JP2015225546A - Object detection device, drive support apparatus, object detection method, and object detection program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an object detection device, a drive support apparatus, an object detection method, and an object detection program, which enable accurate recognition of another vehicle entering a travel lane of a vehicle itself.SOLUTION: An object detection device includes: an imaging part for imaging a surrounding of a vehicle itself; a characteristic line extraction part for extracting first and second characteristic lines from an image picked up by the imaging part, where the first characteristic line is estimated as representing a side lower edge portion of another vehicle and the second characteristic line is estimated as representing a front edge portion of the other vehicle and has an inclination different from the first characteristic line on the image; and a determination part for determining an event of the other vehicle entering a travel road of the vehicle itself on the basis of an intersection between the first and second characteristic lines.

Description

  The present invention relates to an object detection device, a driving support device, an object detection method, and an object detection program.

  Based on the shooting means that is installed in the vehicle and shoots the road conditions ahead of the vehicle, and the image lane that is taken by the shooting means, the vehicle lane in which the vehicle is traveling and the adjacent to the vehicle lane Boundary line detecting means for detecting a boundary line with a lane, and a vehicle for recognizing an adjacent preceding vehicle traveling in an adjacent lane and detecting a traveling form of the adjacent preceding vehicle based on an image situation photographed by the photographing means Based on the detection results of the detection means, the boundary line detection means and the vehicle detection means, an inclination angle detection means for detecting the inclination of the traveling direction of the adjacent preceding vehicle relative to the boundary line, and the traveling direction of the adjacent preceding vehicle relative to the boundary line What is claimed is: 1. A vehicle recognition apparatus comprising: an interrupting vehicle determining means for determining whether an adjacent preceding vehicle is an interrupting vehicle that is about to interrupt an own vehicle lane based on an inclination angle Known (eg patent text Reference 1).

JP-A-9-223235

However, in the conventional apparatus, there is a case where the approach of another vehicle to the traveling lane of the own vehicle cannot be recognized with high accuracy. This is because simply monitoring the inclination angle of the adjacent preceding vehicle with respect to the boundary line may erroneously determine an interruption even if the other vehicle fluctuates.
The present invention has been made in consideration of such circumstances, and an object detection device, a driving support device, an object detection method, and an object detection method capable of accurately recognizing the entry of another vehicle into the traveling lane of the host vehicle, and An object is to provide an object detection program.

  The invention according to claim 1 is an imaging unit (10) that images the periphery of the host vehicle, and a first characteristic line that is estimated to represent a side lower end of another vehicle from an image captured by the imaging unit. , A feature line extraction unit (21, which extracts a second feature line that is estimated to represent a front end portion of the other vehicle and has a different inclination on the image from the first feature line. 22, 23, 24, 25, 30, 31) and a determination unit that determines the approach of another vehicle into the traveling lane of the host vehicle based on the intersection of the first feature line and the second feature line (40, 41, 42).

  Invention of Claim 2 is the object detection apparatus of Claim 1, Comprising: The said determination part is the travel lane of the said own vehicle from which the position of the said intersection can be recognized from the image imaged by the said imaging part, or In response to approaching, touching, or entering the scheduled travel path, the approach of another vehicle to the traveling lane of the host vehicle is determined.

  The invention according to claim 3 is the object detection device according to claim 1 or 2, further comprising a feature point extraction unit (21) that extracts a feature point from an image captured by the imaging unit, and the feature line extraction The unit extracts a set of feature points estimated to represent the same object between the feature points extracted from images of different imaging frames by the feature point extraction unit, and generates the combination of the feature points A feature line that is substantially in the same direction as the straight line is selected as the first feature line.

  Invention of Claim 4 is an object detection apparatus of any one of Claim 1 to 3, Comprising: The said feature line extraction part WHEREIN: When the several 2nd feature line candidate exists, From among a plurality of second feature line candidates, a second feature line candidate closest to the traveling lane of the host vehicle that can be recognized from the image captured by the imaging unit is selected as the second feature line. Is.

