JP2013020295A - Vehicle periphery monitoring device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a vehicle periphery monitoring device in which calculation of an inappropriate actual space position is prevented from being performed by erroneous detection of distance between a vehicle and an object when the actual space position of the object is calculated by combination of a camera and a radar device.SOLUTION: A vehicle periphery monitoring device includes: an image part detection unit 22 which projects an object detected from ranging data of a radar device 12 by an object distance detection unit 21 on a pick up image of a camera 11 to detect an image part of the object; an actual space position calculation unit 23 which calculates an actual space position of the object on the basis of detected distance of the object distance detection unit 21; a translucent object detection unit 25 which detects a grating fence present within a detection range; and an actual space position reliability determination unit 24 which prohibits calculation of the actual space position by the actual space position detection unit 23 for an object present at a position to overlap with the grating fence in the imaging direction of the camera 11.

Description

  The present invention relates to a vehicle periphery monitoring device that monitors an object existing around a vehicle using a radar device and a camera mounted on the vehicle.

  2. Description of the Related Art Conventionally, there is known a vehicle periphery monitoring device that determines the type of an object existing around a vehicle by a radar device and a camera mounted on the vehicle and calculates a real space position of the object (for example, a patent) Reference 1).

  In the conventional vehicle periphery monitoring device, an object existing around the vehicle is detected by a radar device, and the distance between the object and the vehicle is detected. Then, by projecting the object detected by the radar device onto the captured image of the camera, the image portion of the object is detected and its type is discriminated.

  When the object type is a monitoring target such as a pedestrian, the position of the image portion of the object on the captured image is converted into the real space position of the object based on the detection distance between the vehicle and the object. A relative position with respect to the object is calculated.

JP-A-9-264554

  As described above, the inventors of the present invention combine a camera and a radar device to detect an image of an object existing around the vehicle and a distance to the object, and detect the object detected by the camera and the object detected by the radar device. It was discovered that the distance to the object detected by the camera may be detected incorrectly due to inconsistencies. Thus, when the distance to the detected object by the camera is erroneously detected, the calculation of the real space position of the object based on this distance becomes inappropriate.

  The present invention has been made in view of such a background, and when a real space position of an object existing around a vehicle is calculated by a combination of a camera and a radar apparatus, an erroneous detection of a distance between the vehicle and the object by the radar apparatus is made. Accordingly, an object of the present invention is to provide a vehicle periphery monitoring device that prevents an inappropriate actual section position from being used.

  The present invention has been made in order to achieve the above object, and by scanning a detection range around the vehicle with a radar device mounted on the vehicle, an object present in the detection range is detected, and the An object distance detection unit for detecting a distance between an object and the vehicle, and an object detected by the object distance detection unit is projected on a captured image of a camera mounted on the vehicle to detect an image portion of the object And a real space position calculation unit that calculates a real space position of an object in which the image part is detected by the image part detection unit based on a detection distance by the object distance detection unit. It relates to a monitoring device.

  A semi-transparent object detection unit that is within the detection range, shields the beam output from the radar device, and detects a semi-transparent object in which other objects behind the image are captured by the camera And the reliability of the real space position calculated by the real space position calculation unit for an object existing in a position overlapping with the semi-transparent object detected by the semi-transparent object detection unit in the imaging direction of the camera A real space position reliability determination unit is provided (first invention).

  In the first invention, the beam output from the radar device is shielded between the vehicle and an object existing in the vicinity thereof, and the other object behind is imaged when imaged by the camera. An object (such as a lattice fence) may be present. In this case, while the object behind the semi-transparent object is imaged by the camera, only the semi-transparent object is detected by the radar device, and the object behind the semi-transmissive object is not detected by the radar device. There is a case.

  In this case, when the real space position calculation unit calculates the real space position of the object in which the image part is detected by the image part detection unit based on the detection distance by the object distance detection unit, An erroneous real space position based on the detection distance of a different translucent object is calculated.

