JP2018120303A - Object detection device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a technique for reducing a load on image processing.SOLUTION: A control unit 15 acquires a captured image captured by a camera 11 mounted in a vehicle. The control unit 15 sets a plurality of neighboring positions which are the positions separated from a target position by a predetermined distance and identifies a position corresponding to each of the target position and the neighboring position in the acquired captured image as a candidate position based on a relative position of the target position, with an estimated position as a reference. The control unit 15 determines whether or not the candidate position represents the position of a target object in the captured image. When the candidate position represents the position of the target object in the captured image, the control unit 15 causes a memory to record the candidate position as a detection position. The target object is the object arranged along a direction in which a road on which the vehicle travels extends. The neighboring position is the position within a predetermined range including the target position. The candidate position is the position where the target object can exist in the captured image.SELECTED DRAWING: Figure 1

Description

  The present disclosure relates to a technique for detecting a specific object from an image captured by a camera provided in a vehicle.

  2. Description of the Related Art Conventionally, a technique for detecting a specific object from an image captured by a camera provided in a vehicle is known (see Patent Document 1).

Japanese Patent Laid-Open No. 2006-143324

  In the related art as described in Patent Document 1, when detecting an object from an image, the object is detected from the image by image processing such as known pattern matching in the entire range of the image. Is common.

However, in such a conventional technique, image processing such as pattern matching is performed on the entire range of the image, so that there is a problem that the load of image processing is heavy.
One aspect of the present disclosure provides a technique for reducing the load of image processing.

  One aspect of the present disclosure is an object detection device (15) mounted on a vehicle. The object detection device includes a vehicle position part (S115), a target position part (S120), an image acquisition part (S130), a basic part (S135), a determination part (S160), and a recording instruction part (S165). .

  A vehicle position part acquires the estimated position which is the position of the vehicle estimated by the position estimation apparatus from the position estimation apparatus (12) which estimates the position of a vehicle. The target position unit acquires the target position that is the position of the target object from the map storage unit (13) that stores map data including the position of the predetermined target object. An image acquisition part acquires the picked-up image imaged with the camera (11) mounted in the vehicle.

  The basic unit sets a neighboring position with respect to the target position, and based on the relative position of the target position with respect to the estimated position, candidates corresponding to the target position and the neighboring position in the captured image are selected. Specify as location.

  The determination unit determines whether the candidate position represents the position of the object in the captured image. When the candidate position represents the position of the target object in the captured image, the recording instruction unit causes the recording unit (52) to record the candidate position as the detection position.

  The target object is an object arranged along the direction in which the road on which the vehicle travels extends. The proximity position is a position included within a predetermined range from the target position. The candidate position is a position where an object can exist in the captured image.

  According to such a configuration, since the captured image has a configuration for determining whether or not only the candidate position corresponding to the target position and the neighboring position based on the map data represents the position of the target object, image processing is performed. Can be reduced.

  Note that the reference numerals in parentheses described in this column and in the claims indicate the correspondence with the specific means described in the embodiment described later as one aspect, and the technical scope of the present disclosure It is not limited.

The block diagram which shows the structure of a vehicle-mounted system. The flowchart which shows a detection process. The figure explaining the object position and neighborhood position based on map data. The figure explaining an image candidate position. The flowchart which shows the addition process of 1st Embodiment. The figure which shows the reference | standard brightness distribution of a white line. The figure explaining a mode that the degree of matching with reference | standard brightness distribution is detected. The figure explaining the action | operation of a control part. The figure explaining that a new detection position is specified according to the moving amount | distance of the own vehicle. The figure explaining the comparative example by which the image candidate position does not represent the position of the white line which is a target object by the estimation error by a positioning apparatus. The flowchart which shows the addition process of 2nd Embodiment. The figure explaining specification of the present image position based on an estimated line.

Hereinafter, embodiments of the present disclosure will be described with reference to the drawings.
[1. First Embodiment]
[1-1. Constitution]
An in-vehicle system 10 shown in FIG. 1 is a system mounted on a vehicle, and is a system that detects a predetermined object from captured images within a predetermined range around the vehicle. Hereinafter, a vehicle on which the in-vehicle system 10 is mounted is also referred to as a host vehicle. The imaging range referred to below corresponds to a predetermined range around the vehicle.

