JP4539472B2 - Vehicle travel support device - Google Patents

Description

  TECHNICAL FIELD The present invention relates to a vehicular driving support device, and more particularly to a vehicular driving support device that detects pedestrians around the host vehicle, and technical fields for assisting driver recognition, determination, and operation to reduce trouble. Belonging to.

Conventionally, in order to detect an obstacle such as a vehicle or a pedestrian existing around the host vehicle, the obstacle is provided with an imaging means such as a CCD camera or a CMOS camera to assist the driver's recognition, judgment and operation. 2. Description of the Related Art A vehicle travel support device that supports a driver's vehicle operation so as to avoid contact with the vehicle is known. For example, in Patent Literature 1, an object to be imaged in a region corresponding to a person's head to shoulder and a shoulder to arm is taken as a pedestrian among the imaging data around the subject vehicle captured by the camera. Recognizing technology is disclosed. Further, in Patent Document 2, the same space around the host vehicle is imaged with two cameras arranged side by side on the left and right, and the object to be imaged is an obstacle such as a vehicle or a pedestrian that should avoid contact, or A technique for accurately determining whether a non-obstacle such as a white line or a sign on the road surface is disclosed.
JP 2004-303219 A JP 2004-86779 A

  However, the conventional imaging means is provided in a relatively upper part of the host vehicle, for example, in the vicinity of the room mirror, and images a relatively upper space around the host vehicle. Even if it is visible, many non-obstacles such as trees, buildings, and vehicles are reflected in the background. As a result, image processing becomes complicated, and pedestrians that should avoid touching are extracted and specified from the captured data. It took a long time or failed. In addition, as a result of imaging the upper body of the pedestrian including the shoulder from the person's head, for example, the shape and contour to be recognized are more complex than when the lower body of the pedestrian including the leg of the person is imaged, Also from this point, there is a problem that it takes a long time or fails to extract / specify a pedestrian whose contact should be avoided from the imaging data.

  An object of the present invention is to cope with the above-described problems in the vehicle travel support device, and to enable the extraction and identification of pedestrians by the imaging means to be performed accurately, in a short time. To do.

That is, in order to solve the above-mentioned problem, the invention according to claim 1 of the present application is a vehicular travel support device that detects a pedestrian around the host vehicle, and is a lower part that images a lower space around the host vehicle. Spatial imaging means, and pedestrian detection means for detecting the presence of a pedestrian around the host vehicle based on the imaging result of the lower space imaging means, the lower space imaging means being grounded on the road surface The height of the lower space imaging means from the road surface and the depression angle to the pedestrian's feet imaged by the lower space imaging means And a distance measuring means for measuring the distance between the pedestrian and the host vehicle .

  Next, the invention according to claim 2 is the vehicle travel support device according to claim 1, wherein the upper space imaging means for imaging the upper space around the host vehicle and the imaging result of the lower space imaging means And second pedestrian detection means for detecting the presence of a pedestrian around the host vehicle based on the imaging result of the upper space imaging means.

Next, according to a third aspect of the present invention, in the vehicle travel support apparatus according to the first or second aspect , the lower space imaging means is provided below the vehicle body floor panel so as to be rotatable in the horizontal direction. It is characterized by being.

Next, according to a fourth aspect of the present invention, in the vehicular driving support apparatus according to any one of the first to third aspects, the lower space imaging means is provided at a lower portion of the host vehicle, and the vehicle speed. Lens protection means for protecting the lens of the lower space imaging means is provided when the vehicle speed exceeds a predetermined vehicle speed.