  A fifth aspect of the invention includes the object detection device according to any one of the first to fourth aspects, and a driving support unit that supports driving of the vehicle based on a determination result of the determination unit. It is a driving assistance device (1).

  According to a sixth aspect of the present invention, there is provided an imaging unit that captures an image of the periphery of the host vehicle, and an image captured by the imaging unit. A feature line extraction unit that extracts a feature line of the vehicle, a determination unit that determines an approach of another vehicle to the traveling lane of the host vehicle based on an intersection of the first feature line and the second feature line, , An object detection apparatus.

  According to the seventh aspect of the present invention, the object detection device captures an image of the periphery of the own vehicle, and the first feature line is estimated from the captured image to represent a side lower end of the other vehicle, and the other vehicle. A second feature line estimated to represent the front end of the first feature line and a second feature line having a different inclination on the image are extracted, and the first feature line and the second feature line are extracted. This is an object detection method for determining the approach of another vehicle to the traveling lane of the host vehicle based on the intersection with the feature line.

  According to an eighth aspect of the present invention, the control computer of the object detection apparatus has a first characteristic line estimated to represent a side lower end portion of another vehicle from an image obtained by imaging the periphery of the own vehicle, and a front end of the other vehicle. And extracting the first feature line and the second feature line that are different from the first feature line and having a different inclination on the image. This is an object detection program for determining the approach of another vehicle to the traveling lane of the host vehicle based on the intersection with the line.

  According to the first, seventh, and eighth aspects of the invention, the feature line extraction unit is estimated to represent the first feature line that is estimated to represent the side lower end portion of the other vehicle and the front end portion of the other vehicle. A second feature line that is different from the first feature line and having a different inclination on the image, and the determination unit determines whether the first feature line and the second feature line are Since the approach of the other vehicle to the traveling lane of the host vehicle is determined based on the intersection, the approach of the other vehicle to the traveling lane of the host vehicle can be accurately recognized.

  According to the second aspect of the present invention, the determination unit approaches, touches, or enters the travel lane or the planned travel path of the host vehicle whose position of the intersection is recognizable from the image captured by the imaging unit. Accordingly, since the approach of the other vehicle to the traveling lane of the host vehicle is determined, the approach of the other vehicle to the traveling lane of the host vehicle can be recognized with higher accuracy.

  According to the invention described in claim 3, the feature line extraction unit is a set of feature points estimated to represent the same object between the feature points extracted from images of different imaging frames by the feature point extraction unit. And the first feature line that is substantially in the same direction as the straight line generated by connecting the set of feature points is selected as the one that represents the side lower end of the other vehicle. Lines can be extracted. Thereby, the approach of the other vehicle to the traveling lane of the own vehicle can be recognized with higher accuracy.

  According to the fourth aspect of the present invention, the feature point extraction unit is the second closest to the traveling lane of the host vehicle that can be recognized from the image captured by the imaging unit among the plurality of second feature line candidates. By selecting the feature line candidate as the second feature line, the second feature line can be extracted with higher accuracy. Thereby, the approach of the other vehicle to the traveling lane of the own vehicle can be recognized with higher accuracy.

  According to the fifth aspect of the present invention, it is possible to provide suitable driving assistance based on the approach of another vehicle to the traveling lane of the host vehicle recognized with high accuracy by the object detection device.

It is a figure which shows typically an example of a structure of the driving assistance apparatus 1 containing the object detection apparatus 5 which concerns on one Embodiment of this invention. 2 is a diagram illustrating an example of a functional configuration of a driving support device 1 including an object detection device 5. FIG. It is the figure which illustrated feature point CP and vanishing point VP in captured image IM. It is a figure which expands and shows captured image IM, and is the figure which illustrated the adjacent vehicle flow in the captured image IM, the 1st feature line, the 2nd feature line, etc. It is a figure for demonstrating the example which selects an adjacent vehicle vertical edge from a some vertical edge. It is a flowchart which shows an example of the flow of the process performed by the control apparatus 20 of this embodiment.

