JP2010122821A - Vehicle driving support device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To reliably prevent an accident upon sudden meeting at a crossing or curve by making a mirror image imaged in a curve mirror more recognizable to a driver on running and by automatically recognizing an approaching object in the curve mirror and enabling alarming and vehicle control to the driver. <P>SOLUTION: A vehicle driving support device includes: an image capturing means 10a which captures an image of the front of a vehicle; a specification means 21a which recognizes and specifies the curve mirror from the front image; acquisition means 21b (10b, 11) which acquire an image in the mirror; an image conversion means 21c which converts the image in the mirror into an astigmatism image which is an image seen from the front; an image distortion correcting means 21d which corrects the image distortion in the periphery of the mirror with regard to the astigmatism image; and a display means 40 (23a) which displays the corrected astigmatism image to the vehicle driver. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a vehicle driving support device that supports driving of a vehicle driver.

  In general, a curve mirror (corner mirror) 100 is installed at an intersection or a curve of a road, for example, as shown in FIG. When a driver of a vehicle such as an automobile looks at an image reflected on the curve mirror 100 and visually recognizes an approaching object such as a pedestrian or a vehicle, the driver starts braking or operating a steering wheel to meet at an intersection or a curve. To prevent accidents.

  Since such a curved mirror 100 is usually a convex mirror, it has a wide viewing angle. However, the image reflected in the central portion of the mirror is small, and if the driver is not very close to the curved mirror 100, the curved mirror 100 is not in the curved mirror 100. It is difficult to visually recognize the approaching object. For this reason, there are many cases in which the driver overlooks an approaching vehicle or a pedestrian reflected in the curve mirror 100 and leads to a head accident. However, no technology has been developed that makes it easy for a traveling driver to see a mirror image reflected in the curve mirror 100.

For example, Patent Document 1 below discloses a correction conversion technique for converting a distorted image captured using a wide-angle lens into an image without distortion. However, this technique merely removes the distortion of the wide-angle lens image, and is not applied to the in-curve mirror image (convex mirror image) to be visually recognized by the driver while the vehicle is traveling.
JP 2004-199350 A

  The present invention was devised in view of such a situation, and makes it easy for a traveling driver to see a mirror image reflected in a curved mirror, or automatically recognizes an approaching object in the curved mirror to the driver. The purpose of this is to make it possible to prevent and prevent accidents at the intersections and curves at the beginning of the vehicle.

  In order to achieve the above object, a vehicle driving support device according to the present invention (Claim 1) recognizes and identifies a curve mirror from an image pickup means for picking up an image in front of the vehicle and a vehicle forward image picked up by the image pickup means. A specular image, an in-mirror image acquisition unit that acquires an in-mirror image of the curved mirror specified by the specification unit, and an orthographic image obtained by viewing the in-mirror image acquired by the in-mirror image acquisition unit from the front. An image converting means for converting, an image distortion correcting means for correcting image distortion at the periphery of a mirror for the orthographic image obtained by the image converting means, and a corrected orthoscopic image obtained by the image distortion correcting means. It is characterized by comprising display means for displaying to the driver of the vehicle.

  The vehicle driving support apparatus according to the present invention (Claim 2) includes the above-described imaging means, identification means, in-mirror image acquisition means, image conversion means, and image distortion correction means, and is obtained by the image distortion correction means. A recognition means for recognizing an approaching object from the corrected orthographic image, the post-correction orthographic image obtained by the image distortion correction means, and the vehicle forward image captured by the imaging means, and the recognition of the vehicle. A risk level calculating means for calculating a possibility of collision with the approaching object recognized by the means as a risk level, and an alarm for the driver of the vehicle or the vehicle according to the risk level calculated by the risk level calculation means And control means for performing risk avoidance control by braking the vehicle or steering the vehicle. In such a vehicle driving support apparatus, the risk degree calculation means is configured to determine a distance between the approaching object and the curve mirror based on the mounting angle of the curve mirror calculated from the vehicle front image and the corrected normal image. In addition, the distance between the vehicle and the curve mirror may be calculated based on the vehicle front image, and the degree of risk may be calculated based on these distances.

  According to the vehicle driving support device of the present invention described above, when a curve mirror is specified from the vehicle front image, the in-mirror image of the curve mirror is acquired, and the in-mirror image is converted into a normal image before the mirror. The peripheral image distortion correction is performed, and a post-correction orthographic image is displayed to the driver. As a result, the mirror image displayed on the curve mirror is displayed to the driver in an easy-to-see manner like a plane mirror, making it easier for the driver to determine the distance to the approaching object reflected on the curve mirror and its approach speed. It is possible to prevent an accident at the head of a car or a curve in advance.