  The object here refers to an object to be detected by the in-vehicle system 10 and disposed along the direction in which the road extends. The objects arranged along the road extending direction may include objects such as white lines and yellow lines on the road, guardrails, curbs, sound barriers, and the like.

  The in-vehicle system 10 includes a control unit 15. The in-vehicle system 10 may include a camera 11, a positioning device 12, a movement amount detection device 13, a map data recording unit 14, and a vehicle control unit 16.

  The camera 11 is installed at the front end of the host vehicle so as to capture the front of the host vehicle including the road surface in the traveling direction. Here, the front refers to the direction in which the windshield is present when viewed from the center of the host vehicle. A predetermined angle range including the front in the traveling direction of the host vehicle is set as the imaging range of the camera 11. In addition, the installation position and imaging range of the camera 11 are not limited to this. The camera 11 can be installed at an arbitrary position such as a rear end or a side end of the host vehicle. In addition, the imaging range may include an arbitrary range around the host vehicle, such as behind the host vehicle or the side of the host vehicle.

  The positioning device 12 outputs the estimated position to the control unit 15. The estimated position is the current position of the host vehicle estimated by the positioning device 12. In the present embodiment, the positioning device 12 has a GNSS receiving device (not shown). The GNSS receiver identifies the position of the host vehicle based on the GNSS signal, and outputs the identified position to the control unit 15 as an estimated position. GNSS is an abbreviation for Global Navigation Satellite System.

  The positioning device 12 is not limited to this, and may have a GPS receiver instead of the GNSS receiving device. GPS is an abbreviation for Global Positioning System. The positioning device 12 may have an autonomous sensor including sensors such as a gyroscope, a vehicle speed sensor, and a G sensor (not shown). That is, the positioning device 12 can be configured to calculate the estimated position by a known autonomous navigation based on the detection signal from the autonomous sensor.

  The movement amount detection device 13 calculates the movement amount of the host vehicle and outputs it to the control unit 15. The moving amount of the host vehicle here refers to a distance traveled by the host vehicle and a change in the traveling direction of the host vehicle. The movement amount detection device 13 calculates the movement amount every predetermined time. Specifically, the movement amount detection device 13 includes a vehicle speed sensor and a yaw rate sensor (not shown). The vehicle speed sensor detects the vehicle speed of the host vehicle. The yaw rate sensor detects the yaw rate of the host vehicle. The movement amount detection device 13 calculates the movement amount based on the detection results of the vehicle speed sensor and the yaw rate sensor.

  The map data recording unit 14 is composed of an HDD and records map data 141 and the like. HDD is an abbreviation for hard disk drive. The map data 141 includes information about the road such as the position of the road, the shape of the road, the number of lanes on the road, the position of the intersection, and the like.

  The map data 141 includes information representing the position of the object detected by the in-vehicle system 10. For example, for the objects arranged along the road extending direction, such as lane markings, guardrails, curbs, etc., the map data 141 includes position information for every several to several tens of meters. The position information is represented by latitude and longitude.

The vehicle control unit 16 controls the engine, brakes, steering, and the like of the host vehicle in accordance with a signal from the control unit 15.
The control unit 15 is mainly configured by a known microcomputer having a CPU 51 and a semiconductor memory (hereinafter, memory 52) such as a RAM, a ROM, and a flash memory. Various functions of the control unit 15 are realized by the CPU 51 executing a program stored in a non-transitional physical recording medium. In this example, the memory 52 corresponds to a non-transitional tangible recording medium that stores a program. Also, by executing this program, a method corresponding to the program is executed. The number of microcomputers constituting the control unit 15 may be one or plural.

  The control part 15 implement | achieves various functions, such as detection of a target object, when CPU51 runs programs, such as a detection process mentioned later. The method of realizing these elements constituting the control unit 15 is not limited to software, and some or all of the elements may be realized using one or a plurality of hardware. For example, when the above function is realized by an electronic circuit that is hardware, the electronic circuit may be realized by a digital circuit including a large number of logic circuits, an analog circuit, or a combination thereof.