〔Constitution〕
Hereinafter, an embodiment of a driving support device 1 including an object detection device 5 according to an embodiment of the present invention will be described with reference to the drawings.
FIG. 1 is a diagram schematically illustrating an example of a configuration of a driving support device 1 including an object detection device 5 according to an embodiment of the present invention. The driving support device 1 is a device mounted on the host vehicle M, for example, and includes a camera 10 and a control device 20. Hereinafter, a vehicle on which the object detection device 5 is mounted is referred to as “own vehicle”. Further, a lane adjacent to the traveling lane of the host vehicle is referred to as an “adjacent vehicle” that is traveling in the same direction as the host vehicle.

  The camera 10 is attached to the upper part of a front windshield, the back surface of a room mirror, etc., for example. The camera 10 is a digital camera using a solid-state image sensor such as a CCD (Charge Coupled Device) or a CMOS (Complementary Metal Oxide Semiconductor). For example, the camera 10 repeatedly captures an image of the front of the vehicle at a predetermined period and outputs data of the captured image to the control device 20. In addition, the camera 10 estimates the posture of the camera 10 according to the movement of the feature points and feature lines extracted by the control device 20.

  The control device 20 includes, for example, a processor such as a CPU (Central Processing Unit), a ROM (Read Only Memory), a RAM (Random Access Memory), a HDD (Hard Disk Drive), an EEPROM (Electrically Erasable Programmable Read-Only Memory), a flash, and the like. A computer device in which a storage device such as a memory, a communication interface for communicating with other devices in a vehicle, and the like are connected by an internal bus.

  FIG. 2 is a diagram illustrating a functional configuration example of the driving support device 1 including the object detection device 5. The control device 20 of the object detection device 5 includes a feature point extraction unit 21, a feature point correspondence determination unit 22, a posture estimation processing unit 23, an adjacent vehicle flow extraction unit 24, a flow grouping processing unit 25, and a feature line. An extraction unit 30, an adjacent vehicle feature line selection unit 31, an intersection coordinate calculation unit 40, a vehicle position calculation unit 41, and an interrupt determination unit 42 are provided. In addition, these functional units are software functional units that function, for example, when a processor executes a program stored in a storage device. These functional units may be integrated or divided as appropriate. The program executed by the processor may be stored in the storage device in advance when the host vehicle M is shipped, or the program stored in the portable storage medium may be installed in the storage device of the control device 20. Further, the program may be downloaded from another computer device by the in-vehicle Internet facility and installed in the storage device of the control device 20. Also, some or all of the functional units may be hardware functional units such as LSI (Large Scale Integration) and ASIC (Application Specific Integrated Circuit). The control device 20 is connected to the vehicle control unit 50 and the warning display control unit 52. The vehicle control unit 50 and the warning display control unit 52 may be functions of the same computer as the control device 20, or may be functions of a computer separate from the control device 20.

  Hereinafter, functions of the respective functional units of the control device 20 will be described with reference to FIG. The feature points, the first feature lines, the second feature lines, and their intersections will be described with reference to FIGS. 3 and 4 described later.

  The feature point extraction unit 21 extracts feature points from an image captured by the camera 10 (hereinafter referred to as “captured image”). A feature point is, for example, a pixel whose average value of brightness gradient between adjacent pixels in the top, bottom, left, and right is larger than a predetermined value, and contrasts image data captured in time series such as intersections of objects and intersections of lines. It is easy to take correspondence when you do. As a technique for extracting feature points, a known technique such as a Harris operator or a SUSAN operator, or a new technique can be used as appropriate. The functional unit corresponding to the feature point extracting unit 21 may be realized by a computer device built in or attached to the camera 10. FIG. 3 is a diagram illustrating feature points CP and vanishing points VP in the captured image IM. As shown in the drawing, the feature point CP tends to be extracted from the edge or corner of the vehicle body of the adjacent vehicle OB traveling around the subject vehicle that is the detection target.