  According to the vehicle driving support device of the present invention, when a curve mirror is specified from the vehicle front image, the in-mirror image of the curve mirror is acquired, and the in-mirror image is converted into a normal image before the mirror. Peripheral image distortion correction is performed. Then, when an approaching object is recognized from the corrected orthographic image, the possibility of collision between the vehicle and the approaching object is calculated as a risk based on the corrected orthographic image and the vehicle front image, and driving according to the risk A risk avoidance control is performed by an alarm to the person, braking of the vehicle, or steering of the vehicle. Thus, an image recognition process is performed and an approaching object is easily detected. Therefore, it is possible to automatically recognize an approaching object in the curve mirror and perform a warning and vehicle control for the driver, and it is possible to prevent an accident at the head of the vehicle at an intersection or a curve.

Embodiments of the present invention will be described below with reference to the drawings.
[1] Configuration of this Embodiment FIG. 1 is a block diagram showing a functional configuration of a vehicle driving support apparatus as an embodiment of the present invention. The vehicle driving support apparatus 1 of this embodiment shown in FIG. It is equipped with a vehicle of this type and assists the driving of the vehicle driver. The vehicle driving support apparatus 1 according to this embodiment includes a first camera 10a, a second camera 10b, a camera controller 11, a processing unit 20, a database 30, a display 40, an LED lamp 51, and a speaker 52. The image display in the mirror, the warning for the driver using the display 40, the LED lamp 51, and the speaker 52, and the vehicle control by the brake 61, the steering system 62, and the engine 63 are performed.

  The database 30 includes (a1) a curve mirror pattern (shape / size for a plurality of kinds of standardized curve mirrors) used for recognizing and specifying a curve mirror from a vehicle front image as described later. (A2) Information on the specular curvature for multiple types of standardized curve mirrors, (a3) Templates used for recognizing approaching objects as described later, (a4) Risk levels as described later Reference information for calculation / determination and (a5) a table (see FIG. 5) defining alarm / control modes (actions) according to the degree of risk as described later are registered in advance.

The first camera (imaging means) 10a captures the front of the vehicle widely while the vehicle is running, and acquires a vehicle front image (see, for example, FIG. 2). As the first camera 10a, a dedicated one for the vehicle driving support apparatus may be provided, or for example, a camera for a drive recorder can also be used.
As will be described later, the second camera 10b is for zooming and capturing the image of the curved mirror 100. The second camera 10b has a narrower field of view than the first camera 10a but has a high resolution.

  The camera controller (in-mirror image acquisition means) 11 is provided between the processing unit 20 and the first camera 10a and the second camera 10b, and controls the imaging operation of the first camera 10a and the second camera 10b. It is. In particular, the camera controller 11 controls the second camera 10b according to the processing results of the specifying unit 21a and the in-mirror image acquiring unit 21b described later, and the zoom image (for example, the curve mirror 100) is transmitted to the second camera 10b. (See FIG. 3).

The processing unit (CPU) 20 functions as an image recognition processing unit 21, a determination processing unit 22, and an output control system 23. As the processing unit 20, a dedicated one for the vehicle driving support device may be provided, or for example, a vehicle control CPU or a car navigation system CPU may also be used.
The image recognition processing unit 21 has functions as a specifying unit 21a, an in-mirror image acquiring unit 21b, an image converting unit 21c, and an image distortion correcting unit 21d.

  The specifying means 21a recognizes and specifies the curve mirror 100 from the vehicle front image captured by the first camera 10a. More specifically, the specifying unit 21a performs, for example, an area where the curve mirror 100 exists from the vehicle front image as shown in FIG. 2 by pattern matching processing using the curve mirror pattern (a1) registered in the database 30. Search for A1. Since the curve mirror 100 is standardized and has a predetermined shape and size, the curve mirror 100 can be easily searched for by pattern matching processing based on the shape and size characteristics. Further, since the curve mirror 100 is usually installed slightly above the line of sight, the pattern matching process is performed on a region slightly above the line of sight in the vehicle front image, so that it is quick and efficient. Thus, the curve mirror 100 can be recognized and specified. FIG. 2 is a diagram illustrating an example of a vehicle front image captured by the first camera 10a of the present embodiment.

  The in-mirror image acquisition unit 21b acquires an in-mirror image in the curve mirror 100 specified by the specifying unit 21a. More specifically, the in-mirror image acquisition unit 21b causes the second camera 10b to zoom and image an area A1 (see FIG. 2) where the curve mirror 100 exists in the vehicle front image via the camera controller 11. Then, an image shown in the curve mirror 100 (image in the mirror; image in the area A2 in FIG. 3) is acquired. FIG. 3 is a diagram illustrating an example of a zoom image of the curve mirror 100 captured by the second camera 10b of the present embodiment.

  Here, since the own vehicle (vehicle equipped with the vehicle driving support device 1) normally travels while approaching the curve mirror 100, the camera controller 11 performs the second camera 10b according to the traveling state. The zoom magnification is changed or the direction of the second camera 10b is adjusted. Further, when the resolution of the first camera 10a is sufficiently high, it is not necessary to provide the second camera 10b. Further, as will be described later, even when the first camera 10a has a zooming function, the in-mirror image in the curve mirror 100 can be obtained using the zooming function of the first camera 10a without providing the second camera 10b. It is possible to obtain.

