the distance from the preceding vehicle 20 is calculated.

The inter-vehicle distance from the preceding vehicle 20 can be calculated by any of the following methods. W0, W1, K0, and K1 are the width and viewing angle shown in FIG.
Distance between cars = W0 / tan (K0) -3m
Distance between cars = W1 / tan (K1) -3m
Distance between cars = 1.5m / tan (K0 + K1) -3m

In the present embodiment, as shown in FIG. 3, timer interruption processing is performed, for example, a low level alarm is issued when the arrival time becomes 2.7 seconds, and the arrival time becomes 1.9 seconds. Sometimes a high level alarm is issued. The interrupt interval and frequency of the alarm are performed by the timer interrupt process shown in FIG.
Alternatively, a low level alarm is issued when the arrival time reaches 3.5 seconds, and a high level alarm is issued when the arrival time reaches 2.7 seconds.
These times can be increased or decreased by changes in viewing angle, vehicle speed, and the like.
The high level alarm means, for example, increasing the volume, narrowing the alarm interval, or increasing the frequency of the alarm sound.
Timer interrupt processing is performed by interrupting the time to count the alarm buzzer frequency and alarm interval timing, for example, count processing every 1/1000 second.
Since the vehicle speed calculation interrupt process and the timer interrupt process are extremely short, even if this interrupt process is included, it can be processed within 1/60 second as a whole.

Next, calculation of the inter-vehicle distance with the preceding vehicle 20 having a narrow width such as a motorcycle will be described. As shown in FIG. 4, when the arrival time with the preceding vehicle 20 exceeds a predetermined time, the inter-vehicle distance is calculated by measuring in the vertical direction in consideration of the possibility of a motorcycle or the like. The predetermined time is, for example, about 2.7 seconds.

As shown in FIG. 8, the distance between the preceding vehicle 20 and the own vehicle 30 measured in the vertical direction is measured with the front monitoring camera 10 attached to the center of the front surface of the own vehicle 30, and the rear tire of the preceding vehicle 20 is measured. Calculation is based on the viewing angle B between the ground contact position with the road surface and the horizontal direction of the front monitoring camera 10.

The viewing angle B is calculated by the number of pixels between the rear tire of the preceding vehicle 20 in the image captured by the front monitoring camera 10 and the ground contact position of the road surface.
For example, as in the case of the viewing angle A, when the vertical image size is 720 pixels and the vertical angle of view is 90 degrees, one pixel is 0.125 degrees, so the viewing angle A is the number of pixels × 0. It can be calculated at 125 degrees.

In the present embodiment, the position where the front monitoring camera 10 of the own vehicle 30 is attached is, for example, 1.4 m from the ground, and the viewing angles at both ends of the vehicle width of the preceding vehicle 20 are measured from the image, and the preceding vehicle The distance to 20 is calculated.
That is, the distance between the vehicles can be calculated by (1.4 m) ÷ tan (viewing angle B) −3 m. The numerical value of 1.4 m is changed depending on the position of the front monitoring camera 10.

For this reason, the inter-vehicle distance can be reliably calculated even in a motorcycle having a small vehicle width. Based on this inter-vehicle distance, the time for the host vehicle 30 to reach the preceding vehicle 20 is calculated in the same manner as shown in FIG. 7, and when the arrival time is within a predetermined time, as shown in FIG. Issue 1 alarm. For this reason, it is possible to accurately issue a warning necessary for collision prevention according to the speed of the own vehicle. When the arrival time exceeds the predetermined time, the vehicle can travel while maintaining the braking and stopping distance of the host vehicle 30 with respect to the preceding vehicle 20.

Next, as illustrated in FIG. 4, the moving object determination unit 5 determines the moving object 31. The moving object 31 is, for example, a vehicle, a person, a large animal, or the like that is close to the host vehicle 30 from the lateral direction.
As shown in FIGS. 10 and 11, the moving object determination unit 5 measures and determines changes during traveling of the viewing angle before and after the moving object 31 and the angle between the center of the moving object 31 and the traveling center point. .

The viewing angle transition of the moving object 31 changes differently from the viewing angle transition of the stationary object 21. As shown in FIG. 10, the viewing angle K1 indicating the width of the moving object 31 is the viewing angle K0 between the traveling direction of the host vehicle 30 and the moving object 31 (the angle between the center of the moving object 31 and the traveling center point of the image). In the case where the viewing angle K2 indicating the width of the moving object 31 after a predetermined time becomes large despite the fact that it does not change or hardly changes, there is a possibility of a collision. As shown in FIG. Issue 4 alarm.