First, according to the first aspect of the present invention, the lower space imaging means for imaging the lower space around the host vehicle is provided, and a pedestrian exists around the own vehicle based on the imaging result of the lower space imaging means. It is possible to detect that many non-obstacles such as standing trees, buildings and vehicles are reflected in the background of the pedestrian, and the background of the pedestrian is almost occupied by the road surface. Processing becomes simple and simple, and extraction and identification of pedestrians whose contact should be avoided from the imaging data can be performed accurately and in a short time. In addition, as a result of imaging the lower body of the pedestrian including the person's legs, the shape and contour to be recognized are simpler and simpler than when the upper body of the pedestrian including the shoulder is captured from the head of the person. Therefore, also from this point, it becomes possible to extract and specify a pedestrian whose contact should be avoided from the imaging data, reliably and accurately in a short time.

Further, according to the present invention, the lower space imaging means is provided at a position where the foot of a pedestrian who is in contact with the road surface can be imaged, and the height of the lower space imaging means from the road surface, Since the distance between the pedestrian and the subject vehicle is measured based on the depression angle to the pedestrian's foot imaged by the lower space imaging means, for example, distance measurement such as laser radar or millimeter wave radar There is no need to provide a separate device for Even if the pedestrian is close to the host vehicle, an accurate distance from the host vehicle to the pedestrian can be obtained.

  Next, according to the second aspect of the present invention, the apparatus further includes an upper space imaging unit that images the upper space around the host vehicle, and the imaging result of the lower space imaging unit and the imaging result of the upper space imaging unit Based on the above, since the presence of pedestrians around the vehicle is detected, the presence of pedestrians can be comprehensively detected from more data (data fusion: data fusion). ), The recognition accuracy of being a pedestrian can be improved.

According to the third aspect of the present invention, since the lower space imaging means is provided below the vehicle body floor panel and is provided so as to be rotatable in the horizontal direction, it is possible to remove all adjacent objects and pedestrians around the host vehicle. It is possible to monitor in the direction, detect obstacles when starting forward or starting backward, and avoid troubles such as contact when starting.

According to the invention described in claim 4 , since the lower space imaging means is provided in the lower part of the host vehicle and the vehicle speed is equal to or higher than a predetermined vehicle speed, the lens of the lower space imaging means is protected. In a situation where the lens of the image pickup unit is easily soiled, the lens can be prevented from being soiled. Hereinafter, the present invention will be described in more detail through embodiments of the invention.

  FIG. 1 is a layout diagram of each component of the vehicle travel support apparatus 1 according to the present embodiment. The driving support apparatus 1 includes an upper camera 20 as an upper space imaging unit that images an upper space around the host vehicle X, and includes a lower camera 30 as a lower space imaging unit that images a lower space around the host vehicle X. It is out. The upper camera 20 is provided in the vicinity of the rear-view mirror at the front of the vehicle compartment, and the lower camera 30 is provided at the front bumper at the front end of the vehicle. These cameras 20 and 30 are constituted by image sensors such as a CCD camera and a CMOS camera, for example. These cameras 20 and 30 capture the surroundings of the host vehicle X, and extract and specify obstacles such as other vehicles and pedestrians that should avoid contact from the captured data.

  In addition, the driving support device 1 includes a radar 40 such as a laser radar or a millimeter wave radar as a distance measuring unit that measures a distance between an obstacle around the host vehicle X and the host vehicle X. The radar 40 is provided on the front grill at the front end of the vehicle. Although not shown in detail, the radar 40 includes a light emitting unit that emits light toward the obstacle, and a light receiving unit that receives the light emitted from the light emitting unit and reflected by the obstacle. The distance from the own vehicle X to the obstacle is measured based on the time from when the light is emitted from the part until the light is received by the light receiving part.

  The travel support device 1 also includes a display and alarm device 50. This display and alarm device 50 is provided on the instrument panel in front of the driver's seat. The display and alarm device 50 displays obstacles such as other vehicles and pedestrians around the host vehicle X detected by the cameras 20 and 30 and the radar 40, and avoids contact with the obstacles. In addition, a visual and auditory warning is issued to the driver. Furthermore, the driving support apparatus 1 includes a vehicle speed sensor 60 that detects the vehicle speed of the host vehicle X, and a lens cover actuator 70 as lens protection means that protects the lens of the lower camera 30.