  The feature point correspondence determination unit 22 determines the correspondence between the feature points extracted by the feature point extraction unit 21 between a plurality of captured images (for example, between two consecutive frames of captured images). The feature point correspondence determination unit 22 associates the feature point of the previous frame extracted by the feature point extraction unit 21 with the feature point of the current frame. As a method for determining the correspondence between feature points, the luminance distribution of a pixel group around the feature points extracted in the previous frame (for example, about 5 × 5 pixels or 7 × 7 pixels around the feature points) is obtained. It is possible to use a block matching method that searches for a pixel group having the same distribution as the luminance distribution around the same position of the image and extracts the center of the pixel group having the most consistent luminance distribution as the current feature point. In addition, a known technique of KLT (Kanade Lucas Tomasi) tracking and a new technique can be used as appropriate. Further, the feature point correspondence determining unit 22 calculates the three-dimensional coordinates of the feature points based on the correspondence relationship between the feature points of the image over a plurality of frames.

  FIG. 4 is an enlarged view of the captured image IM, and illustrates the adjacent vehicle flow, the first feature line, the second feature line, and the like in the captured image IM. In the example of FIG. 4, the adjacent vehicle OB captured at different times is represented by one frame for convenience. The adjacent vehicle OB0 is imaged at time T0, and the adjacent vehicle OB1 is imaged at time T1 after time T0.

  CPa0 and CPb0 in FIG. 4 are feature points representing different portions of the adjacent vehicle OB0 imaged at time T0. Further, CPa1 and CPb1 in FIG. 4 are characteristic points representing different portions of the adjacent vehicle OB1 imaged at time T1. The feature points CPa0 and CPa1 and the feature points CPb0 and CPb1 are feature points that represent the same part of the adjacent vehicle OB, respectively. The feature point correspondence determining unit 22 extracts a set of feature points that are estimated to represent the same object (or a specific portion of the object) between feature points extracted from images of different imaging frames. In the example of FIG. 4, the feature point correspondence determination unit 22 associates feature points CPa0 and CPa1 and feature points CPb0 and CPb1 to form a pair.

  The posture estimation processing unit 23 estimates the posture (pitch angle, pan angle, roll angle, etc.) of the camera 10 from the correspondence between the feature points between the plurality of captured images. In the example of FIG. 4, for example, the posture estimation processing unit 23 estimates the posture of the camera 10 from the relationship between the feature points CPa0 and CPa1 and the relationship between the feature points CPb0 and CPb1 that are associated by the feature point correspondence determination unit 22, for example. .

  The adjacent vehicle flow extraction unit 24 generates a straight line by connecting a set of feature points associated by the feature point correspondence determination unit 22. Hereinafter, this straight line is referred to as “adjacent vehicle flow”. Here, the adjacent vehicle flow is a directional line segment in space by connecting feature points associated by the feature point correspondence determining unit 22 between a plurality of captured images. For example, the adjacent vehicle flow extraction unit 24 has an inclination of a line segment within a predetermined angle with respect to a direction from the position of the line segment toward the vanishing point VP from the directed line segments on the space connecting the feature points. A directed line segment that is present and below the vanishing point VP on the image is extracted as an adjacent vehicle flow.

  In the example of FIG. 4, the adjacent vehicle flow extraction unit 24 connects the respective feature points “CPa0 and CPa1” and “CPb0 and CPb1” that are associated by the feature point correspondence determination unit 22 between the two captured images. “Fa” and “Fb”, which are directed line segments in space, are extracted as adjacent vehicle flows. The adjacent vehicle flows “Fa” and “Fb” are in the direction from the periphery of the image toward the vanishing point, and exist below the vanishing point on the image.

  The flow grouping processing unit 25 groups the adjacent vehicle flows extracted by the adjacent vehicle flow extraction unit 24. The flow grouping processing unit 25 groups the feature point flows by, for example, clustering processing. The flow grouping processing unit 25 performs grouping by clustering adjacent vehicle flows that can be regarded as the same object from the coordinate position, length, inclination, and the like of each adjacent vehicle flow extracted by the adjacent vehicle flow extraction unit 24. Since various methods are already known about this, description is abbreviate | omitted. In the example of FIG. 4, the flow grouping processing unit 25 can group each adjacent vehicle flow “Fa” and “Fb” as one group by clustering based on the coordinate position, length, and inclination. .