As shown in FIGS. 10 and 11, the traveling direction of the host vehicle 30 is increased while the viewing angle K2 indicating the width of the moving object 31 after a predetermined time increases from the viewing angle K1 indicating the width of the moving object 31. When the viewing angle K0 between the vehicle and the moving object 31 decreases after a predetermined time, the moving object 31 passes in front of the host vehicle 30.

Although the viewing angle K2 indicating the width of the moving object 31 after a predetermined time increases from the viewing angle K1 indicating the width of the moving object 31, the viewing angle K0 between the traveling direction of the host vehicle 30 and the moving object 31 is the predetermined time. When increasing later, the moving object 31 passes behind the host vehicle 30.
In addition, when the viewing angle K2 indicating the width of the moving object 31 after a predetermined time does not change from the viewing angle K1 indicating the width of the moving object 31, the distance between the moving object 31 and the host vehicle 30 is long and a collision is possible. There is no sex.

Next, the stationary object determination means 4 in FIG. 4 will be described.
First, based on FIGS. 12 and 13, the stationary object determination means 4 calculates a distance W (indicated by W in FIGS. 12 and 13) between the own vehicle 30 and the stationary object 21 in the width direction. If the distance W in the width direction is within 1 m, the distance between the stationary object 21 and the host vehicle 30 is further calculated because there is a possibility of a collision.

The speed of the own vehicle 30 is detected from the speed data of the own vehicle 30 and the time to reach the stationary object 21 is calculated. If the arrival time is within a predetermined time, for example, within 3 seconds, an alarm of level 2 To emit. For this reason, it is possible to quickly determine whether or not the stationary object 21 has a possibility of collision, and to issue an alarm with sufficient time for avoiding the collision.

A method for calculating the distance W in the width direction between the host vehicle 30 and the stationary object 21 will be described.
As shown in FIG. 12, the viewing angle K0 between the stationary object 21 and the front of the host vehicle 30 is measured. Thereafter, when the image is updated, the viewing angle K1 between the next stationary object 21 and the front surface of the host vehicle 30 is measured. When the frame rate of the image is 60 fps, the approach distance is calculated from the vehicle speed calculation of the host vehicle 30 because it is updated every 1/60 seconds. The vehicle speed calculation is performed by the interruption process shown in FIG.

As shown in FIG. 13, the distance W in the width direction is calculated by the following calculation formula.
L2 = sin (K0) × L0
L3 = L2 / tan (K1-K0)
L1 = cos (K1) × L3
W = sin (K0) × L3

Next, correction when turning a curve will be described.
As shown in FIG. 14, when the host vehicle 30 is traveling straight, the viewing angle with respect to the left and right stationary objects 21 changes in the same way as the traveling proceeds, and the point at which the viewing angle change is divided into left and right is the traveling center point. is there. In addition, since the moving object 31 and the preceding vehicle 20 make another movement, they are not used for curve correction.

15 and 16 show a case where the host vehicle 30 curves to the left. When turning to the left, as shown in FIG. 15, the change in the viewing angle of the stationary object 21 on the right side of the center of the screen is larger than the change in the viewing angle of the stationary object 21 on the left side.
Therefore, as shown in FIG. 16, the viewing angle of the deviation between the front of the host vehicle 30 and the center point of the image is added and subtracted as a correction value to correct the correct viewing angle as shown in FIG. The above warning is calculated based on the corrected viewing angle.

When the traveling direction of the host vehicle 30 moves up and down due to changes in the road surface and the image shifts up and down, the horizontal viewing angle is less affected, but the vertical viewing angle as shown in FIG. 8 is measured. Then, when calculating the inter-vehicle distance or the like, an error occurs, so that the apparent horizontal line height is corrected as necessary.
As shown in FIG. 17, when the image does not shift vertically, the horizontal analysis line does not fluctuate.

In the correction in the vertical shift, the height of the apparent horizontal line is corrected by the branch point of the stationary object 21 with respect to the traveling center point. As shown in FIG. 18, when the stationary object 21 moves in the upward direction and is lost in the middle, the horizontal line is down and corrected in that direction, and the stationary object 21 is moved in the downward direction. Searches one pixel above and below, and when rediscovered, changes the apparent horizontal line in the opposite direction.
For this reason, when the inter-vehicle distance is determined from the angle in the vertical direction, the advancing center point and the horizontal direction can be matched, and correction in the vertical vibration can be performed.