  Therefore, the real space position reliability determination unit determines the reliability of the real space position calculated by the real space position calculation unit for an object existing at a position overlapping the semi-transparent object detected by the semi-transparent object detection unit. Decrease sex. Thereby, it is possible to prevent an inappropriate real space position from being used for an object that exists at a position overlapping with the semi-transmissive object detected by the semi-transmissive object detection unit.

  In the first aspect of the present invention, a blur degree calculation unit that calculates a degree of blur of the image portion detected by the image portion detection unit is provided, and the real space position reliability determination unit is configured such that the image portion is detected by the image portion detection unit. For a detected object, a relationship between the degree of blur of the image portion calculated by the blur degree calculation unit and the distance between the object detected by the object distance detection unit and the vehicle is set in advance. When the reference relationship between the degree and the distance deviates by a first predetermined level or more, the real space position calculation unit prohibits the calculation of the real space position of the object (second invention).

  According to the second invention, the degree of blurring of the image portion of the object detected by the image portion detection unit changes according to the distance between the object and the vehicle. Therefore, the real space position reliability determination unit is a relationship between the degree of blur of the image portion of the object calculated by the degree of blur calculation unit and the distance between the object detected by the object distance detection unit and the vehicle. However, when the predetermined reference level relationship between the degree of blur and the distance deviates by a first predetermined level or more, the real space position calculation unit prohibits the calculation of the real space position of the object. Thereby, when the correspondence between the image portion and the detection distance is incorrect, it is possible to prevent the inappropriate real space position from being used due to the inappropriate real space position being calculated by the real space position calculating unit. it can.

  As described above, the process of prohibiting the calculation of the real space position of the object by the real space position calculation unit corresponds to lowering the reliability of the real space position of the object calculated by the real space position calculation unit.

  In the first or second aspect of the invention, the apparatus further comprises a displacement state calculation unit that calculates a displacement state of an image portion of the same object detected by the image portion detection unit from a time-series captured image by the camera. The spatial position reliability determination unit is configured such that the displacement status calculated by the displacement status calculation unit is a change in distance from the vehicle detected by the object distance detection unit for the same object, and a traveling speed of the vehicle. The real space position of the object is prohibited from being calculated by the real space position calculation unit when the estimated displacement state calculated based on is deviated by a second predetermined level or more (third invention).

  According to the third invention, when the displacement state calculated by the displacement state calculation unit deviates from the estimated displacement state by the second predetermined level or more, an image portion is detected by the image portion detection unit. It can be determined that the correspondence between the object and the distance detected by the object distance detection unit is incorrect. Therefore, in this case, the real space position reliability determination unit prohibits the calculation of the real space position of the object by the real space position calculation unit. This prevents the real space position calculation unit from calculating an inappropriate real space position and preventing the use of an inappropriate real space position for the object whose image portion has been detected by the image part detection unit. Can do.

  Further, in any one of the first to third inventions, the rate of change of the size of the image part of the same object detected by the image part detection unit is calculated from the time-series captured images by the camera, An estimated distance calculating unit that calculates an estimated distance between the same object and the vehicle based on the rate of change and the traveling speed of the vehicle, and the real space position reliability determining unit includes the object for the same object; When the distance to the vehicle detected by the distance detector deviates from the estimated distance calculated by the estimated distance calculator by a third predetermined level or more, the real space position calculator calculates the actual object The calculation of the spatial position is prohibited (fourth invention).

  According to the fourth aspect of the present invention, when the distance detected by the object distance detector deviates from the estimated distance calculated by the estimated distance calculator by the third predetermined level or more, the image portion detector detects an image. It can be determined that the correspondence between the object whose portion is detected and the distance detected by the object distance detection unit is incorrect. Therefore, in this case, the real space position reliability determination unit prohibits the calculation of the real space position of the object by the real space position calculation unit. This prevents the real space position calculation unit from calculating an inappropriate real space position and preventing the use of an inappropriate real space position for the object whose image portion has been detected by the image part detection unit. Can do.

