JP2016176736A - Image distance measuring device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image distance measuring device for improving distance measuring accuracy in calculating a distance from a self vehicle to a target by using information where a coordinate position in an image is associated with a three-dimensional distance.SOLUTION: An image distance measuring device includes: a target detection section for detecting a target M in the circumference of a self vehicle N; an image acquisition section for acquiring an image 17 including the target M; a target area recognition section for recognizing a target area A1 being the area of the target M in the image 17; a shadow area recognition section for recognizing a shadow area A2 being the area having contact with the target area A1 in the image 17 and also the area of the shadow K of the target M in the image 17; a reference point setting section for setting a reference point P1 on a boundary L between the target area A1 and the shadow area A2; a storage section for storing information where a coordinate position in the image 17 is associated with a three-dimensional distance; and a distance calculation section for calculating a distance D from the self vehicle N to the target M.SELECTED DRAWING: Figure 2

Description

  The present invention relates to an image ranging device.

  As a parameter used for vehicle travel control, there is a distance between the host vehicle and an object existing around the host vehicle. For example, Patent Document 1 describes a distance measuring device that calculates a distance from a host vehicle to an object using an image captured by an in-vehicle camera of the host vehicle. This apparatus calculates the distance to an object by stereo processing using two images respectively acquired at the same timing by two cameras. If the parallax corresponding to all points that can be calculated from the object area that is the area of the object in the image is calculated, the calculation load is high and processing time is required. Only the parallax corresponding to the focused point of interest is calculated.

  On the other hand, if the setting density of the attention points is constant, the number of parallaxes calculated when the area of the object region in the image is small decreases, and the accuracy of the representative parallax of the object region calculated from these parallaxes is reduced. descend. Therefore, this distance measurement device sets the attention point setting density higher when the area of the object region in the image is small than when the area of the object region in the image is large.

JP 2009-180661 A

  By the way, as a distance measurement method using an image, the coordinate position of the reference point set for the object in the image is obtained from the own vehicle using information in which the coordinate position in the image is associated with the three-dimensional distance. There is a method for converting to a distance (three-dimensional distance) to an object. According to this method, it is possible to avoid performing complicated stereo processing as in a conventional distance measuring device. However, as has been a problem in the conventional distance measuring device, when the area of the object in the image is small, the accuracy of the distance from the host vehicle to the object decreases.

  Therefore, the image ranging apparatus of the present invention aims to improve the ranging accuracy in calculating the distance from the host vehicle to the object using information in which the coordinate position in the image is associated with the three-dimensional distance. And

  One aspect of the present invention is an object detection unit that detects an object around a host vehicle by using a detection result of a radar mounted on the host vehicle or an image captured by an in-vehicle camera mounted on the host vehicle. An image acquisition unit that acquires an image including an object using an in-vehicle camera mounted on the host vehicle, and an object that is a region of the object in the image using a detection result of the object detection unit Using the recognition result of the object area recognition unit and the object area recognition unit for recognizing the area, a shadow area that is in contact with the object area in the image and is a shadow area of the object in the image A shadow area recognizing unit for recognizing, a reference point setting unit for setting a reference point on a boundary between the object region and the shadow region using the recognition result of the object region recognizing unit and the recognition result of the shadow region recognizing unit; , Describe the information related to the coordinate position in the image and the 3D distance. Provided with a storage unit and, by using the stored information of the coordinate position and the storage unit of a reference point in the image, a distance calculating unit for calculating a distance from the vehicle to the object, the.

  According to the image ranging device of the present invention, it is possible to improve the ranging accuracy in calculating the distance from the host vehicle to the object using information in which the coordinate position in the image is associated with the three-dimensional distance. .

It is a block diagram which shows the image ranging apparatus which concerns on one form of this invention. (A) is a figure which shows the relationship between the own vehicle and a target object, (b) is an example of the image acquired in the image acquisition part. It is a flowchart which shows operation | movement of an image ranging apparatus. (A) is an example of an image acquired by the image acquisition unit, (b) is an example of an image showing a result of recognizing an object area, and (c) is an image of an image showing a result of recognizing a shadow area. It is an example and (d) is an example of an image showing a result of setting a reference point. It is a figure which shows another example of the image acquired in the image acquisition part.