DESCRIPTION OF SYMBOLS 1 Image pickup means 2 Image recognition means 3 Inter-vehicle distance determination means 4 Stationary object determination means 5 Moving object determination means 10 Forward monitoring camera 20 Predecessor vehicle 21 Stationary object 26 Moving object 30 Self-vehicle

Claims (8)

The inter-vehicle distance warning and the collision warning device have an imaging means for imaging a front vehicle, a stationary object or a moving object, which is an object,
Image recognition means for recognizing an image of the front vehicle, the stationary object or the moving object, which is the object imaged by the imaging means;
An inter-vehicle distance determining unit that determines an inter-vehicle distance from the vehicle in front recognized by the image recognizing unit;
Having a stationary object judging means for judging the stationary object in front imaged by the imaging means;
Moving object judging means for judging the moving object in front imaged by the imaging means;
The image recognizing unit exchanges the hue, saturation, and luminance for the image of the object having a predetermined width in the vertical direction from the traveling center point of the image of the front vehicle imaged by the imaging unit. By identifying the object in order from the center point in the horizontal direction by adding a serial number,
The inter-vehicle distance determining means, the stationary object determining means, and the moving object determining means measure the identified object from the image, and sequentially calculate the distance from the object identified as the own vehicle,
The speed of the self-vehicle is detected from the speed data of the self-vehicle, the time for the self-vehicle to reach the object is calculated, and if the arrival time is within a predetermined time, an inter-vehicle distance that issues an alarm Alarm and collision alarm device. The inter-vehicle distance determination means regards the vehicle width of the front vehicle positioned at the traveling center point of the image as 1.5 m, measures the viewing angles of both ends of the vehicle width of the front vehicle from the image, 2. The inter-vehicle distance according to claim 1, wherein a distance from the preceding vehicle is calculated to calculate a time for the host vehicle to reach the preceding vehicle, and an alarm is issued when the arrival time is within a predetermined time. Distance alarm and collision alarm device. The inter-vehicle distance determination means measures the angle of the horizontal plane between the position where the tire of the front vehicle located at the center point contacts the road surface and the imaging means from the image, and calculates the road surface of the imaging means. The distance from the vehicle ahead is calculated based on the height from the vehicle, the time for the vehicle to reach the vehicle ahead is calculated, and if the arrival time is within a predetermined time, a warning is issued. Item 4. The inter-vehicle distance warning and the collision warning device according to Item 1. The stationary object judging means measures a change in viewing angle between the moving center point of the image and the stationary object, detects the speed of the own vehicle from the speed data of the own vehicle, and Calculate the travel distance, calculate the distance to the traveling lane of the own vehicle from the travel distance of the own vehicle and the change in the viewing angle between the traveling center point of the image and the stationary object, and the distance Is less than 1 m, it is assumed that there is a possibility of collision, and further, the distance between the stationary object and the own vehicle is calculated,
The inter-vehicle distance according to claim 1, wherein the speed of the self-vehicle is detected from the speed data of the self-vehicle, the time to reach the stationary object is calculated, and an alarm is issued when the arrival time is within a predetermined time. Distance alarm and collision alarm device. The moving object determining means measures a change in visibility of the width between both sides of the moving object and the angle between the center of the moving object and the advancing center point of the image, and the movement A warning may be issued if there is a possibility of a collision when the visibility between both sides of the object is increased and the angle between the center of the moving object and the advancing center point of the image hardly changes. The inter-vehicle distance warning and the collision warning device described. The inter-vehicle distance alarm and the collision alarm device according to any one of claims 1 to 5, wherein the inter-vehicle distance alarm and the alarm issued by the collision alarm device are changed so that the alarm level increases as the estimated collision time approaches. . The correction when turning a curve is to measure the viewing angle between the front of the host vehicle and the center of the image relative to the stationary object with respect to the center point of travel of the image. The inter-vehicle distance alarm and the collision alarm device according to any one of claims 1 to 6, wherein the inter-vehicle distance alarm and the collision alarm device are corrected. The inter-vehicle distance warning and the collision according to any one of claims 1 to 7, wherein the correction in the vertical direction change is performed by correcting a height of a horizontal line by a boundary where the stationary object moves with respect to an apparent horizontal line of an image. Alarm device.

Images