  Further, in any one of the first to fourth inventions, the navigation apparatus mounted on the vehicle includes a host vehicle position recognition unit that recognizes a position of the vehicle on a map, and the real space position reliability is provided. The determination unit detects the pedestrian's image portion by the image portion detection unit, and determines the position on the map of the vehicle calculated by the own vehicle position calculation unit and the distance detected by the object distance detection unit. When the position of the pedestrian on the map recognized based on the map is off the road, the calculation of the real space position of the pedestrian by the real space position calculation unit is prohibited (fifth invention) ).

  According to the fifth invention, when the position of the pedestrian detected by the object position detection unit is an unnatural situation where the position on the map of the vehicle by the navigation device is off the road, The space position reliability determination unit prohibits the calculation of the real space position by the real space position calculation unit for the pedestrian. As a result, for the pedestrian whose image part is detected by the image part detection unit, an inappropriate real space position is calculated by the real space position calculation unit and an inappropriate real space position is not used. be able to.

  In any one of the first to fifth inventions, the semi-transmissive object detection unit detects a lattice fence as the semi-transmissive object (sixth invention).

  According to the sixth aspect of the present invention, the lattice fence in which the object behind the image is imaged by the camera through the lattice and the beam output from the radar device is shielded by the lattice and does not reach the object behind the half is provided. It can be detected as a transparent object.

  In the sixth invention, the translucent object detection unit detects that the object is a lattice fence when the width of the object calculated from the distance measurement data of the radar device is equal to or larger than a predetermined value. (Seventh invention).

  According to the seventh aspect of the present invention, it is possible to easily detect the presence of a grid fence having a width larger than that of a pedestrian or the like that is a monitoring target from the distance measurement data of the radar device.

Explanatory drawing of the attachment aspect to the vehicle of a vehicle periphery monitoring apparatus. Explanatory drawing of the scanning range of the left-right direction of a radar apparatus. The block diagram of a vehicle periphery monitoring apparatus. Explanatory drawing of the condition where a lattice fence interposes between a vehicle and an object. Explanatory drawing of the captured image of a lattice fence and the object which exists behind it. Explanatory drawing of an example of the process which detects a lattice fence. Explanatory drawing of the process which judges the appropriateness of a detection distance by the blurring degree of an image part. Explanatory drawing of the process which judges the suitability of detection distance by the optical flow of an image part. Explanatory drawing of the process which judges the suitability of detection distance by the estimated distance based on the change rate of the magnitude | size of an image part. Explanatory drawing of the process which judges the suitability of detection distance with the information from a navigation apparatus.

  An embodiment of a vehicle periphery monitoring device of the present invention will be described with reference to FIGS.

  Referring to FIG. 1, a vehicle periphery monitoring device 10 is mounted on a vehicle (own vehicle) 1 having an infrared camera 11 and a radar device 12 mounted on the front side. The infrared camera 11 has a characteristic that the output level increases (the luminance increases) as the temperature of the imaged object increases. Instead of the infrared camera 11, a camera whose sensitivity is adjusted to other wavelength regions such as visible light may be used.

  Further, in the real space coordinate system of the object detected by the radar device 12, the front portion of the vehicle 1 is the origin 0, the X axis is the left and right direction of the vehicle 1, the Y axis is the up and down direction of the vehicle 1, and the Z axis is the vehicle 1 It is defined as the front-rear direction.

  The radar device 12 is attached to an upper position of the infrared camera 11 on the front side of the vehicle 1. As shown in FIG. 2, the radar device 12 generates a millimeter-wave (electromagnetic wave) beam BM that spreads in a narrower width than the imaging range L1 to L2 of the infrared camera 11 when viewed from above the vehicle 1. The transmission angle of the beam BM having a constant intensity is shifted by a predetermined angle so as to be scanned in the left-right direction of the vehicle 1 and output.