  DESCRIPTION OF EMBODIMENTS Hereinafter, embodiments for carrying out the present invention will be described in detail with reference to the accompanying drawings. In the description of the drawings, the same elements are denoted by the same reference numerals, and redundant description is omitted.

  FIG. 1 is a block diagram showing an image distance measuring apparatus 1. The image ranging device 1 is mounted on a vehicle such as an automobile, for example, and measures the distance between the host vehicle on which the image ranging device 1 is mounted and an object existing around the host vehicle. Examples of the object include pedestrians, bicycles, animals, and other vehicles. The periphery of the host vehicle includes, for example, a front region of the host vehicle, a rear region and a side region of the host vehicle. The distance between the host vehicle and the object measured by the image ranging device 1 is used, for example, in a driving support device that supports driving of the host vehicle. In the driving support device, for example, the distance between the host vehicle and the target is used for collision avoidance support between the host vehicle and the target.

  The image ranging apparatus 1 includes an ECU [Engine Control Unit] 2 that performs processing related to ranging using an image. The ECU 2 is an electronic control unit including a CPU [Central Processing Unit], a ROM [Read Only Memory], a RAM [Random Access Memory], and the like. The ECU 2 loads a program stored in the ROM into the RAM and executes it with the CPU, thereby executing processing related to distance measurement. The ECU 2 may be composed of a plurality of electronic control units. The ECU 2 is connected to the vehicle external information acquisition unit 3, the display unit 5, the sound generation unit 6, and the actuator 7.

  The vehicle external information acquisition unit 3 has a function of acquiring information related to the outside around the host vehicle. Specifically, the vehicle external information acquisition unit 3 has a function of acquiring various types of information (object information) such as a structure that forms a blind spot area around the host vehicle, moving objects such as other vehicles, pedestrians, and bicycles. Have. The vehicle external information acquisition unit 3 includes, for example, an in-vehicle camera 3a such as a stereo camera or a monocular camera, and a radar 3b such as a millimeter wave radar. The vehicle external information acquisition unit 3 detects the presence / absence of a ranging object such as another vehicle by the radar 3b. Moreover, the vehicle external information acquisition part 3 acquires the image containing the target object detected by the radar 3b. In the present embodiment, the range imaged by the in-vehicle camera 3a is included in the detection range of the radar 3b.

  The display unit 5 includes, for example, a monitor or a head-up display, and has a function of displaying information for driving support. The sound generation unit 6 is configured by a speaker, a buzzer, and the like, and has a function of generating sound for driving support and a buzzer sound. The actuator 7 includes, for example, a brake actuator and a steering actuator (electric power steering system actuator). The brake actuator is an actuator that controls the braking force of the host vehicle, and the braking force is controlled according to a control signal from the ECU 2. The steering actuator is a steering actuator that controls the steering torque of the host vehicle, and the steering torque is controlled in accordance with a control signal from the ECU 2.

  Next, the functional configuration of the ECU 2 will be described. The ECU 2 includes an object detection unit 10, an image acquisition unit 11, an object region recognition unit 12, a shadow region recognition unit 13, a reference point setting unit 14, a storage unit 15, and a distance calculation unit 16. Have. Some of the functions of the ECU 2 described below may be executed in a computer or portable information terminal of a facility such as an information management center that can communicate with the host vehicle.

  The object detection unit 10 detects the object M present in front of the host vehicle N (see FIG. 2A). Here, in the following description, it is assumed that the bicycle Mb and the person Ma who rides the bicycle Mb are the object M. The object detection unit 10 detects the presence or absence of the object M using, for example, an image captured by the in-vehicle camera 3a or a detection result of the radar 3b included in the vehicle external information acquisition unit 3.

  The image acquisition part 11 acquires the image 17 (refer FIG.2 (b)) in which the target object M detected in the target object detection part 10 was contained using the vehicle-mounted camera 3a.

  The object area recognition unit 12 recognizes the object area A1 (see FIG. 2B) using the detection result of the object detection unit 10 and known image processing. Here, the object area A1 refers to an area occupied by the object M in the image 17 acquired by the image acquisition unit 11. The object area recognition unit 12 performs a plurality of candidate areas that are assumed to be occupied by a three-dimensional object included in the image 17 by performing known image processing such as edge extraction processing and pattern recognition on the image 17, for example. Extract. Then, using the detection result of the object detection unit 10, the area of the object M that is a distance measurement object is recognized as the object area A1 from the plurality of extracted candidate areas.