[1-2. processing]
Next, detection processing executed by the control unit 15 will be described with reference to the flowchart of FIG. The detection process is a process of detecting an object from the captured image of the camera 11. Specifically, the detection process is a process of specifying a detection position that is a position of an object in a captured image. Below, the example which detects the white line which is a lane marking in a road as a target object is demonstrated. The control unit 15 repeatedly executes the detection process for each processing cycle that is a predetermined cycle while the vehicle-mounted system 10 is powered on.

  First, in S110, the control unit 15 specifies the type of the object detected by the in-vehicle system 10. The type of the object refers to the type of the object such as a white line, a yellow line, a curb, a guardrail, and the like. In the present embodiment, information indicating the type of the object (hereinafter referred to as type information) is recorded in the memory 52 in advance. The control unit 15 acquires the type information and specifies the type of the object. In the present embodiment, the type information indicates that the type of the object is a white line.

  In addition, it is not limited to this, The kind of target object detected by the vehicle-mounted system 10 can be input using the input device which is not illustrated by the driver | operator of a vehicle, for example. And the control part 15 can pinpoint the kind of target object based on the information input using the input device.

In S115, the control unit 15 acquires the estimated position from the positioning device 12. As shown in FIG. 3, the estimated position P is represented using latitude and longitude, for example (x ₀ , y ₀ ).

  In S120, the control unit 15 acquires the target position from the map data recording unit 14. The target position is the position of the target based on the map data 141. In the present embodiment, as shown in FIG. 3, the control unit 15 includes a white line 202 that is a lane line between the own lane 101 that is the lane in which the host vehicle 1 travels and an adjacent lane 102 that is adjacent to the own lane 101. A plurality of white lines such as a white line 201 opposite to the white line 202 in 101 and a white line 203 opposite to the white line 202 in the adjacent lane 102 are set as objects.

  The position of the white line is recorded in the map data 141 every several meters. The control unit 15 acquires, as target positions, the positions of the white lines recorded in the map data 141 and the positions of a plurality of white lines within the imaging range Z of the camera 11 mounted on the host vehicle 1. In FIG. 3, the target position A is indicated by a black circle within the range surrounded by the dotted line 300.

  In S125, the control unit 15 sets a vicinity position with respect to the target position A. The vicinity position is a position included in a predetermined range including the target position A. In FIG. 3, the vicinity position B is indicated by a white circle within the range surrounded by the dotted line 300.

  In the present embodiment, as shown in FIG. 3, the control unit 15 sets a position included within a range that is a predetermined distance Dx that is a predetermined distance from the white line to the right and left in the horizontal direction as the vicinity position B. To do. That is, the white line is configured by arranging a plurality of target positions A, and the control unit 15 determines positions included in a range that is separated from the target position A that configures the white line by a specified lateral distance Dx to the right and left in the horizontal direction. Set as the proximity position B.

  The lateral direction here refers to a direction perpendicular to the direction in which the road extends on the plane represented by the map. In the present embodiment, an arbitrary value smaller than the error of the estimated position P by the positioning device 12 (hereinafter, estimated error by the positioning device 12) is set as the designated lateral distance Dx. The designated lateral distance Dx is recorded in the memory 52 in advance.

The control unit 15 sets at least one neighboring position B for each white line in the imaging range Z.
In addition, when the control part 15 sets the some vicinity position B with respect to a white line, it is the distance between the adjacent vicinity positions B, Comprising: The distance (henceforth, vertical distance) dy in the road extending direction is arbitrary values. Can be set.

  In the present embodiment, the control unit 15 sets the vertical distance dy between adjacent neighboring positions B to a random value that is less than the estimation error by the positioning device 12. However, the present invention is not limited to this. For example, the vertical distance dy between adjacent neighboring positions B can be set to a predetermined value that is less than the estimation error by the positioning device 12.

Returning to FIG. In S130, the control unit 15 acquires a captured image captured by the camera 11.
In S135, the control unit 15 specifies an image candidate position. Based on the relative position of the target position A relative to the estimated position P and the relative position of the neighboring position B relative to the estimated position P, the control unit 15 performs the target position A expressed in latitude and longitude. And the position of the vicinity position B is converted into image coordinates. Image coordinates are coordinates in a captured image. The control unit 15 performs conversion into image coordinates using the camera parameters of the camera 11. The camera parameter represents the mounting position and mounting posture of the camera 11 and is recorded in the memory 52 in advance.