  In the present embodiment, the in-mirror image acquisition unit 21b described above also functions as an attachment angle calculation unit that calculates the attachment angle of the curve mirror 100. That is, the in-mirror image acquisition unit 21b calculates the angle of the curve mirror 100 viewed from the own vehicle from the appearance of the curve mirror 100 based on the zoom image of the curve mirror 100 as shown in FIG. As described above, the curve mirror 100 is standardized, and the shape (hereinafter referred to as “straight-view shape”) and the size when the mirror 100 is viewed from the front are registered in the database 30 for each type of curve mirror. Yes. Therefore, the in-mirror image obtaining unit 21b determines from which angle the zoom image is viewed from the curve mirror 100 based on the shape of the curve mirror 100 in the zoom image and the orthographic shape registered in the database 30. That is, the mounting angle of the curve mirror 100 can be calculated. At this time, as the mounting angle, for example, a depression angle (vertical mounting angle; angle around the horizontal axis) and an angle of the mirror 100 with respect to the traveling direction of the vehicle (left and right mounting angle; angle around the vertical axis) are included. Calculated. This attachment angle is used when the in-mirror distance is calculated in the risk level calculation means 22b described later.

  The image converting means 21c converts the in-mirror image (see area A2 in FIG. 3) acquired by the in-mirror image acquiring means 21b into a normal image viewed from the front. More specifically, the image conversion means 21c acquires the specular shape of the curve mirror 100 specified this time from the database 30, and the in-mirror image acquired by the in-mirror image acquisition means 21b (the image in the area A2 in FIG. 3). ) Is mapped to the acquired normal-view shape region, for example, coordinate transformation by affine transformation is performed. By this image conversion means 21c, the in-mirror image (see area A2 in FIG. 3) acquired by the in-mirror image acquisition means 21b is converted from a state seen from an oblique direction to a state seen from the front (an orthographic image). It will be.

  The image distortion correction unit 21d performs image distortion correction on the periphery of the mirror for the orthographic image obtained by the image conversion unit 21c. The orthographic image obtained by the image conversion means 21c is in a state where the curve mirror 100 is viewed from the front. However, distortion caused by the curve mirror 100 being a convex mirror occurs in the periphery of the mirror of the image. ing. The image distortion correction unit 21d performs correction for removing such distortion on the normal image obtained by the image conversion unit 21c.

  As the correction method, for example, Seiya Takatsuki et al. “Lens distortion correction method using gray code pattern” (URL: http://www.am.sanken.osaka-u.ac.jp/~echigo/miru2005. pdf # search = 'Lens distortion correction method using gray code pattern'), Katsumi Seki and others "Image correction technology in digital cameras" (Ricoh Technical Report No.31, pp.103-110, DECEMBER, 2005) The image conversion technique (image correction technique) disclosed in the above can be used. In the above correction, the image distortion correcting unit 21d can read and use the mirror curvature rate (a2) of the type of curve mirror corresponding to the curve mirror 100 registered in the database 30.

  The image distortion correction means 21d corrects the distortion of the peripheral part of the mirror due to the convex mirror, and the coordinate conversion of the image reflected on the curve mirror 100 into an easy-to-view image (for example, see FIG. 4) as reflected on the plane mirror. Will be. FIG. 4 is a diagram showing an example of the post-correction orthographic image displayed on the display 40 in the present embodiment. As shown in FIG. 4, the post-correction orthographic image obtained by the image distortion correcting means 21d is The information is displayed on the display (display means) 40 by the display control unit 23a of the output control system 23 described later.

The determination processing unit 22 has functions as an approaching object recognition unit 22a, a risk level calculation unit 22b, and a control unit 22c.
The approaching object recognizing means (recognizing means) 22a recognizes the approaching object 101 (see FIG. 4) such as a vehicle, a motorcycle, or a pedestrian from the corrected front-view image obtained by the image distortion correction means 21d. The approaching object recognizing means 22a includes various existing object recognition methods such as template matching (matching processing using a template (a3) registered in the database 30 in advance), a shape recognition method such as a neuron, and a moving object recognition method. Is used to recognize the approaching object 101 from the corrected orthographic image reflected on the curve mirror 100. As a recognition technique applied here, a technique generally known as a technique for recognizing a moving object (a car-like shape) from an image obtained by a camera that images the front of the vehicle can be used.