  The shadow area recognition unit 13 recognizes the shadow area A2 (see FIG. 2B) using the recognition result of the object area recognition unit 12. Here, the shadow area A2 is an area in contact with the object area A1 in the image 17 and an area of the shadow K of the object M in the image 17. The shadow area recognition unit 13 recognizes a plurality of shadow candidate areas in the image 17 using the luminance and chromaticity of each pixel constituting the image 17. Then, the shadow area recognition unit 13 recognizes the shadow area A2 of the object M from the plurality of shadow candidate areas using the recognition result of the object area recognition unit 12. For example, the shadow area recognition unit 13 may select a shadow area that exists in the vicinity of the object area recognition unit 12 as the shadow area A2 of the object M. The shadow area recognition unit 13 uses a position of a light source such as the sun or a street light to target a shadow area that is in the vicinity of the object area recognition unit 12 and exists in an expected range from the position of the light source. The shadow area A2 of the object M may be selected. The shadow area recognition unit 13 recognizes the position of the light source by, for example, known image processing.

  The reference point setting unit 14 uses the recognition result of the object region recognition unit 12 and the recognition result of the shadow region recognition unit 13 to use the reference point P1 (FIG. 2) on the boundary L between the object region A1 and the shadow region A2. (B) is set. Here, the reference point P1 is a point serving as a reference for obtaining a distance to the object M.

  The reference point setting unit 14 recognizes the boundary L between the object region A1 and the shadow region A2 using the recognition result of the object region recognition unit 12 and the recognition result of the shadow region recognition unit 13. The boundary L is, for example, a straight line passing through a plurality of contact points between the object area A1 and the shadow area A2. The boundary L may be a straight line that minimizes the distance to each contact point when there are three or more contact points between the object region A1 and the shadow region A2. Then, the reference point setting unit 14 sets the reference point P1 at an arbitrary position on the boundary L.

  The storage unit 15 stores information in which the coordinate position in the image 17 is associated with the three-dimensional distance. Here, the coordinate position refers to information indicating a position specified based on a two-dimensional coordinate system in the image 17. For example, the position in the image 17 is the x coordinate when the horizontal direction of the image 17 is the X axis direction, the vertical direction of the image is the Y axis direction, and the lowest point on the left side of the image 17 is set as the origin OP. And y-coordinates. The three-dimensional distance refers to a distance from the host vehicle N (the in-vehicle camera 3a) in a three-dimensional space (three-dimensional coordinate system). Here, the distance D shown in FIG. 2A is a three-dimensional distance along the traveling direction of the host vehicle N from the in-vehicle camera 3a of the host vehicle N to the point P1a where the object M and the road surface R contact. The point P1a is a point on the road surface R in the three-dimensional space, and is the point closest to the host vehicle N among the points where the object M and the road surface R are in contact.

  The information in which the coordinate position is associated with the three-dimensional distance is information for converting the coordinate position into the three-dimensional distance. In other words, the coordinate position of the reference point P1 set on the road surface R is a tertiary along the road surface R up to the point P1a on the road surface R with which the object M is in contact, based on the information recorded in the storage unit 15. It is associated with the original distance. Here, the relationship between the coordinate position and the three-dimensional distance will be described. As shown in FIG. 2, the in-vehicle camera 3 a is fixed to the host vehicle N, and the height of the in-vehicle camera 3 a relative to the road surface R and the angle of view of the in-vehicle camera 3 a are determined while the host vehicle N is traveling. Maintained. Then, for example, when the relative position between the host vehicle N and the object M does not change, the object M exists at the same position in the image 17. Therefore, if the relationship between the coordinate position in the image 17 and the distance D from the host vehicle N is obtained in advance by actual measurement or design, the distance D from the host vehicle N to the object M is used using the coordinate position in the image 17. Can be calculated. The information associating the coordinate position with the three-dimensional distance may be, for example, a function having the coordinate position as an independent variable and the three-dimensional distance as a dependent variable. The information that associates the coordinate position with the three-dimensional distance may be a database such as a conversion table in which the three-dimensional distance corresponding to the coordinate position is combined.