  The control unit 15 specifies these image coordinates as image candidate positions that are positions corresponding to the target position A and the neighboring position B in the captured image acquired in S130. The image candidate position represents a position where a white line can exist in the captured image. The control unit 15 records the specified image candidate position in the memory 52.

  In the captured image shown in FIG. 4, the image candidate position M is indicated by a white circle within a range surrounded by a dotted line 301. That is, after the end of this step, the position of each white circle within the range surrounded by the dotted line 301 in FIG. 4 is recorded in the memory 52 as the image candidate position M.

  Returning to FIG. In S140, the control unit 15 determines whether or not the captured image acquired in S130 is the first frame. The first frame represents a captured image captured for the first time and represents that the detection position has not been detected yet. Here, in the case of the first frame, the control unit 15 shifts the processing to S150 without setting anything as an image addition position to be described later. On the other hand, if it is not the first frame, the control unit 15 shifts the process to S145.

  In S145, the control unit 15 performs an additional process. The additional processing is processing for specifying the current image position and the image preliminary position in the captured image. The controller 15 repeatedly records the detected position in the memory 52 for each processing cycle by executing this detection process. The current image position is a position corresponding to a past detection position recorded in the memory 52 in the captured image acquired in S130. The image preliminary position is a position within a predetermined range including the current image position. After recording the current image position and the image preliminary position in the memory 52 as the image addition position, the control unit 15 shifts the processing to S150.

Here, the addition process will be described with reference to the flowchart of FIG.
In S200, the control unit 15 acquires the movement amount of the host vehicle from the movement amount detection device 13. In the present embodiment, the movement amount detection device 13 is configured to output a movement amount for each processing cycle, and the cycle in which the movement amount is output is synchronized with the cycle in which the movement amount is acquired in S200. .

  In S <b> 210, the control unit 15 acquires the detection position of the white line detected in the “past processing cycle” from the memory 52. Hereinafter, the detection position of the white line is also simply referred to as a detection position. In addition, the detection position detected in the past processing cycle is also simply referred to as the past detection position.

  In the present embodiment, the above-mentioned “past processing cycle” refers to a past processing cycle one prior to the current processing cycle (hereinafter, the current cycle) in which the main detection processing is executed. In addition, the above-described past detection position refers to a detection position detected in a past processing cycle immediately before the current cycle. The detection position is represented by image coordinates in the captured image.

In S220, the control unit 15 specifies the current image position based on the movement amount of the host vehicle acquired in S200 and the past detection position acquired in S210. Specifically, as shown in FIG. 8, which will be described later, the control unit 15 sets the past detection position C, which is the detection position in the past processing cycle T _n−1 acquired in S210, as the amount of movement of the host vehicle. Is reflected and converted to the current image position D, which is the position on the captured image in the captured image acquired in S130. The current image position D is represented by latitude and longitude. Hereinafter, the captured image acquired in S130 is also referred to as a captured image of the current period.

In S230, the control unit 15 records each of the current image positions D specified in S220 in the memory 52 as image addition positions.
Further, in S240, the control unit 15 sets an image preliminary position. The control unit 15 sets a position within a predetermined range including the current image position D as the image preliminary position E.

  As shown in FIG. 8, the predetermined range J here refers to a range indicated by a circle whose radius is a predetermined image designated distance centered on the current image position D. The designated image distance is a distance between pixels in a captured image.

  In this embodiment, the image designation distance is a distance between pixels in the captured image and is set to a size corresponding to an estimation error by the positioning device 12. However, it is not limited to this. For example, the predetermined range can be set to an arbitrary shape. Further, the processing range can be set to an arbitrary size.

  The control unit 15 can set at least one image preliminary position E within a predetermined range J for one current image position D. In the present embodiment, the control unit 15 sets a plurality of image preliminary positions E within a predetermined range J with respect to the current image position D. FIG. 8 shows only one image preliminary position E with respect to the current image position D.