  The risk level calculation means 22b is based on the corrected front-view image obtained by the image distortion correction means 21d and the vehicle front image taken by the first camera 10a, and the approaching object recognized by the own vehicle and the approaching object recognition means 22a. The possibility of collision with 101 is calculated as the degree of danger. More specifically, the risk level calculation unit 22b is based on the mounting angle of the curved mirror 100 calculated from the vehicle front image (the mounting angle calculated by the function of the in-mirror image acquisition unit 21b) and the corrected normal image. The distance d1 between the approaching object 101 and the curve mirror 100 is calculated, the distance d2 between the host vehicle and the curve mirror 100 is calculated based on the vehicle front image, and the risk is calculated based on these distances d1 and d2 ( (Danger level determination).

  Here, the distance d2 between the own vehicle and the curve mirror 100 is the vehicle front image captured by the first camera 10a as shown in FIG. 2 if the mounting position and the directing direction of the first camera 10a are known. It is possible to calculate from the Y-axis coordinates. The mounting position and the directing direction of the first camera 10a are usually provided as known values. Further, when a stereo camera is used as the first camera 10a, the distance d2 between the own vehicle and the curve mirror 100 can be calculated from the parallax (viewing angle).

  In addition, the distance d1 between the approaching object 101 and the curve mirror 100 can be determined as follows if the mounting angle of the curve mirror 100 is known, as shown in FIG. It can be calculated from the axis coordinates. As described above, the mounting angle of the curve mirror 100 is calculated by the mounting angle calculation function of the in-mirror image acquisition means 21b. Then, the time differential value of the distance d1 becomes the approach speed v1 of the approaching object 101.

Based on the distances d1 and d2, the approach speed v1 and the speed v2 of the host vehicle (for example, a value obtained from the speed sensor), the risk level calculation unit 22b may calculate the risk (possibility) of the collision, for example, The following five levels of risk levels (b1) to (b5) are used.
Danger level (b1): No danger Danger level (b2): Danger transition process (state that is moving to a dangerous state)
Danger level (b3): Dangerous state (a state where a collision occurs if you run without doing anything)
Danger level (b4): Criticality (a condition that will lead to a collision if evasive action is not taken urgently)
Danger level (b5): Collision inevitable (physical collision avoidance is impossible)

  Here, the correspondence between the distances d1, d2, the approach speed v1 and the speed v2 of the host vehicle and the above five levels of risk levels (b1) to (b5) is registered in the database 30 as reference information (a4) in advance. Has been. Actually, the distance D (= d1 + d2) between the own vehicle and the approaching object 101, the approach speed V (= v1 + v2) of the approaching object 101 with respect to the own vehicle, and the above five risk levels (b1) to (b5). ) Is registered in advance as reference information (a4).

  The risk level calculation means 22b illuminates the current distance D and the approach speed V between the host vehicle and the approaching object 101 with reference to the reference information (a4), and the current situation is any of the risk levels (b1) to (b5). It is calculated and judged whether it corresponds to. For example, when the distance D is not less than a predetermined value and the speed V is sufficiently slow (for example, the distance D is not less than 50 m and the speed V is not more than 10 km / h), the risk level (b1) is determined as “no danger”. . When the distance D is less than the predetermined value and the speed V is high (for example, the distance D is less than 50 m and the speed V is 60 km / h or more; in this case, the collision occurs within 5 seconds), the risk level (b5) “collision” Inevitable ”. As described above, by appropriately setting the reference information (a4) stepwise for the distance D and the approach speed V, the risk level calculation means 22b calculates and determines the risk level based on the distance D and the speed of approach V. It is possible to do.

  The control means 22c performs a warning avoidance control by a warning to the driver of the own vehicle or braking / steering / engine control of the own vehicle according to the risk level calculated by the risk calculating means 22b. The correspondence relationship between the five levels of risk levels (b1) to (b5) and the alarm / control mode (action) corresponding to each risk level is stored in the database 30 in a table (a5) as shown in FIG. As registered in advance. The control unit 22c refers to the table (a5), recognizes the alarm / control mode (action) corresponding to the risk level obtained by the risk level calculation unit 22b, and determines the recognized alarm / control mode (action). This is executed via the output control system 23. FIG. 5 is a diagram showing an example of a table (a5) that defines an alarm / control mode (action) according to the degree of risk in the present embodiment.

  According to the definition of the table (a5) shown in FIG. 5, the control means 22c performs, for example, the following control operations (c1) to (c5) according to the risk levels (b1) to (b5). It is like that. In the table (a5) shown in FIG. 5, “◎” indicates an essential action, and “Δ” indicates an action that may be performed as necessary.

(c1) If the risk level is (b1) “No danger”, no control action is performed.
(c2) Risk level (b2) In the case of “dangerous transition process”, information is provided to notify the driver that the state is about to transition to a dangerous state. Specifically, the control means 22c controls the display 40 to display that it is in a state of shifting to a dangerous state via the display control unit 23a of the output control system 23, and provides information to the driver. At the same time, the information is provided by controlling the color / flashing / lighting of the LED lamp 51 through the alarm control unit 23b of the output control system 23 or controlling the speaker 52 to output a beep sound or artificial sound.