  The feature line extraction unit 30 has a “slanted edge” that is a feature line having an inclination close to the vanishing point VP in an IM range close to the adjacent vehicle flow group, and an inclination close to the vertical direction in the captured image IM. The feature lines “vertical edges” are extracted. “Near” means, for example, that an angle formed by a direction and an inclination is within a predetermined angle range. A straight line or a curve having a certain curvature or less is selected as the oblique edge and the vertical edge. From the positional relationship between the host vehicle and the adjacent vehicle, the side lower end portion of the adjacent vehicle is likely to be extracted as an oblique edge, and the front end portion of the adjacent vehicle is likely to be extracted as a vertical edge. The oblique edge and the vertical edge are characteristic lines having different inclinations. An oblique edge is, for example, a line in which pixels having a luminance gradient in a direction orthogonal to the direction toward the vanishing point in a region near the adjacent vehicle flow group in the captured image IM are larger than a predetermined value. Further, the vertical edge is, for example, a line connecting pixels in which the luminance gradient in the horizontal direction (horizontal direction) of the image is larger than a predetermined value in a region near the adjacent vehicle flow group in the captured image IM. Further, the feature line may be a line in which the feature points extracted by the feature point extraction unit 21 are connected, or may be a line extracted from the image by another method. As a technique for extracting feature lines, for example, a Canny technique, which is a known technique for detecting a line or contour line by feature point enhancement, or a new technique can be used as appropriate.

  The feature line extraction unit 30 extracts diagonal edges and vertical edges in a region close to the grouped adjacent vehicle flow groups “Fa” and “Fb” corresponding to the adjacent vehicle OB. Note that the feature line extraction unit 30 sequentially selects all the grouped adjacent vehicle flow groups, extracts diagonal edges and vertical edges in an area close to the selected adjacent vehicle flow group, and processing to be described later May be repeated.

  The adjacent vehicle feature line selection unit 31 is an adjacent vehicle oblique edge (first feature line) that is an oblique edge that is estimated from the oblique edge extracted by the feature line extraction unit 30 to represent the lateral lower end of the adjacent vehicle. Select. The “lower end” of the other vehicle corresponds to, for example, a rocker panel portion, a contact surface between the road surface and the other vehicle, a shadow of the other vehicle, and the like. In the example of FIG. 4, the characteristic line “A1” that has a slope close to the grouped adjacent vehicle flows “Fa” and “Fb” corresponding to the adjacent vehicle OB and is estimated to indicate the side lower end portion of the adjacent vehicle is , Selected as an adjacent vehicle diagonal edge.

  The adjacent vehicle feature line selection unit 31 is a vertical edge that is estimated from the vertical edge extracted by the feature line extraction unit 30 to represent the front end portion of the adjacent vehicle, and has a different inclination on the image from the adjacent vehicle oblique edge. The adjacent vehicle vertical edge (second feature line) is selected. Here, the “front end” of the adjacent vehicle OB is a portion formed near the front grill and / or near the front bumper of the adjacent vehicle OB. In the example of FIG. 4, the adjacent vehicle feature line selection unit 31 selects the feature line “B1” as the adjacent vehicle vertical edge.

  When only one vertical edge is extracted, the adjacent vehicle feature line selection unit 31 selects the one vertical edge as the adjacent vehicle vertical edge. However, when a plurality of vertical edges are extracted, The adjacent vehicle vertical edge is selected by the method illustrated in FIG. FIG. 5 is a diagram for describing an example in which an adjacent vehicle vertical edge is selected from a plurality of vertical edges. In the figure, A2 is an adjacent vehicle oblique edge selected by the adjacent vehicle feature line selection unit 31, and B2 and B3 are vertical edges extracted by the feature line extraction unit 30. In this case, the adjacent vehicle feature line selection unit 31 selects the vertical edge B2 closest to the traveling lane of the host vehicle that can be recognized from the image captured by the imaging unit 10 as the adjacent vehicle vertical edge.