  In S250, the control unit 15 records each of the image addition positions recorded in the memory 52 in S230 and each of the image preliminary positions E set in S240 in the memory 52 as image addition positions. And this addition process is complete | finished. As a result, each of the current image position D and each of the plurality of preliminary image positions E set around the current image position D are recorded in the memory 52 as image addition positions.

  Returning to FIG. In S150, the control unit 15 re-records each of the image candidate positions M specified in S135 and each of the image addition positions specified in S145 in the memory 52 as the image candidate positions M. That is, each of the image candidate positions M specified in S135, each of the current image positions D, and each of the image preliminary positions E are recorded in the memory 52 as the image candidate positions M.

  In S155, the control unit 15 acquires a reference luminance distribution according to the type of the object to be detected. The type of the object to be detected is specified in S110. The reference luminance distribution is a luminance distribution (hereinafter referred to as luminance distribution) in the captured image, and is a luminance distribution determined in advance according to the type of the object. In the present embodiment, for example, reference luminance distributions for a plurality of types of objects such as white lines, yellow lines, curbs, guardrails, and the like are recorded in the memory 52 in advance.

  In the present embodiment, the reference luminance distribution is a luminance in a direction (hereinafter, image horizontal direction) perpendicular to the image vertical direction in a captured image in which the road extending direction is the vertical direction (hereinafter image vertical direction). This is a distribution and is represented as a luminance distribution of pixels within a predetermined range (hereinafter referred to as a detection range). As described above, since it is specified in S110 that the white line is an object, the control unit 15 acquires a reference luminance distribution of the white line, for example, as illustrated in FIG.

  In S160, the control unit 15 determines whether each of the image candidate positions M recorded in the memory 52 represents the position of the white line in the captured image, based on the reference luminance distribution acquired in S155. To do.

  Specifically, as illustrated in FIG. 7, the control unit 15 acquires a luminance distribution within a predetermined detection range including the candidate position M in the captured image acquired in S <b> 130. Here, the captured image acquired in S <b> 130 is an image in which the traveling direction of the host vehicle 1 is a vertical direction and a direction perpendicular to the vertical direction is a horizontal direction. The position in the horizontal direction in the captured image acquired in S130 is referred to as a horizontal position. In the present embodiment, the luminance distribution within the range equal to the detection range at the horizontal position including the image candidate position M is acquired.

  Based on the degree of matching between the acquired luminance distribution and the reference luminance distribution, the control unit 15 captures the image candidate position M acquired in S130 when the degree of matching is equal to or greater than a predetermined threshold. It is determined that the position of the white line is represented by. The degree of matching is larger when the degree of matching is larger than when the degree of matching is small. Since such a determination method is a well-known method as pattern matching, detailed description is omitted.

  The control unit 15 shifts the process to S165 when the image candidate position M represents the position of the white line in the captured image, and proceeds to S170 when the image candidate position M does not represent the position of the white line in the captured image. Transition.

In S165, the control unit 15 causes the image candidate position M to be recorded in the memory 52 as a detection position for the white line.
In S <b> 170, the control unit 15 determines whether it is determined whether or not all image candidate positions M are white line positions. Here, the control part 15 makes a process transfer to S160, when determination is not performed about all the image candidate positions M, and repeats the process of S160-S170. On the other hand, the control part 15 complete | finishes this detection process, when determination is performed about all the image candidate positions M. FIG.

That is, the control unit 15 configured as described above operates as follows.
(1) As shown in FIG. 8, the control unit 15 converts the target position A based on the map data 141 and a plurality of neighboring positions B set around the target position A into image coordinates in the captured image of the current period T _n. The image candidate position M, which is the converted position, is recorded in the memory 52. In FIG. 8, a circle within a range surrounded by a dotted line 301 indicates the image candidate position M here.

(2) As shown in FIG. 8, the control unit 15 converts the past detection position C into the image coordinates in the captured image of the current period T _n based on the movement amount of the host vehicle 1. D is recorded in the memory 52. In FIG. 8, a circle in the range surrounded by the dotted line 302 indicates the detection position C. Further, a black circle within a range surrounded by a dotted line 303 indicates the current image position D.