  FIG. 2 is a control system diagram centering on the control unit 10 of the driving support apparatus 1. Although not shown, the control unit 10 has a central processing unit composed of a CPU. The control unit 10 inputs a signal from the upper camera 20 (upper space image), a signal from the lower camera 30 (lower space image), and a signal from the radar 40 (measurement distance), and based on these signals. And detecting that there is a pedestrian around the host vehicle X (pedestrian detection means, second pedestrian detection means), and outputs a control signal to the display and alarm device 50 according to the detection result, Inform the driver of the presence of a pedestrian. As a result, the driver's vehicle operation is assisted so as to assist the driver's recognition / judgment / operation and avoid contact with the pedestrian.

  The control unit 10 also receives a signal (vehicle speed) from the vehicle speed sensor 60 and outputs a control signal to the lens cover actuator 70 based on this signal to protect the lens of the lower camera 30 from contamination.

  FIG. 3 is an explanatory diagram showing the imaging ranges of the upper camera 20 and the lower camera 30 in this embodiment (first embodiment) in which the lower camera 30 is disposed in the front bumper portion. As described above, since the upper camera 20 images the upper space around the host vehicle X, it captures the upper body including the head and shoulders of the pedestrian Y existing in front of the host vehicle X. Since the camera 30 captures an image of the lower space around the host vehicle X, the camera 30 captures the lower body of the pedestrian Y existing in front of the host vehicle X including the legs and toes.

  As a result, as shown in FIG. 4, the image (upper space image) W <b> 1 captured by the upper camera 20 includes a large number of non-obstacles such as standing trees in addition to the upper body of the pedestrian Y that should avoid contact Therefore, there is a concern that it takes a long time or fails to extract and specify the pedestrian Y from the imaging data. Moreover, since the upper body of the pedestrian Y has a more complicated shape and contour than the lower body, it takes a long time to extract / specify the pedestrian Y from the imaging data, or the pedestrian Y fails. There is a concern that.

  On the other hand, as shown in FIG. 5, an image (lower space image) W2 captured by the lower camera 30 includes a large number of non-obstacles such as standing trees on the background of the lower body of the pedestrian Y that should avoid contact. Since the background of the lower body of the pedestrian Y is almost occupied by the road surface R, the image processing becomes simple and simple, and the extraction and identification of the pedestrian Y from the imaging data is surely accurate. Well, it can be done in a short time. Also, since the lower body of the pedestrian Y is simpler and simpler than the upper body, the shape and contour are simpler and simpler. From this point, the extraction and identification of the pedestrian Y from the imaging data is surely and accurately performed It can be done in a short time.

FIG. 6 is an explanatory diagram of the principle of measuring the distance L between the host vehicle X and the pedestrian Y by the lower camera 30. The lower camera 30 can position the foot of the pedestrian Y who is in contact with the road surface R (for example, the front bumper as in the first embodiment or the floor panel as in the second embodiment). (Lower part). And the distance L between the pedestrian Y and the own vehicle X can be measured based on following Formula (1).
Distance to pedestrian Y L = H × (f / y)
= H × tan (90−θ) (1)
Here, H is the height of the lower camera 30 from the road surface R, f is the focal length of the lower camera 30, and y is the position of the foot of the pedestrian Y on the imaging surface s of the lower camera 30 (y (Coordinate position) and θ are depression angles to the pedestrian Y's foot. Therefore, if the radar 40 is the first distance measuring means, it can be said that the lower camera 30 is the second distance measuring means.

  Next, a specific example of the pedestrian recognition operation by the upper camera 20 performed by the control unit 10 will be described with reference to the flowchart of FIG. First, the control unit 10 acquires the upper space image W1 in step S1, and determines whether any object is detected in the upper space image W1 in step S2. As a result, when it is detected, in step S3, recognition of the pedestrian Y is executed using the upper space image W1, and when it is not detected, the process proceeds to step S24 in FIG. Recognition of the pedestrian Y is performed using the lower space image W2 imaged by the camera 30.