  The scanning range of the beam BM in the left-right direction is set so as to include all the imaging areas L1 to L2 of the infrared camera 11 in the horizontal direction. Further, the vertical width of the beam BM is set so as to include the vertical visual field of the imaging range L1 to L2 of the infrared camera 11.

  The radar apparatus 12 receives this millimeter wave reflected wave, that is, the millimeter wave reflected by the object existing in front of the vehicle 1 by a receiving antenna (not shown) arranged in the vertical direction. The radar apparatus 12 detects an object present in the transmission direction of the beam BM based on a reflected wave having a predetermined intensity or more among the received reflected waves.

  Further, the radar apparatus 12 detects the distance between the object reflecting the beam BM and the vehicle 1 based on the time difference between the transmission wave and the reception wave. In addition, as a radar apparatus, you may employ | adopt the radar apparatus which uses elastic vibration waves, such as an ultrasonic wave, besides electromagnetic waves (laser light etc.), such as a millimeter wave.

  Next, referring to FIG. 3, the vehicle periphery monitoring device 10 is an electronic unit composed of a CPU, a memory, an interface circuit, etc. (not shown), and digitally outputs an analog video signal output from the infrared camera 11. The CPU has a function of converting the data into an image memory (not shown) and performing various arithmetic processes by the CPU on the front image of the vehicle 1 captured in the image memory.

  The vehicle periphery monitoring device 10 also includes distance measurement data including reflection intensity data indicating the reflection intensity of the millimeter wave reflected by the object from the radar device 12 and distance data indicating the distance between the object and the vehicle 1. Is entered. Further, the vehicle 1 includes a yaw rate sensor 13 that detects the yaw rate of the vehicle 1, a speed sensor 14 that detects the traveling speed of the vehicle 1, and a brake sensor 15 that detects an operation state of a brake pedal (not shown) of the vehicle 1. The detection signals of these sensors are input to the vehicle periphery monitoring device 10.

  In addition, the vehicle 1 includes an audio output device 16, an image output device 17, a braking device 18, and a navigation device 19, and operations of these devices are controlled by a control signal output from the vehicle periphery monitoring device 10. The As the image output device 17, a HUD (head-up display) that displays an image on the front window of the vehicle 1, a display meter that indicates the traveling state of the vehicle 1, a display device that constitutes a navigation device, and the like are employed.

  The vehicle periphery monitoring device 10 detects an object located in front of the vehicle 1 based on distance measurement data output from the radar device 12 by executing a vehicle periphery monitoring program stored in the memory by the CPU. The object distance detecting unit 21 that detects the distance between the object and the vehicle 1, the image part detecting unit 22 that detects the image part of the object detected by the object distance detecting unit 21 from the image captured by the infrared camera 11, and the image The real space position calculation unit 23 that calculates the real space position of the object from which the image portion is detected by the partial detection unit 22, and the real space that determines the reliability of the real space position of the object calculated by the real space position calculation unit 23 The position reliability determination unit 24 functions.

  Further, the vehicle periphery monitoring device 10 detects a semi-transmissive object (for example, a lattice fence such as a wire mesh) based on a captured image obtained by the infrared camera 11 or distance measurement data output from the radar device 12. 25. Displacement status for calculating the optical flow of the image part of the same object between the time-series captured images by the infrared camera 11 and the blur degree calculating part 26 for calculating the degree of blur of the image part detected by the image part detecting unit 22 It functions as an estimated distance calculation unit 28 that calculates the rate of change of the size of the same object between the calculation unit 27 and the time-series captured images by the infrared camera 11.

  The object distance detection unit 21 detects the distance between the object and the vehicle 1 by scanning the detection range with the radar device 12 and detecting an object existing in the detection range for each predetermined control cycle. The image part detection unit 22 projects the detection position of the object by the object distance detection unit 21 onto the captured image of the infrared camera 11, sets the detection range of the image part of the object, and sets the image of the object within the detection range. Detect part.