(3) As shown in FIG. 8, the control unit 15 records in the memory 52 image preliminary positions E that are a plurality of positions set around the current image position D in the captured image of the current period T _n . In FIG. 8, a white circle in the range surrounded by the dotted line 303 indicates the image preliminary position E. In the present embodiment, a plurality of image preliminary positions E are set around the current image position D.

(4) The control unit 15 records the image candidate position M in (1), the current image position D in (2), and the image preliminary position E in (3) in the memory 52 as the image candidate position M. cure. The control unit 15 in the captured image of the current period T _{n for} the image candidate position M of (1), the current image position D of (2), and the image preliminary position E of (3) as the image candidate position M. It is determined whether or not the position of the object is represented. Control unit 15 records in the memory 52 the image candidate position M, which represents the position of the object in the captured image of the current period T _n of the plurality of image candidate position M as a detection position N.

  In FIG. 8, the black circle in the range surrounded by the dotted line 304 and the dotted line 305 indicates the detection position N. The detection position N within the range surrounded by the dotted line 304 indicates the detection position N detected by making a determination based on the image candidate position M in (1). The detection position N within the range surrounded by the dotted line 305 indicates the detection position N detected by making a determination based on the current image position D in (2) and the image preliminary position E in (3). .

[1-3. effect]
According to the first embodiment described in detail above, the following effects are obtained.
(1a) In S115, the control unit 15 acquires an estimated position P that is the position of the host vehicle 1 estimated by the positioning device 12 from the positioning device 12 that estimates the position of the host vehicle. In S120, the control unit 15 acquires the target position A, which is the position of the target object, from the map data recording unit 14 that stores the map data 141 including the position of the predetermined target object.

  In S130, the control unit 15 acquires a captured image captured by the camera 11 mounted on the vehicle. In S135, the control unit 15 sets the vicinity position B with respect to the target position A, and based on the relative position of the target position A with the estimated position P as a reference, the target position A and the vicinity A position corresponding to each of the positions B is specified as an image candidate position M.

  In S160, the control unit 15 determines whether the image candidate position M represents the position of the object in the captured image. In S165, when the image candidate position M represents the position of the object in the captured image, the control unit 15 records the image candidate position M in the memory 52 as the detection position.

  The target object is an object arranged along the direction in which the road on which the host vehicle 1 travels extends. The proximity position B is a position included within a predetermined range from the target position A. The image candidate position M is a position where an object can exist in the captured image.

  According to this, since it is determined whether or not the position of the object is represented only for the image candidate position M in the captured image, for example, whether or not the position of the object is represented for each of all the positions included in the captured image. The processing load for image processing can be reduced as compared with the case of determining.

  Further, not only the target position A but also a neighboring position B located within a predetermined range from the target position A is set as an image candidate position M, and it is determined whether or not the neighboring position B also represents an object. Therefore, even if there is an error in the estimated position P by the positioning device 12, the detected position can be specified in the captured image.

  In the present embodiment, the designated lateral distance Dx, which is the distance in the lateral direction between the target position A and the neighboring position B, is set to be less than the estimation error by the positioning device 12, so that the detected position based on the neighboring position B is detected. Can be accurately identified.

Further, in the present embodiment, as shown in FIG. 9, the range outside the imaging range Z of the camera 11 in the past processing cycle T _n−1 is also within the imaging range Z of the camera 11 in the current cycle T _n . In such a range, it is possible to newly specify the detection position N. That is, new detection positions N can be identified one after another according to the amount of movement of the host vehicle 1.

  (1b) In S165, the control unit 15 repeatedly records the detected position in the memory 52 at a predetermined processing cycle. In S220, the control unit 15 specifies a current image position D that is a position corresponding to the past detection position C recorded in the memory 52 in the captured image acquired in S130. In S160, the control unit 15 determines each of the image candidate positions M using the image candidate position M specified in S135 and the current image position D as the image candidate positions M.

  Here, since the past detection position C is a position specified as the position of the target object in the past captured image by image processing such as pattern matching, the accuracy is high. That is, the past detection position C accurately represents the position of the object in the past captured image even if there is an error in the measurement position P by the positioning device 12.