  Here, the “vertical edge closest to the traveling lane of the host vehicle” is, for example, from a plurality of vertical edges (here, line segments having lengths) B2 and B3 to a lane line L1 that divides the traveling lane of the host vehicle. The shortest distances r2 and r3 are obtained, and the shortest distance (FIG. 5) is determined by selecting the vertical edge B2. The vertical edge including the point closest to the center line in the left-right direction on the captured image IM may be selected as the adjacent vehicle vertical edge.

  The intersection coordinate calculation unit 40 calculates the intersection of the adjacent vehicle diagonal edge and the adjacent vehicle vertical edge. In the example of FIG. 4, the intersection coordinate calculation unit 40 calculates an intersection X1 between the adjacent vehicle oblique edge A1 and the adjacent vehicle vertical edge B1.

  The vehicle position calculation unit 41 calculates the three-dimensional position of the intersection based on the position on the image of the intersection calculated by the intersection coordinate calculation unit 40 and the posture of the camera 10 estimated by the posture estimation processing unit 23.

  The interruption determination unit 42 determines whether or not the adjacent vehicle OB is an interruption vehicle based on the three-dimensional position of the intersection calculated by the vehicle position calculation unit 41. For example, the interrupt determination unit 42 determines whether or not the calculated three-dimensional position of the intersection and the planned travel path of the host vehicle (area partitioned by a line extending forward from the left and right ends of the host vehicle) overlap. If it is overlapped, it is determined that the adjacent vehicle OB is an interrupting vehicle. Moreover, the interruption determination unit 42 may determine that the adjacent vehicle OB is an interruption vehicle when the intersection crosses the lane line L1 (FIGS. 4 and 5) that divides the lane in which the host vehicle is traveling.

  The interrupt determination unit 42 may calculate an index value (continuous value or step value) indicating the possibility that the adjacent vehicle OB is an interrupted vehicle, instead of the above processing. In this case, the interrupt determination unit calculates a larger index value as the distance (FIG. 4, α and / or β) becomes shorter. Further, the interrupt determination unit 42, for example, the difference between the distance α and the distance β (that is, the temporal change in the distance between the intersection and the lane marking L1) and the degree of the change, the ratio between the distance α and the distance β, the moving direction of the intersection, and the like. May be combined to determine the possibility of the adjacent vehicle OB entering the lane in which the host vehicle is traveling.

  When the interrupt determination unit 42 calculates an index value or the like, the interrupt determination unit 42 may correct the index value based on various state values. For example, when the distance in the traveling direction between the host vehicle and the adjacent vehicle or the collision time (TTC; Time To Collision) is long, the interrupt determination unit 42 greatly corrects the possibility of the adjacent vehicle entering the host lane. If the distance in the traveling direction between the own vehicle and the adjacent vehicle or the collision time is short, the possibility of the adjacent vehicle entering the own lane may be corrected to be small. In this way, when the host vehicle and the adjacent vehicle are close to each other and the urgency is high, the possibility of the adjacent vehicle entering the lane on which the host vehicle is traveling is greatly corrected, thereby controlling the host vehicle. It is possible to determine the lane change of the adjacent vehicle in consideration of whether or not it is necessary.

  The vehicle control unit 50 performs various safety controls based on the position information of the adjacent vehicle calculated by the interrupt determination unit 42 so that the driver of the vehicle can drive safely. For example, the speed of the host vehicle may be controlled so as to keep the distance from the adjacent vehicle constant according to the position and speed of the adjacent vehicle detected by the control device 20, or based on the position of the detection object. Thus, automatic brake control and automatic steering control of the host vehicle may be performed.