  (c3) In case of danger level (b3) “Dangerous state”, alert the driver to recognize that it is a state that will cause a collision if you run without doing anything. Specifically, the control means 22c causes the display 40 to display the above warning through the display control unit 23a of the output control system 23, and the LED lamp 51 via the alarm control unit 23b of the output control system 23. Color / flashing / lighting control is performed, or control for outputting a beep sound or artificial sound from the speaker 52 is performed to call out the above-mentioned attention. At this time, for example, the LED lamp 51 and the speaker 52 are alerted and activated so that the driver has a higher degree of urgency (risk) than when the information is provided. The beep sounding interval by the speaker 52 is made shorter than that at the time of providing the information. Further, as necessary, the control means 22c can provide braking control of the brake 61, steering control of the steering system 62, and control of the engine 63 via the braking control unit 23c, the steering control unit 23d, and the engine control unit 23e of the output control system 23. You may perform collision avoidance control (intervention control) by performing operation control.

  (c4) When the risk level is (b4) “Danger criticality”, a warning is given to the driver to recognize that a collision will occur if no evasive action is taken urgently. Specifically, the control means 22c causes the display 40 to display the alarm via the display control unit 23a of the output control system 23, and the color of the LED lamp 51 via the alarm control unit 23b of the output control system 23. The above alarm is given by performing flashing / lighting control or controlling the speaker 52 to output a beep sound or artificial sound. At this time, the alarm operation by the LED lamp 51 or the speaker 52 is, for example, the blinking interval of the LED lamp 51 or the speaker so that it becomes clear to the driver that the urgency level (risk level) is higher than that at the time of calling attention. The beep sounding interval by 52 is further shortened compared to the above alerting. Also at this time, as necessary, the control means 22c controls the braking of the brake 61 and the steering control of the steering system 62 via the braking control unit 23c, the steering control unit 23d, and the engine control unit 23e of the output control system 23. Alternatively, operation control of the engine 63 may be performed to perform collision avoidance control (intervention control).

  (c5) When the risk level is (b5) “collision inevitable”, collision avoidance is physically impossible, so forced collision avoidance control (intervention control) is performed as well as an alarm. Specifically, the control means 22c outputs an alarm indicating that the collision is inevitable by the display 40 via the display control unit 23a of the output control system 23, and via the alarm control unit 23b of the output control system 23. The alarm is performed by controlling the color / flashing / lighting of the LED lamp 51 or controlling the speaker 52 to output a beep sound or artificial sound. Simultaneously with this alarm control, the control means 22c, via the braking control unit 23c, the steering control unit 23d and the engine control unit 23e of the output control system 23, performs braking control of the brake 61, steering control of the steering system 62, and operation of the engine 63. Control is performed to forcibly perform collision avoidance control (intervention control).

The output control system 23 has functions as a display control unit 23a, an alarm control unit 23b, a braking control unit 23c, a steering control unit 23d, and an engine control unit 23e.
The display control unit 23a controls the display state of the display (display unit) 40, and displays the corrected orthographic image obtained by the image distortion correction unit 21d on the display 40 so as to be presented to the driver. In addition to having a function, it has a function of performing display control of the display 40 in accordance with an instruction from the control means 22c accompanying the control operations (c2) to (c5).

  The alarm control unit 23b controls the operation of the LED lamp 51 and the speaker 52, and the operation of the LED lamp 51 and the speaker 52 according to the instruction from the control means 22c accompanying the control operations (c2) to (c5). It has a function to perform control.

  The braking control unit 23c controls braking by the brake 61, and controls braking of the brake 61 (intervention for collision avoidance) according to an instruction from the control means 22c accompanying the control operations (c3) to (c5). Control).

  The steering control unit 23d controls the steering by the steering system 62, and the steering control by the steering system 62 (for collision avoidance) according to the instruction from the control means 22c accompanying the control operations (c3) to (c5). Has a function to perform intervention control).

  The engine control unit 23e controls the operation of the engine 63. The engine control unit 23e controls the operation of the engine 63 (intervention for collision avoidance) according to the instruction from the control means 22c accompanying the control operations (c3) to (c5). Control).

[2] Operation of the present embodiment Next, the operation of the vehicle driving support device 1 of the present embodiment configured as described above will be described according to the flowchart (steps S11 to S32) shown in FIG.
While the host vehicle is traveling, the vehicle front image captured by the first camera 10a is always input to the processing unit 20 (specification unit 21a) via the camera controller 11, and the specification unit 21a causes the curve in the vehicle front image to be displayed. The presence or absence of a mirror is checked (step S11).

  When the region A (see FIG. 2) where the curve mirror 100 is present in the vehicle front image is specified by the specifying unit 21a (YES route in step S11), the camera controller 11 performs the second operation based on the specifying result of the specifying unit 21a. The camera 10b is directed toward the curve mirror 100 (step S12), and the camera controller 11 zooms the second camera 10b onto the curve mirror 100 (step S13). (See FIG. 3).