  Next, in step S4, it is determined whether or not the pedestrian Y has been recognized. If so, the distance L to the pedestrian Y is measured using the radar 40 in step S5, and in step S6, When it is determined that the measurement distance L is shorter than the predetermined distance Lo, it is determined in step S7 that the pedestrian Y is an object (pedestrian Y) that needs to be re-recognized, and will be described later with reference to FIG. Proceeding to step S31, re-recognition of the pedestrian Y is executed using both the upper space image W1 imaged by the upper camera 20 and the lower space image W2 imaged by the lower camera 30 (data fusion). On the other hand, when it is determined in step S6 that the measured distance L is longer than the predetermined distance Lo, in step S8, the pedestrian Y is still walking at a remote position from the own vehicle X that is not likely to contact. It is determined that the user is Y, and the process returns (shifts to processing of the next frame).

  On the other hand, when the pedestrian Y cannot be recognized in step S4, the distance L to the object is measured using the radar 40 in step S9, and the measurement distance L is determined in step S10. If it is determined that the distance is shorter than the distance Lo, it is determined in step S7 that the object is an object that needs to be re-recognized, and the process proceeds to step S31 in FIG. Object re-recognition is executed using both the spatial image W1 and the lower spatial image W2 captured by the lower camera 30 (data fusion). On the other hand, when it is determined in step S10 that the measurement distance L is longer than the predetermined distance Lo, it is determined in step S11 that the object is an object other than the pedestrian Y, and the process returns (processing of the next frame). To).

  Next, a specific example of the pedestrian recognition operation by the lower camera 30 performed by the control unit 10 will be described with reference to the flowchart of FIG. First, the control unit 10 acquires the lower space image W2 in step S21, and determines whether any object is detected in the lower space image W2 in step S22. As a result, if it is detected, it is determined in step S23 whether or not there is an object detected by the upper camera 20, and if not, in step S24, the pedestrian Y is recognized using the lower space image W2. On the other hand, if there is, in step S31, recognition of the pedestrian Y is performed using both the upper space image W1 and the lower space image W2 ("re-recognition" when moving from step S7 in FIG. 7) ( Data fusion).

  In any case, in step S25, it is determined whether or not the pedestrian Y has been recognized. If so, the distance L to the pedestrian Y is measured using the lower camera 30 in step S26. If it is determined in step S27 that the measured distance L is shorter than the predetermined distance Lo, in step S28, the pedestrian Y is required to avoid contact with the host vehicle X in the highest priority. It determines with it being the pedestrian Y who exists in a position, and returns (it transfers to the process of the following flame | frame).

  On the other hand, when it is determined in step S27 that the measured distance L is longer than the predetermined distance Lo, in step S29, the pedestrian Y is still in a remote position from the own vehicle X that is unlikely to contact. It is determined that the user is Y, and the process returns (shifts to processing of the next frame).

  On the other hand, when the pedestrian Y cannot be recognized in step S25, it is determined in step S30 that the object is an object other than the pedestrian Y, and the process returns (shifts to processing of the next frame). ). Furthermore, if no object is detected in step S22, the process returns as it is (transition to the next frame).

  As described above, the present embodiment includes the lower camera 30 that images the lower space around the host vehicle X, and the pedestrian Y exists around the host vehicle X based on the imaging result W2 of the lower camera 30. (Step S24 in FIG. 8), it is suppressed that many non-obstacles such as standing trees, buildings and vehicles are reflected in the background of the pedestrian Y, and the background of the pedestrian Y Is almost occupied by the road surface R (W2 in FIG. 5), and the image processing becomes simple and simple, and the extraction and identification of the pedestrian Y that should avoid contact from the imaging data is surely and accurately performed. It can be done in a short time.