  The real space position calculation unit 23 converts the position (coordinates) of the image part detected by the image part detection unit 22 into a real space position based on the distance detected by the object distance detection unit 21, thereby generating an image. The real space position of the object from which the part is detected is calculated.

  Note that details of the process of converting the position of the image portion in the captured image into the real space position based on the distance between the object and the vehicle 1 are described in Japanese Patent Laid-Open No. 9-264554 described above. Therefore, the description is omitted here.

  The vehicle periphery monitoring device 10 determines the possibility of contact between the vehicle 1 and the object whose real space position is calculated by the real space position calculation unit 23. In addition, regarding the determination process of the possibility of contact between the vehicle 1 and the object, for example, as described in Japanese Patent Application Laid-Open No. 2001-6096, the detected movement vector of the object is set in front of the vehicle. When it is determined from the determination area to the approach determination area (the object approaches the vehicle 1), it is determined that there is a possibility that the object (approaching object) and the vehicle 1 are in contact with each other.

  When it is determined that there is a possibility that the object and the vehicle 1 are in contact with each other, the vehicle periphery monitoring device 10 executes a contact avoidance process. Specifically, the vehicle periphery monitoring apparatus 10 determines whether or not the driver is operating a brake based on a detection signal from the brake sensor 15. When the brake operation is not performed, an alarm from the audio output device 16 and the image output device 17 is immediately output.

  On the other hand, when the driver performs a brake operation, the vehicle periphery monitoring device 10 determines whether or not contact is avoided by the operation of the brake device 18 according to the brake operation. When it is determined that it will be avoided, an alarm is not output so that the driver is not bothered excessively.

  When it is not determined that the vehicle is to be avoided, the vehicle periphery monitoring device 10 outputs an alarm from the audio output device 16 and the image output device 17. Furthermore, the vehicle periphery monitoring device 10 operates the braking device 18 so that contact is avoided as necessary.

  Next, the real space position reliability determination unit 24 determines the distance between the object and the vehicle 1 detected by the object distance detection unit 21 when at least one of the following first to fifth conditions is satisfied. Based on the above, a process of prohibiting the calculation of the real space position of the object by the real space position calculation unit 23 is performed. Hereinafter, this process will be described with reference to FIGS. As described above, the process of prohibiting the calculation of the real space position of the object by the real space position calculation unit 23 corresponds to lowering the reliability of the real space position of the object calculated by the real space position calculation unit 23. .

[First Condition] Conditions Based on Detection of Lattice Fence As shown in FIG. 4A, when a lattice fence 50 (such as a wire fence) is interposed between the vehicle 1 and an object 60 (such as an advertising board). In addition, the radar apparatus 12 detects the lattice fence 50 existing in the foreground. However, the object 60 that exists behind the lattice fence 50 may not be detected by the radar device 12.

  In this case, the distance between the object 60 from which the image portion 60a is detected and the vehicle 1 is not a true distance of 40 m as shown in FIG. If it is, it will be detected by mistake.

  Therefore, the real space position reliability determination unit 24 performs the real space position calculation unit for the object 60 that overlaps the lattice fence 50 when viewed from the vehicle 1 when the lattice fence 50 is detected by the translucent object detection unit 25. 23, the calculation of the real space position based on the distance (20 m) between the object 60 and the vehicle 1 detected by the object distance detection unit 21 is prohibited. This prevents an inappropriate real space position from being calculated based on an erroneous detection distance (20 m) between the object 60 and the vehicle 1.

  If the real space position of the object 60 is erroneously calculated at a position 20 m from the vehicle 1, the size of the image portion 60 a of the object 60 is close to the size assumed for a pedestrian 20 m away from the vehicle 1. There is a case. In this case, the contact avoidance process by the vehicle periphery monitoring device 1 is executed, which may cause the driver to feel uncomfortable or uneasy. However, an inappropriate real space position is calculated according to the first condition. It is possible to prevent such a sense of incongruity and anxiety from occurring.