  Thereafter, the mounting angle of the curve mirror 100 is calculated by the function of the in-mirror image acquisition unit 21b as the mounting angle calculation unit (step S15). At that time, based on the shape of the curve mirror 100 in the zoom image acquired in step S14 and the orthographic shape registered in the database 30, the zoom image obtained by viewing the curve mirror 100 from which angle, that is, As the mounting angle of the curve mirror 100, the depression angle and the angle of the mirror 100 with respect to the traveling direction of the vehicle are calculated.

  On the other hand, when the in-mirror image shown by area A2 in FIG. 3, for example, is acquired from the zoom image acquired in step S14 by the in-mirror image acquisition means 21b (step S16), the image conversion means 21c in step S16 The acquired in-mirror image (see region A2 in FIG. 3) is coordinate-converted and converted from a state viewed from an oblique direction to a state viewed from the front (an orthographic image) (step S17).

  Then, the image distortion correction means 21d performs coordinate conversion of the orthographic image obtained in step S17 into an image of a plane mirror, whereby the image distortion correction of the mirror peripheral portion is performed on the orthographic image, and the mirror periphery of the in-mirror image Distortion caused by the curved mirror 100 being a convex mirror occurring in the part is removed (step S18).

  The corrected stereographic image obtained in step S18 is displayed on the display 40 by the display control unit 23a and presented to the driver (step S19). As a result, the mirror image reflected on the curve mirror 100 is displayed to the driver in an easy-to-view manner like a plane mirror.

  Further, the approaching object recognition means 22a uses, for example, a shape recognition method such as template matching or neurology or a moving object recognition method, and the vehicle, motorcycle, pedestrian, etc. from the corrected front-view image (converted image) obtained in step S18. The approaching object 101 (see FIG. 4) is recognized (step S20). Here, when the approaching object 101 is not recognized, for example, the processing returns to step S11 and the same processing as described above is repeated.

  When the approaching object 101 is recognized in step S20, the distance d1 between the approaching object 101 and the curve mirror 100 is calculated by the risk level calculation means 22b based on the Y-axis coordinates of the corrected orthographic image obtained in step S18. In addition, the speed v1 of the approaching object 101 is calculated as a time differential value of the distance d1 (step S21).

  Further, the risk level calculation means 22b calculates the distance d2 between the host vehicle and the curve mirror 100 based on the Y-axis coordinates of the vehicle front image captured by the first camera 10a, and the traveling speed v2 of the host vehicle is calculated. It is obtained by the speed sensor of the own vehicle (step S22).

  Thereafter, the current distance D (= d1 + d2) and the approach speed V (= v1 + v2) between the host vehicle and the approaching object 101 are compared with the reference information (a4) in the database 30 by the risk level calculation means 22b. Which of the risk levels (b1) to (b5) corresponds to the possibility of collision is calculated and determined (step S23).

  Then, according to the risk level calculated in step S23, the control means 22c executes warning for the driver of the own vehicle and risk avoidance control by braking / steering / engine control of the own vehicle (steps S24 to S32). ).

  That is, in the case of the danger level (b5) “inevitable collision” (YES route of step S24), since collision avoidance is physically impossible, not only alarm but also forced collision avoidance control (intervention control) Is done. Specifically, as described above as the control operation (c5), an alarm is given to the driver by the display 40, the LED lamp 51, and the speaker 52, the brake control of the brake 61, the steering control of the steering system 62, and the engine. 63 operation control is performed, and collision avoidance control (intervention control) is forcibly performed (step S25).

  In the case of the danger level (b4) “danger criticality” (NO route in step S24 and YES route in step S26), an alarm is issued to make the driver recognize that a collision will occur if no avoidance action is taken urgently. . Specifically, as described above as the control operation (c4), a warning is given to the driver by the display 40, the LED lamp 51, and the speaker 52 (step S27). At this time, the collision avoidance control (intervention control) may be performed simultaneously with the alarm by performing braking control of the brake 61, steering control of the steering system 62, and operation control of the engine 63.

  In the case of the danger level (b3) “dangerous state” (NO route in step S26 and YES route in step S28), a warning is issued to let the driver recognize that it is in a state where a collision will occur if nothing is done as it is. It is. Specifically, as described above as the control operation (c3), the driver is alerted by the display 40, the LED lamp 51, and the speaker 52 (step S29). Also at this time, the collision avoidance control (intervention control) may be performed by performing the braking control of the brake 61, the steering control of the steering system 62, and the operation control of the engine 63 simultaneously with the alerting.

  In the case of the risk level (b2) “danger transition process” (NO route in step S28 and YES route in step S30), information is provided to notify the driver that the vehicle is in a dangerous state. Specifically, as described above as the control operation (c2), information is provided to the driver through the display 40, the LED lamp 51, and the speaker 52 (step S31).