According to this, the control unit 15 sets the current image position D corresponding to the past detection position C with high accuracy as a candidate, and determines whether the current image position D is also the position of the object. . Therefore, even if the estimated error by the positioning device 12 has occurred, it is possible to accurately identify the detection position N in the captured image of the current period T _n. Note that the term “candidate” here means a candidate for the position of the object in the captured image.

  For example, FIG. 10 shows a comparative example in which the estimation error by the positioning device 12 is too large and none of the image candidate positions M obtained by converting the target position A into image coordinates represents the position of the white line as the target. ing.

  According to the present embodiment, even in the case where the estimation error by the positioning device 12 is large as in the comparative example, the past detection position as in the detection position N within the range surrounded by the dotted line 305 in FIG. Based on C, the detection position N can be specified with high accuracy.

  (1c) In S145, the control unit 15 sets a plurality of image preliminary positions E with respect to the current image position D in the captured image acquired in S130. In S160, the control unit 15 determines each of the image candidate positions M by setting each of the image candidate position M, the current image position D, and the image preliminary position E specified in S135 as the image candidate positions M. . The image preliminary position E is a position within a predetermined range J including the current image position D.

  According to this, the control unit 15 sets a plurality of image preliminary positions E around the current image position D with respect to the current image position D, and whether or not the image preliminary position E is also the position of the object. Determine whether. For this reason, even if an estimation error caused by the positioning device 12 is larger, a new detection position N can be specified in the captured image.

  (1d) In S155, the control unit 15 acquires a reference luminance distribution representing a predetermined luminance distribution for the object. In S160, the control unit 15 acquires the luminance distribution of the pixels within a predetermined range including the image candidate position M, and based on the degree of coincidence between the acquired luminance distribution and the reference luminance distribution, the image candidate position M is acquired. It is determined whether or not represents the position of the object.

  The degree of coincidence with the reference luminance distribution is performed based on image processing such as pattern matching. According to this, since the control unit 15 determines the degree of coincidence with the reference luminance distribution only for the image candidate position M, it is possible to reduce the load of image processing.

  (1e) The reference luminance distribution is set for each type of object. According to this, since the reference luminance distribution is set for each type of object, the detection position can be specified with high accuracy for each object.

  (1f) The control unit 15 specifies the type of the object in S110. And the control part 15 acquires the reference | standard brightness distribution according to the kind of target object specified by S110 in S155.

  According to this, even if there are a plurality of types of objects to be detected, a luminance reference distribution corresponding to the target is acquired for each target, so that the detection positions of the plurality of types of objects in the captured image are accurately determined. It becomes possible to specify well.

[2. Second Embodiment]
[2-1. Difference from the first embodiment]
Since the basic configuration of the second embodiment is the same as that of the first embodiment, differences will be described below. Note that the same reference numerals as those in the first embodiment indicate the same configuration, and the preceding description is referred to.

  In the first embodiment described above, the current image position D is specified by reflecting the movement amount of the host vehicle with respect to the past detection position C. On the other hand, the second embodiment is different from the first embodiment in that the position on the estimated line that is a line represented by the arrangement of the past detection positions C is specified as the current image position D.

[2-2. processing]
Next, an additional process executed by the control unit 15 of the second embodiment instead of the process shown in FIG. 5 of the first embodiment will be described with reference to the flowchart of FIG. Note that the processes in S210 and S230 to S250 in FIG. 11 are the same as the processes in S210 and S230 to S250 in FIG.

In S210, the control unit 15 acquires the past detection position C for the white line recorded in the memory 52, as in the above embodiment.
In S215, the control unit 15 identifies an estimated line. The estimated line is a line represented by a sequence of past detection positions C. As shown in FIG. 12, the control unit 15 uses the least square method or the like to calculate, as the estimated line L _n−1 , a line passing through the most past detection positions C among the plurality of past detection positions C. . The estimated line L _n-1 can be a straight line or a curve.

In S220, the control unit 15 specifies the current image position D. In the present embodiment, the control unit 15 specifies the current image position D based on the estimated line L _n−1 specified in S215.
Specifically, as shown in FIG. 12, the control unit 15 specifies an estimated line L _n corresponding to the estimated line L _{n −1} in the captured image of the current period T _n , and performs arbitrary processing on the estimated line L _n . The position is specified as the current image position D. Control unit 15 sets a plurality of current image position D on the estimated line L _n. Current image position D is represented by image coordinates in the captured image of the current period T _n.