  Further, as shown in FIG. 5, the translucent object detection unit 25 detects the image portion 50b having a lattice pattern from the captured image Im2 of the infrared camera 11 as an image portion of the wire mesh fence.

  Note that the distribution of the reflection intensity based on the distance measurement data output from the radar device 12 may be detected as a lattice fence for an object that is strong at the upper and lower ends and weak at the middle.

  Further, as shown in FIG. 6, the width in the horizontal direction (X-axis direction) is larger than the width of the pedestrian (the width of the pedestrian) based on the image data captured by the infrared camera 11 or the distance measurement data output from the radar device 12. When an object 50 that is equal to or greater than (set to a value) is detected, it may be detected that the object 50 is a lattice fence.

[Second Condition] Condition According to Degree of Blur of Image Part Referring to FIG. 7, the degree of blur of captured image 50c of object 50 in captured image Im3 of infrared camera 11 increases as the distance between object 50 and vehicle 1 increases. Become. Therefore, the blur degree calculation unit 26 calculates the blur degree for the image portion of the object detected by the image portion detection unit 22. Note that the degree of blur is calculated based on the luminance distribution of the image portion and the like.

  The real space position reliability determination unit 23 sets in advance a correspondence relationship between the distance between the object detected by the object distance detection unit 21 and the vehicle 1 and the degree of blur calculated by the degree-of-blur calculation unit 26. When the actual space position calculation unit 23 deviates more than a first predetermined level from the reference value of the distance-blur degree (set as the distance OOm-blur degree △ Δ%), the real space position calculation unit 23 performs the real space of the object. Prohibit position calculation.

  This prevents an inappropriate real space position from being calculated based on an erroneous detection distance (20 m) between the object 60 and the vehicle 1.

[Third Condition] Optical Flow Condition The displacement state calculation unit 27 is an optical flow of the image portion of the same object between time-series images captured for each control cycle by the infrared camera 11 (corresponding to the displacement state of the present invention). ) Is calculated. Note that the process of extracting the image portion of the same object from each time-series captured image is performed using, for example, the identity determination process described in Japanese Patent Laid-Open No. 2001-6096.

  As shown in FIG. 8A, when the infrared camera 11 is used with the lattice fence 50 (such as a wire fence) interposed between the object 60 and the infrared camera 11 and the radar device 12 of the vehicle 1. When the optical flow of the image portion between the series of captured images is calculated, it is as shown in V1 of FIG.

  V 2 in FIG. 8B is an estimated optical flow (based on the distance detected by the object distance detector 21, the distance between the vehicles 1, and the traveling speed of the vehicle 1 detected by the speed sensor 14 ( This corresponds to the estimated displacement state of the present invention).

  Here, when the detection distance by the object distance detection unit 21 is the distance between the object 60 and the vehicle 1, V1 and V2 are substantially equal. On the other hand, when the detection distance by the object distance detection unit 21 is the distance between the lattice fence 50 and the host vehicle 1, V1 is different from V2.

  Therefore, the real space position reliability determination unit 24 detects the object 60 by the real space position calculation unit 23 when the optical flow V1 calculated by the displacement state calculation unit 27 deviates from the estimated optical flow V2 by a second predetermined level or more. The calculation of the real space position of is prohibited. This prevents an inappropriate real space position from being calculated based on an erroneous detection distance between the object 60 and the vehicle 1 (detection distance between the lattice fence 50 and the vehicle 1).

[Fourth Condition] Condition Based on the Rate of Change of Size of Image Part The estimated distance calculation unit 28 is the rate of change of the size of the image part of the same object between time-series images captured for each control cycle by the infrared camera 11. Is calculated. Note that the process of extracting the image portion of the same object from each time-series captured image is performed using, for example, the identity determination process described in Japanese Patent Laid-Open No. 2001-6096.