  The distance calculation unit 16 calculates the distance D from the host vehicle N to the object M using the coordinate position of the reference point P1 in the image 17 and the information stored in the storage unit 15. The distance calculation unit 16 acquires the position coordinates of the reference point P1 set by the reference point setting unit 14. Next, the distance calculation unit 16 refers to the storage unit 15 and acquires a function or conversion table for converting the position coordinates of the reference point P1 into a three-dimensional distance. And the distance calculation part 16 calculates the distance D from a position coordinate using a function or a conversion table.

  Next, the operation of the image ranging apparatus 1 will be described with reference to FIG. FIG. 3 is a flowchart showing the operation of the image ranging apparatus 1 according to the embodiment. The flowchart shown in FIG. 3 is repeatedly executed while the host vehicle N is traveling, for example. When the driver finishes driving the vehicle N, the flowchart shown in FIG.

  In step S <b> 1, the image ranging device 1 acquires vehicle external information around the host vehicle N by the vehicle external information acquisition unit 3. Next, in step S2, the ECU 2 detects the object M by the object detection unit 10 using the vehicle external information. Specifically, when the object detection unit 10 determines that the object M is not detected using the vehicle external information acquired by the vehicle external information acquisition unit 3 (S2: NO), step S1 is performed. , S2 is performed again. If the object detection unit 10 determines that the object M has been detected (S2: YES), the object detection unit 10 proceeds to the next step S3.

  In step S3, the ECU 2 of the image ranging device 1 acquires the image 17 by the image acquisition unit 11 (see FIG. 4A). Subsequently, in step S4, the ECU 2 recognizes the object area A1 by the object area recognition unit 12 (see FIG. 4B). Specifically, the object area recognition unit 12 recognizes the object area A1 in the image 17 using the vehicle external information.

  The ECU 2 of the image ranging apparatus 1 recognizes the shadow area A2 by the shadow area recognition unit 13 in step S5 (see FIG. 4C). Specifically, the shadow area recognition unit 13 executes a recognition process of the shadow area A2 of the object M in the image 17 using the recognition result of the object area A1. When the shadow area recognition unit 13 can recognize the shadow area A2 in the image 17 (S5: YES), the process proceeds to the next step S8. On the other hand, when the shadow area recognition unit 13 cannot recognize the shadow area A2 in the image 17 (S5: NO), the process proceeds to another next step S6.

  First, when the result of step S5 is NO, the ECU 2 of the image ranging apparatus 1 recognizes the road surface area occupied by the road surface in the image 17 in step S6. The road surface area can be recognized by, for example, known image processing. Subsequently, the ECU 2 sets a reference point in step S7. Specifically, the ECU 2 recognizes the boundary between the object area A1 and the road surface by known image processing using the object area A1 and the road surface area. Subsequently, the ECU 2 sets a point closest to the host vehicle N at the boundary (that is, a point closest to the lower end of the image 17 on the boundary line) as a reference point.

  On the other hand, if the result of step S5 is YES, the ECU 2 of the image ranging apparatus 1 sets the reference point P1 on the boundary L between the object area A1 and the shadow area A2 by the reference point setting unit 14 in step S8. Set. Specifically, the reference point setting unit 14 recognizes the boundary L between the object area A1 and the shadow area A2 by using the object area A1 and the shadow area A2. Subsequently, the ECU 2 sets, on the boundary L, a point closest to the host vehicle N (that is, a point closest to the lower end of the image 17 on the boundary line) as the reference point P1. In step S <b> 9, the ECU 2 of the image ranging device 1 acquires the coordinate position of the reference point P <b> 1 by the distance calculation unit 16. In step S9, the reference point setting unit 14 may acquire the coordinate position of the reference point P1.

  In step S <b> 10, the ECU 2 of the image distance measuring device 1 calculates a distance D from the host vehicle N to the object M by the distance calculation unit 16. The distance calculation unit 16 uses the information in which the coordinate position recorded in the storage unit 15 and the three-dimensional distance are associated with each other to convert the position coordinate of the reference point P1 into the three-dimensional distance. A distance D to the object M is calculated.