  Further, the result of imaging the lower body of the pedestrian Y including the person's legs (W2 in FIG. 5), for example, compared to the case where the upper body of the pedestrian Y including the shoulder from the person's head is imaged (FIG. 4). W1), the shape / contour to be recognized becomes simple and simple. Also from this point, it is possible to reliably and accurately extract the pedestrian Y whose contact should be avoided from the imaging data in a short time. Now you can do it.

  The camera further includes an upper camera 20 that captures an upper space around the host vehicle X, and walks around the host vehicle X based on an imaging result W2 of the lower camera 30 and an imaging result W1 of the upper camera 20. Since the presence of the person Y is detected (step S31 in FIG. 8), the presence of the pedestrian Y can be comprehensively detected from more data (data fusion). This improves the recognition accuracy.

  In that case, when the radar 40 that measures the distance L between the subject imaged by the upper camera 20 and the host vehicle X is provided, and the distance L measured by the radar 40 is shorter than the predetermined distance Lo. Only (steps S6 and S10 in FIG. 7), the presence of a pedestrian Y around the host vehicle X is detected based on the imaging result W2 of the lower camera 30 and the imaging result W1 of the upper camera 20. Therefore, in a situation where the recognition accuracy of being a pedestrian Y is required to be higher, the presence of a nearby pedestrian Y can be comprehensively detected from more data (data fusion), The recognition accuracy of being a nearby pedestrian Y can be improved, and the verification result / reconfirmation of being a nearby pedestrian Y can be performed including the imaging result W2 of the lower camera 30.

  Further, the lower camera 30 is provided at a position where the foot of the pedestrian Y who is in contact with the road surface R can be imaged. The height H of the lower camera 30 from the road surface R and the lower camera 30 Since the distance L between the pedestrian Y and the host vehicle X is measured based on the imaged depression angle θ to the foot of the pedestrian Y (FIG. 6), a device for distance measurement ( The need for a separate radar 40) is eliminated. Even if the pedestrian Y is close to the host vehicle X, the accurate distance L from the host vehicle X to the pedestrian Y can be calculated.

  By the way, when the lower camera 30 is provided in the lower part of the own vehicle X in order to image the lower space around the own vehicle X, there is a concern about contamination of the lens. Therefore, when the vehicle speed detected by the vehicle speed sensor 60 is equal to or higher than the predetermined vehicle speed, the lens cover actuator 70 may be operated to cover the lens with the cover to protect the lens. Thereby, it becomes possible to prevent the lens of the lower camera 30 from being soiled in a situation where the lens is easily soiled. As another example of the lens protection means, when the vehicle speed is equal to or higher than a predetermined vehicle speed, the camera may be stored inside the vehicle body.

  Next, FIG. 9 is an explanatory diagram showing respective imaging ranges of the upper camera 20 and the lower camera 30 in the second embodiment in which the lower camera 30 is disposed at the lower part of the vehicle body floor panel. As in the first embodiment, the upper camera 20 images the upper body including the head and shoulders of the pedestrian Y existing in front of the host vehicle X in order to capture the upper space around the host vehicle X. Thus, the lower camera 30 images the lower body including the legs and feet of the pedestrian Y existing in front of the host vehicle X in order to capture the lower space around the host vehicle X.

  FIG. 10 is an image diagram of image data (lower space image) W3 captured by the lower camera 30 in the second embodiment. Since the lower camera 30 is disposed at the lower part of the vehicle body floor panel, the upper part of the image W3 is occupied by the floor panel (same as X of the host vehicle), and the lower part of the image W3 is occupied by the road surface R. Then, with the road surface R in the background, the legs and toes of the pedestrian Y that should avoid contact are imaged. Therefore, it is completely prevented that non-obstacles such as standing trees and buildings are reflected in the background of the pedestrian Y, and since the background of the pedestrian Y is entirely occupied by the road surface R, the image processing becomes much simpler and simpler, and imaging Extracting and specifying the pedestrian Y from the data can be performed more reliably, accurately and in a short time.