  9A and 9B show a lattice fence 50 (such as a wire fence) between the object 60 and the infrared camera 11 and the radar device 12 of the vehicle 1, as shown in FIG. 8A. The image Im5 imaged at time t1 and the image Im6 imaged at time t2 (t2 = t1 + Δt) are shown.

  In this case, the estimated distance calculation unit 28 divides the height h2 of the image portion 60f of the object 60 at Im6 by the height h1 of the image portion 60e of the object 60 at Im5 to obtain the size of the image portion of the object 60. The rate of change (= h2 / h1) is calculated.

  Then, the real space position reliability determination unit 24 determines the distance between the object 60 and the vehicle 1 based on the change rate of the size of the image portion of the object 60 and the traveling speed of the vehicle 1 detected by the speed sensor 14. Is estimated.

  Here, when the detection distance by the object distance detection unit 21 is the distance between the object 60 and the vehicle 1, the detection distance is substantially equal to the estimated distance. On the other hand, when the detection distance by the object distance detection unit 21 is the distance between the lattice fence 50 and the host vehicle 1, this detection distance is different from the estimated distance.

  Therefore, the real space position reliability determination unit 24 detects when the detection distance by the object distance detection unit 21 deviates from the estimated distance between the object 60 and the vehicle 1 calculated by the estimated distance calculation unit 28 by a third predetermined level or more. The calculation of the real space position of the object 60 by the real space position calculation unit 23 is prohibited. This prevents an inappropriate real space position from being calculated based on an erroneous detection distance between the object 60 and the vehicle 1 (detection distance between the lattice fence 50 and the vehicle 1).

[Fifth Condition] Condition Based on Information of Navigation Device When the real space position reliability determination unit 24 identifies that the object from which the image part is detected by the image part detection unit 22 is a pedestrian, it is shown in FIG. As described above, walking on the map data Ma based on the position 71 of the vehicle 1 on the map data Ma obtained from the data from the navigation device 19 and the distance between the pedestrian and the vehicle 1 detected by the object distance detection unit 21. The position 72 of the person is calculated.

  When the pedestrian position 72 on the map data Ma is off the road, the real space position reliability determination unit 24 prohibits the real space position calculation unit 23 from calculating the pedestrian's real space position. This prevents an inappropriate real space position from being calculated based on an erroneous detection distance between the pedestrian and the vehicle 1.

  In this embodiment, the real space position reliability determination unit 24 determines whether or not the first to fifth conditions are satisfied, and the real space position calculation unit 23 determines the real space position based on the erroneous detection distance. Although the calculation is prevented from being performed, it is not always necessary to determine all the conditions, and the effect of the present invention can be obtained by determining at least whether or not the first condition is satisfied.

  In the present embodiment, the real space position reliability determination unit 24 determines whether or not the first condition to the fifth condition are satisfied, and prohibits the real space position calculation unit 23 from calculating the real space position of the object. The reliability of the real space position of the object calculated by the real sky position calculation unit 23 according to the determination result may be lowered. Then, it may be determined whether or not to adopt the real space position of the object calculated by the real space position calculation unit 23 according to the degree of reliability.

  In the present embodiment, an example in which a lattice fence is detected as a translucent object that shields a beam output from the radar device 12 and images other objects behind when captured by the infrared camera 11 is used. Although shown, the present invention can be applied to other types of translucent objects.

  DESCRIPTION OF SYMBOLS 1 ... Vehicle (own vehicle), 10 ... Vehicle periphery monitoring device, 11 ... Infrared camera, 12 ... Radar device, 16 ... Sound output device, 17 ... Image output device, 18 ... Braking device, 19 ... Navigation device, 21 ... Object Distance detection unit 22 ... Image part detection unit 23 ... Real space position calculation unit 24 24 Real space position reliability determination unit 25 25 Translucent object detection unit 26 26 Degree degree calculation unit 27 27 Displacement state calculation unit 28 ... Estimated distance calculation unit, 50 ... Lattice fence (semi-transparent object).