  In the image ranging device 1 described above, the distance calculation unit 16 uses the information in which the coordinate position in the image 17 is associated with the three-dimensional distance to determine the coordinate position of the reference point P1 from the own vehicle N (vehicle camera 3a). It is converted into a distance D to the object M. The reference point P1 is set on the boundary L between the object area A1 and the shadow area A2 in the image 17. Since this boundary L is formed on the road surface R, it can be said that the reference point P1 is set on the road surface R. Thereby, in the image ranging apparatus 1, by using the reference point P1 set on the road surface R, the camera of the back of the person Ma riding the bicycle Mb by the conventional ranging method (recognized by the camera of the object M). The distance from the vehicle N to the object M compared to the case of using a reference point set to an area that is easy to carry out), that is, the case of using a reference point floating away from the road surface R. The accuracy of calculating D can be improved. Therefore, according to the image ranging device 1, the ranging accuracy is improved in the calculation of the distance from the own vehicle N to the object M using the information that associates the coordinate position in the image 17 with the three-dimensional distance. be able to.

  Also, in the object ranging process, if the object is a thin rod-shaped solid object or a relatively soft three-dimensional object, sufficient ranging accuracy may not be obtained with a radar such as a millimeter wave radar. possible. According to the image ranging device 1 described above, since ranging to an object is performed using an image, it is possible to improve the ranging accuracy even for a thin rod-shaped three-dimensional object or a relatively soft three-dimensional object. Can do. For example, as shown in FIG. 5, even in the case of an animal 19 such as a deer having thin legs and a soft torso, an object region 19a occupied by the animal 19 and a shadow region 19c occupied by a shadow 19b of the animal 19 are provided. By recognizing and setting the reference point P2 at the boundary between the object area 19a and the shadow area 19c, the ranging accuracy up to the animal 19 can be improved.

  In addition, as in the conventional distance measuring device disclosed in Patent Document 1 described above, when the area of the object in the image is small, an image with a high resolution is required to set the setting density of the attention point and the corresponding point high. Required. A configuration for obtaining an image with a high resolution and a configuration for processing the image are disadvantageous in cost because high performance is required. On the other hand, according to the image distance measuring device 1, since the configuration and performance required for the distance measuring device of Patent Document 1 are not required, an increase in manufacturing cost can be suppressed.

  The present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the gist of the present invention.

  For example, the image ranging device 1 detects the presence or absence of the object M using the radar 3b. However, the presence or absence of the object M is detected using an image captured by the in-vehicle camera 3a. Also good. In addition, the camera that acquires the image for detecting the object M and the camera that acquires the image for recognizing the object area A1 and the shadow area A2 may be the same camera, or different It may be a camera.

DESCRIPTION OF SYMBOLS 1 ... Image ranging device, 2 ... ECU, 3 ... Vehicle external information acquisition part, 3a ... Car-mounted camera, 3b ... Radar, 10 ... Object detection part, 11 ... Image acquisition part, 12 ... Object area recognition part, 13 ... shadow area recognition section, 14 ... reference point setting section, 15 ... storage section, 16 ... distance calculation section, N ... own vehicle, M ... object, A1 ... object area, A2 ... shadow area, P1, P2 ... reference point.

Claims (1)

An object detection unit that detects an object around the host vehicle using a detection result of a radar mounted on the host vehicle or an image captured by an in-vehicle camera mounted on the host vehicle;
Using an in-vehicle camera mounted on the host vehicle, an image acquisition unit that acquires an image including the object;
Using the detection result of the object detection unit, an object region recognition unit that recognizes an object region that is the region of the object in the image;
A shadow area recognition unit that recognizes a shadow area that is in contact with the object area in the image and is a shadow area of the object in the image by using a recognition result of the object area recognition unit. When,
A reference point setting unit that sets a reference point on a boundary between the object region and the shadow region using the recognition result of the object region recognition unit and the recognition result of the shadow region recognition unit;
A storage unit storing information associated with a coordinate position and a three-dimensional distance in the image;
A distance calculation unit that calculates a distance from the host vehicle to the object using the coordinate position of the reference point in the image and the information stored in the storage unit;
An image distance measuring device.

Images