  Based on the position information of the adjacent vehicle calculated by the interrupt determination unit 42, the warning display control unit 52 causes the display device such as a liquid crystal display to display information indicating caution or warning. In addition, information indicating attention or alerting may be transmitted to the driver by tightening a seat belt, an alarm sound, vibration, or the like.

[Operation flow]
FIG. 6 is a flowchart illustrating an example of a flow of processing executed by the control device 20 of the present embodiment. With reference to FIG. 6, the process which the driving assistance apparatus 1 performs is demonstrated.

  First, the feature point extraction unit 21 extracts feature points from the captured image (step S100). Next, the feature point correspondence determining unit 22 determines the correspondence relationship between the feature points extracted by the feature point extracting unit 21 between the plurality of captured images (step S102). Next, the posture estimation processing unit 23 estimates the posture (pitch angle, pan angle, roll angle, etc.) of the camera 10 from the correspondence between the feature points between the plurality of captured images. (Step S104). Next, the adjacent vehicle flow extraction unit 24 extracts the adjacent vehicle flow of the adjacent vehicle from the correspondence relationship of the feature points associated by the feature point correspondence determination unit 22. (Step S106). Next, the flow grouping processing unit 25 groups the adjacent vehicle flows extracted by the adjacent vehicle flow extraction unit 24. (Step S108). Then, the feature line extraction unit 30 includes a “diagonal edge” that is a feature line having an inclination close to the vanishing point VP in the captured image IM within the region close to the adjacent vehicle flow group of the selected group, and the captured image. Each “vertical edge”, which is a feature line having an inclination close to the vertical direction in IM, is extracted (step S110).

  Next, the adjacent vehicle feature line selection unit 31 is an adjacent vehicle oblique edge (first feature) that is an oblique edge estimated to represent the side lower end portion of the adjacent vehicle from the oblique edge extracted by the feature line extraction unit 30. Line) and an adjacent vehicle vertical edge (second feature line) estimated to represent the front end portion of the adjacent vehicle from the vertical edges extracted by the feature line extraction unit 30 is selected (step S112). Next, the intersection coordinate calculation unit 40 calculates the intersection of the adjacent vehicle diagonal edge and the adjacent vehicle vertical edge (step S114). Next, the vehicle position calculation unit 41 calculates the intersection coordinate value in the three-dimensional coordinates corresponding to the image from the intersection coordinate value and the posture of the camera 10 estimated by the posture estimation processing unit 23 (step S116). Next, the feature line extraction unit 30 determines whether there is an unselected adjacent vehicle flow group (step S118). If another grouped adjacent vehicle flow group exists, the process returns to step S110, and if another feature point flow group does not exist, the process proceeds to step S120.

  Next, the interruption determination unit 42 determines whether the vehicle is an interruption vehicle from the three-dimensional coordinate value of the intersection of adjacent vehicles (step S120). When it is determined that the vehicle is an interrupting vehicle, the vehicle control unit 50 performs various safety controls based on the position information of the adjacent vehicle calculated by the interrupt determining unit 42 so that the driver of the vehicle can drive safely. (Step S122). Thereby, the process of this flowchart is complete | finished.

  According to the object detection device 5 of the present embodiment described above, the control device 20 has an adjacent vehicle oblique edge (first characteristic line) estimated to represent the side lower end of the other vehicle, and the front end of the other vehicle. An adjacent vehicle vertical edge (second feature line) estimated to represent a part, and an adjacent vehicle oblique edge (first feature line) and a vertical edge (second feature line) having a different inclination on the image, And the interrupt determination unit 42 determines whether the other vehicle is in the traveling lane of the own vehicle based on the intersection of the adjacent vehicle diagonal edge (first feature line) and the adjacent vehicle vertical edge (second feature line). Since the approach and the approach are determined, it is possible to accurately determine the approach of another vehicle to the traveling lane of the host vehicle.

  Moreover, according to the driving assistance apparatus 1 of this embodiment, suitable driving assistance can be performed based on the approach of the other vehicle to the traveling lane of the host vehicle recognized by the object detection apparatus 5 with high accuracy. As a result, it can contribute to safe driving.