Claims (7)

An object distance detection unit that detects an object existing in the detection range by scanning a detection range around the vehicle by a radar device mounted on the vehicle, and detects a distance between the object and the vehicle;
An image part detection unit that projects an object detected by the object distance detection unit onto a captured image of a camera mounted on the vehicle and detects an image part of the object;
In a vehicle periphery monitoring device comprising: a real space position calculation unit that calculates a real space position of an object whose image portion is detected by the image portion detection unit based on a detection distance by the object distance detection unit;
A semi-transparent object detection unit that is within the detection range and shields a beam output from the radar apparatus and detects a semi-transparent object in which another object behind the image is imaged by the camera;
For an object existing in a position overlapping with the semi-transparent object detected by the semi-transparent object detection unit in the imaging direction of the camera, the real space for reducing the reliability of the real space position calculated by the real space position calculation unit A vehicle periphery monitoring device comprising a position reliability determination unit. In the vehicle periphery monitoring device according to claim 1,
A blur degree calculation unit that calculates the degree of blur of the image part detected by the image part detection unit;
The real space position reliability determination unit detects the degree of blur of the image part calculated by the degree-of-blur calculation unit and the object distance detection unit for the object whose image part is detected by the image part detection unit. Further, when the relationship between the distance between the object and the vehicle deviates from a preset reference relationship between the degree of blur and the distance by a first predetermined level or more, the real space position of the object by the real space position calculation unit Vehicle periphery monitoring device, which prohibits calculation of In the vehicle periphery monitoring device according to claim 1 or 2,
A displacement state calculation unit that calculates a displacement state of an image part of the same object detected by the image part detection unit from time-series captured images by the camera;
The real space position reliability determination unit is configured such that the displacement state calculated by the displacement state calculation unit is a change in the distance from the vehicle detected by the object distance detection unit for the same object, and the traveling of the vehicle. A vehicle periphery monitoring device that prohibits calculation of the real space position of the same object by the real space position calculation unit when the estimated displacement state calculated based on the speed deviates by a second predetermined level or more. . In the vehicle periphery monitoring device according to any one of claims 1 to 3,
The rate of change in the size of the image portion of the same object detected by the image portion detection unit is calculated from the time-series captured images by the camera, and the same rate is calculated based on the rate of change and the traveling speed of the vehicle. An estimated distance calculation unit for calculating an estimated distance between the object and the vehicle;
The real space position reliability determination unit is configured such that a distance between the same object and the vehicle detected by the object distance detection unit deviates from the estimated distance calculated by the estimated distance calculation unit by a third predetermined level or more. When this is done, the vehicle surroundings monitoring device prohibits the calculation of the real space position of the same object by the real space position calculation unit. In the vehicle periphery monitoring device according to any one of claims 1 to 4,
A navigation device mounted on the vehicle, including a host vehicle position calculation unit that calculates a position of the vehicle on a map;
The real space position reliability determination unit detects a pedestrian's image part by the image part detection unit, and calculates the position on the map of the vehicle calculated by the own vehicle position calculation unit and the object distance detection unit. When the position of the pedestrian on the map recognized based on the detected distance is off the road, the calculation of the real space position of the pedestrian by the real space position calculation unit is prohibited. A vehicle periphery monitoring device. In the vehicle periphery monitoring device according to any one of claims 1 to 5,
The vehicle periphery monitoring device, wherein the translucent object detection unit detects a lattice fence as the translucent object. In the vehicle periphery monitoring device according to claim 6,
The semi-transparent object detection unit detects that the object is a lattice fence when the width of the object calculated from the distance measurement data of the radar device is equal to or greater than a predetermined value. apparatus.

Images