  When the risk level (b1) is “no danger” (NO route of step S30), if the collision avoidance process of any of steps S25, S27, S29, and S31 has been executed until immediately before, the collision avoidance process is performed. After canceling (step S32), the process returns to the process of step S11. On the other hand, if the collision avoidance process is not performed immediately before, the process returns to the process of step S11 without performing any control operation.

  In the example shown in FIG. 6, the determination and the collision avoidance process are preferentially performed when the risk level requiring high speed is high.

[3] Effects of the Embodiment As described above, according to the vehicle driving support device 1 as an embodiment of the present invention, when the curve mirror 100 is specified from the vehicle front image, the in-mirror image of the curve mirror 100 is identified. Is acquired, and the image in the mirror is converted into a normal image, and then image distortion correction around the mirror is performed, and the corrected normal image is displayed on the display 40 to the driver. As a result, the mirror image reflected on the curve mirror 100 is displayed to the driver in an easy-to-view manner like a plane mirror, so that the driver can easily determine the distance to the approaching object 101 reflected on the curve mirror 100 and the approach speed. Thus, it is possible to prevent an accident at the beginning of a meeting at an intersection or a curve.

  Further, when the approaching object 101 is recognized from the corrected orthographic image, the possibility of collision between the own vehicle and the approaching object 101 is calculated as the risk based on the corrected orthographic image and the vehicle front image, and the risk is Accordingly, warning to the driver or risk avoidance control (collision avoidance control) by braking / steering / engine control of the own vehicle is performed. As described above, the approaching object 101 is easily detected by performing the image recognition processing. Therefore, the approaching object 101 in the curve mirror 100 is automatically recognized, and the vehicle can be forcibly controlled (risk avoidance control / collision avoidance control) regardless of the driver's intention or the driver's intention. It is possible to prevent an accident at the head of a car or a curve in advance and more reliably.

  Further, when the second camera 10b as the in-mirror image acquisition means has a zooming function, the curve mirror 100 is displayed on the zoom image captured by the second camera 10b even when the vehicle approaches the curve mirror 100 as the vehicle travels. The camera controller 11 automatically adjusts the zoom magnification of the second camera 10b so as to always have the same size, thereby performing a process of adjusting the size of the image on the recognition engine (image recognition processing unit 21) side. This is unnecessary, and the load on the recognition engine can be greatly reduced.

[4] Modified Example of Embodiment In the above-described embodiment, the first camera 10a that captures a wide area in front of the vehicle and obtains the vehicle front image, and the second camera that has a narrower field of view than the first camera 10a but has a high resolution. Although the case where two cameras 10b are provided has been described, the same operational effects as described above can be obtained with one camera as described below.

  For example, if the first camera 10a as the imaging means has a zooming function, it is possible to acquire a zoom image of the curve mirror 100 by this zooming function. By the first camera 10a having this zooming function, An in-mirror image in the curve mirror 100 can be acquired. At this time, the zooming function of the first camera 10a is controlled by the camera controller 11, and the curve mirror 100 is captured and followed by the first camera 10a. In this case, since the zooming function is used, not only the second camera 10b is unnecessary, but it is not necessary to use a high resolution camera as the first camera 10a, and the apparatus cost can be greatly reduced.

  In addition, when the first camera 10a as the imaging unit has a zooming function as described above, the curve mirror on the zoom image captured by the first camera 10a even when the vehicle approaches the curve mirror 100 as the vehicle travels. A function (which can be realized by the camera controller 11) for automatically adjusting the zoom magnification of the first camera 10a is provided so that 100 is always the same size. By providing such a function, it is not necessary to adjust the image size on the recognition engine (image recognition processing unit 21) side, and the load on the recognition engine side can be greatly reduced.

  Furthermore, in the case where the first camera 10a as an imaging means for capturing a front image of the vehicle is configured to have a resolution as high as that of the second camera 10b described above, the second camera 10b can be made unnecessary. It is possible to acquire in-mirror images of two or more curved mirrors simultaneously from one high-resolution vehicle front image, and to perform image conversion processing and image distortion correction on the two or more in-mirror images.

[5] Others The present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the spirit of the present invention.
For example, in the above-described embodiment, the case where the shape of the curve mirror 100 is rectangular has been described. However, the present invention is not limited to this, and if the curve mirror is standardized in advance, The shape is not limited to a rectangle, and may be a circle or other shapes. In any case, it is needless to say that the same effects as those of the above-described embodiment can be obtained. In the above-described embodiment, the case where the risk level is the five levels (b1) to (b5) has been described. However, the present invention is not limited to such five levels.