  As mentioned above, although the form for implementing this invention was demonstrated using embodiment, this invention is not limited to such embodiment at all, In the range which does not deviate from the summary of this invention, various deformation | transformation and substitution Can be added.

DESCRIPTION OF SYMBOLS 1 ... Driving assistance apparatus, 5 ... Object detection apparatus, 10 ... Camera, 20 ... Control apparatus, 21 ... Feature point extraction part, 22 ... Feature point corresponding | compatible determination part, 23 ... Posture estimation process part, 24 ... Neighboring vehicle flow extraction part 25 ... Flow grouping processing unit, 30 ... Feature line extraction unit, 31 ... Adjacent vehicle feature line selection unit, 40 ... Intersection coordinate calculation unit, 41 ... Vehicle position calculation unit, 42 ... Interrupt determination unit, 50 ... Vehicle control unit 52 ... Warning display control unit, M ... own vehicle, IM ... captured image, OB ... adjacent vehicle, CP ... feature point, VP ... vanishing point, Fa, Fb ... adjacent vehicle flow, A0, A1, A2 ... adjacent vehicle oblique edge (First feature line), B0, B1 ... adjacent vehicle vertical edge (second feature line), X0, X1 ... intersection, L1 ... partition line, B2, B3 ... vertical edge (second feature line candidate)

Claims (8)

An imaging unit for imaging the periphery of the vehicle;
A first feature line estimated from the image captured by the imaging unit to represent a side lower end portion of another vehicle and a second feature line estimated to represent a front end portion of the other vehicle; A feature line extraction unit for extracting the first feature line and a second feature line having a different inclination on the image;
A determination unit that determines an approach of another vehicle to the traveling lane of the host vehicle based on an intersection of the first feature line and the second feature line;
An object detection apparatus comprising: The determination unit determines whether the position of the intersecting point approaches, touches, or enters the travel lane or planned travel path of the host vehicle that can be recognized from the image captured by the imaging unit. Determining the entry of another vehicle into the vehicle lane,
The object detection apparatus according to claim 1. A feature point extraction unit that extracts a feature point from the image captured by the imaging unit;
The feature line extraction unit extracts a set of feature points estimated to represent the same object between feature points extracted from images of different imaging frames by the feature point extraction unit, and the set of feature points A feature line that is substantially in the same direction as a straight line generated by connecting the two is selected as the first feature line.
The object detection apparatus according to claim 1 or 2. When there are a plurality of second feature line candidates, the feature line extraction unit travels the host vehicle that can be recognized from an image captured by the imaging unit from among the plurality of second feature line candidates. Selecting the second feature line candidate closest to the lane as the second feature line;
The object detection device according to claim 1. The object detection device according to any one of claims 1 to 4,
A driving support unit that supports driving of the vehicle based on the determination result of the determination unit;
A driving support apparatus comprising: An imaging unit for imaging the periphery of the vehicle;
A feature line extraction unit that extracts a first feature line and a second feature line having a different inclination on the image from the image captured by the imaging unit;
A determination unit that determines an approach of another vehicle to the traveling lane of the host vehicle based on an intersection of the first feature line and the second feature line;
An object detection apparatus comprising: Object detection device
Take an image of the area around your vehicle,
From the captured image, a first feature line that is estimated to represent a lower side end portion of another vehicle and a second feature line that is estimated to represent a front end portion of the other vehicle, the first feature line And a second feature line having a different inclination on the image,
Based on the intersection of the first feature line and the second feature line, it is determined whether the other vehicle has entered the traveling lane of the host vehicle.
Object detection method. In the control computer of the object detection device,
A first feature line that is estimated to represent a lower side portion of a side of the other vehicle from an image of the periphery of the host vehicle, and a second feature line that is estimated to represent a front end of the other vehicle, Extracting a first feature line and a second feature line having a different inclination on the image;
Based on the intersection of the first feature line and the second feature line, the approach of another vehicle to the traveling lane of the host vehicle is determined.
Object detection program.

Images