  In the second embodiment, the lower camera 30 may be provided so as to be rotatable in the horizontal direction. It is possible to monitor omnidirectional objects and pedestrians Y around the host vehicle X. In particular, it is possible to detect obstacles when starting forward or starting backward, and contact with obstacles when starting. Etc. can be avoided.

  Although the present embodiment is the best embodiment of the present invention, it goes without saying that various modifications and changes can be made without departing from the scope of the claims. For example, the distance L to the object to be imaged captured by the upper camera 20 (steps S5 and S9 in FIG. 7) and the distance L to the object to be imaged captured by the lower camera 30 (step S26 in FIG. 8). Only when they are substantially the same, the pedestrian Y may be recognized and re-recognized using both the upper space image W1 and the lower space image W2 (step S31 in FIG. 8). Data fusion will be smooth and recognition accuracy is expected to increase.

  As described above in detail with reference to specific examples, the present invention is a vehicle driving support device that assists the driver's recognition, determination, and operation. It can be carried out in a short time, and is expected to have a wide range of industrial applicability in the technical field of vehicular travel support devices.

FIG. 2 is a layout diagram of each component of the vehicle travel support apparatus according to the best embodiment of the present invention. It is a control system figure centering on the control unit of the said travel support apparatus for vehicles. It is explanatory drawing which shows each imaging range of an upper camera and a lower camera at the time of arrange | positioning a lower camera to a front bumper part (1st Embodiment). It is an image figure of the imaging data (upper space image) by the upper camera of the said 1st Embodiment. It is an image figure of the imaging data (lower space image) by a lower camera equally. It is a principle explanatory view of distance measurement to a pedestrian by a lower camera. It is a flowchart which shows one specific example of the pedestrian recognition operation | movement by an upper camera which the said control unit performs. It is a flowchart which shows one specific example of the pedestrian recognition operation | movement by a lower camera which the said control unit performs. It is explanatory drawing which shows each imaging range of an upper camera and a lower camera at the time of arrange | positioning a lower camera to a floor panel lower part (2nd Embodiment). It is an image figure of the imaging data (lower space image) by the lower camera of the said 2nd Embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Vehicle travel assistance apparatus 10 Control unit (pedestrian detection means, 2nd pedestrian detection means)
20 Upper camera (upper space imaging means)
30 Lower camera (lower space imaging means, second distance measuring means)
40 Radar (distance measuring means)
50 Vehicle speed sensor 60 Display and alarm device 70 Lens cover actuator (lens protection means)
R Road surface X Own vehicle Y Pedestrian

Claims (4)

A vehicle travel support device that detects pedestrians around the host vehicle,
Lower space imaging means for imaging the lower space around the host vehicle;
Pedestrian detection means for detecting the presence of a pedestrian around the host vehicle based on the imaging result of the lower space imaging means ,
The lower space imaging means is provided at a position where it is possible to image a pedestrian's toes that are in contact with the road surface,
Distance measuring means for measuring the distance between the pedestrian and the vehicle based on the height from the road surface of the lower space imaging means and the depression angle to the pedestrian's feet imaged by the lower space imaging means A vehicular travel support apparatus characterized by being provided. In the vehicle travel support device according to claim 1,
Upper space imaging means for imaging the upper space around the host vehicle;
And further comprising second pedestrian detection means for detecting the presence of a pedestrian around the host vehicle based on the imaging result of the lower space imaging means and the imaging result of the upper space imaging means. A vehicle driving support device. In the vehicle travel support device according to claim 1 or 2 ,
The vehicular travel support apparatus, wherein the lower space imaging means is provided below the vehicle body floor panel so as to be rotatable in the horizontal direction . In the vehicle travel support device according to any one of claims 1 to 3,
The lower space imaging means is provided at the lower part of the host vehicle ,
A vehicle travel support apparatus, comprising: lens protection means for protecting the lens of the lower space imaging means when the vehicle speed is equal to or higher than a predetermined vehicle speed .

Images