  Further, the above-described image recognition processing unit 21 (identifying unit 21a, in-mirror image obtaining unit 21b, image converting unit 21c and image distortion correcting unit 21d), determination processing unit 22 (approaching object recognizing unit 22a, risk degree calculating unit 22b and The control means 22c) and the output control system 23 (display control unit 23a, alarm control unit 23b, braking control unit 23c, steering control unit 23d, and engine control unit 23e) function (all or part of functions) are computers ( This is realized by a CPU (information processing apparatus and various terminals) including a predetermined application program (vehicle driving support program).

  The program is, for example, a flexible disk, CD (CD-ROM, CD-R, CD-RW, etc.), DVD (DVD-ROM, DVD-RAM, DVD-R, DVD-RW, DVD + R, DVD + RW, etc.), Blu-ray Disc And the like recorded in a computer-readable recording medium. In this case, the computer reads the vehicle driving support program from the recording medium, transfers it to an internal storage device or an external storage device, and uses it. Further, the program may be recorded in a storage device (recording medium) such as a magnetic disk, an optical disk, or a magneto-optical disk, and provided from the storage device to a computer via a communication line.

  Here, the computer is a concept including hardware and an OS (operating system), and means hardware operating under the control of the OS. Further, when the OS is unnecessary and the hardware is operated by the application program alone, the hardware itself corresponds to the computer. The hardware includes at least a microprocessor such as a CPU and means for reading a computer program recorded on a recording medium. The application program as the vehicle driving support program includes a program code that causes the above-described computer to realize the functions as the image recognition processing unit 21, the determination processing unit 22, and the output control system 23 described above. Also, some of the functions may be realized by the OS instead of the application program.

It is a block diagram which shows the function structure of the vehicle driving assistance device as one Embodiment of this invention. It is a figure which shows an example of the vehicle front image imaged with the 1st camera of this embodiment. It is a figure which shows an example of the zoom image of the curve mirror imaged with the 2nd camera of this embodiment. It is a figure which shows an example of the post-distortion normal-view image displayed on a display in this embodiment. It is a figure which shows an example of the table which defines the alarm / control aspect (action) according to the risk in this embodiment. It is a flowchart for demonstrating operation | movement of the vehicle driving assistance device shown in FIG. It is a figure for demonstrating the installation condition of a general curve mirror (corner mirror).

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Vehicle drive assistance apparatus 10a 1st camera (imaging means)
10b Second camera (in-mirror image acquisition means)
11 Camera controller (in-mirror image acquisition means)
20 Processing unit (CPU)
21 Image recognition processing unit 21a Identification unit 21b Image acquisition unit in mirror (attachment angle calculation unit)
21c Image conversion means 21d Image distortion correction means 22 Judgment processing part 22a Approaching object recognition means (recognition means)
22b Risk calculation means 22c Control means 23 Output control system 23a Display control part 23b Alarm control part 23c Braking control part 23d Steering control part 23e Engine control part 30 Database 40 Display (display means)
51 LED lamp 52 Speaker 61 Brake 62 Steering system 63 Engine 100 Curve mirror (corner mirror)
101 Approaching object (vehicle)

Claims (3)

Imaging means for imaging the front of the vehicle;
Identifying means for recognizing and identifying a curve mirror from a vehicle front image captured by the imaging means;
In-mirror image acquisition means for acquiring an in-mirror image in the curved mirror specified by the specifying means;
Image conversion means for converting the in-mirror image acquired by the in-mirror image acquisition means into a normal image viewed from the front;
Image distortion correction means for correcting image distortion at the periphery of the mirror for the orthographic image obtained by the image conversion means;
A vehicle driving support apparatus, comprising: a display means for displaying a corrected front-view image obtained by the image distortion correction means to a driver of the vehicle. Imaging means for imaging the front of the vehicle;
Identifying means for recognizing and identifying a curve mirror from a vehicle front image captured by the imaging means;
In-mirror image acquisition means for acquiring an in-mirror image in the curved mirror specified by the specifying means;
Image conversion means for converting the in-mirror image acquired by the in-mirror image acquisition means into a normal image viewed from the front;
Image distortion correction means for correcting image distortion at the periphery of the mirror for the orthographic image obtained by the image conversion means;
Recognizing means for recognizing an approaching object from the corrected orthographic image obtained by the image distortion correcting means;
Risk of collision between the vehicle and the approaching object recognized by the recognizing unit based on the post-corrected normal image obtained by the image distortion correcting unit and the vehicle front image captured by the imaging unit A risk calculating means for calculating the degree,
Control means for performing warning avoidance control by warning to the vehicle driver or braking of the vehicle or steering of the vehicle according to the risk calculated by the risk calculation means. A vehicle driving support device as a feature.   The risk degree calculating means calculates the distance between the approaching object and the curve mirror based on the mounting angle of the curve mirror calculated from the vehicle front image and the corrected orthographic image, and the vehicle front image. The vehicle driving support device according to claim 2, wherein a distance between the vehicle and the curve mirror is calculated based on the distance, and the risk is calculated based on the distance.

Images