Control unit 15, in S230, each of the current image position D which is specified based on the estimated line L _n at S220, as the image addition position, recorded in the memory 52.
Furthermore, the control unit 15 sets a position within a predetermined range J including the current image position D as the image preliminary position E in S240.

In S250, the control unit 15 performs the same process as in S240 in FIG.
Thereby, in the present embodiment, each of the current image position D set based on the estimated line L _n and each of the image preliminary position E set for the current image position D are the image addition positions. Is recorded in the memory 52.

[2-3. effect]
According to 2nd Embodiment explained in full detail above, there exist the effect (1a)-(1f) of 1st Embodiment mentioned above, and also there exist the following effects.

(2a) control unit 15, in S215, to identify the estimated line L _n is a line represented by a sequence of past detection positions C. Control unit 15, in S220, to identify the position of the putative line L _n a position in the captured image of the current period T _n as the current image position D.

According to this, the control unit 15 determines whether or not the position on the estimated line L represented by the arrangement of the past detection positions C is the position of the object in the captured image. Therefore, even if a large error in the estimated position P by the positioning device 12 has occurred, suppressing the influence of such errors, identify more accurately the new detection position N in the captured image in the current cycle T _n be able to.

In particular, a greater effect can be obtained when the road on which the object is placed is straight.
[3. Other Embodiments]
As mentioned above, although embodiment of this indication was described, this indication is not limited to the above-mentioned embodiment, and can carry out various modifications.

(3a) In the above embodiment, the position included in the range separated from the white line by the designated lateral distance Dx on the right and left sides in the horizontal direction is set as the vicinity position B. However, the present invention is not limited thereto.
For example, the proximity position B can be set to a position that is separated from the target position A by the specified lateral distance Dx for at least some of the plurality of target positions A acquired in S120. Further, at least one neighboring position B can be set with respect to the target position A for which the neighboring position B is set.

  Further, for example, the vicinity position B can be set to an arbitrary position within a range that is a specified lateral distance Dx away from the target position A for at least some of the plurality of target positions A acquired in S120.

  (3b) In the above embodiment, an example has been described in which each of the image candidate positions M specified in S135, each of the current image positions D, and each of the image preliminary positions E are used as the image candidate positions M. However, the present invention is not limited to this.

  For example, the control unit 15 may determine whether or not the image candidate position M represents the position of the target object using only the image candidate position M specified in S135. In this case, the control unit 15 may be configured to execute the process in which S140 to S150 are deleted as the detection process.

  Further, for example, the control unit 15 uses each of the image candidate positions M specified in S135 and each of the current image positions D as the image candidate positions M, and the image candidate positions M represent the positions of the objects. It may be determined whether or not. In this case, the control unit 15 may be configured to execute the process in which S240 to S250 are deleted as an additional process.

(3c) In the above embodiment, an example in which only a white line is detected as an object has been described. However, it is not limited to this.
For example, the control unit 15 may be configured to detect a plurality of types of objects such as a white line and a yellow line. In this case, it is recorded in advance in the memory 52 that the white line and the yellow line should be detected, and the control unit 15 may specify that the white line and the yellow line are detected as objects in S110. .

  Then, the control unit 15 executes S115 to S170 with the white line as the object and records the white line detection position in the memory 52, and then executes S115 to S170 with the yellow line as the object and sets the detection position of the yellow line. It may be configured to record in the memory 52.

Further, for example, the control unit 15 may be configured to detect a curbstone, a guardrail, or the like as an object in addition to the white line and the yellow line. In this case, in addition to the luminance distribution in the detection range in the horizontal direction of the image, the luminance distribution in the vertical direction of the image and the luminance distribution of the pixels in the predetermined range may be preset as the reference luminance distribution. The control unit 15 may be configured to acquire the luminance distribution in the longitudinal direction in the captured image of the current period T _n.

Then, the control unit 15 determines the position of the object in the captured image with the image candidate position M being the current period T _n based on the degree of coincidence between the acquired luminance distribution in the horizontal direction and the vertical direction and the reference luminance distribution. It may be determined whether or not it is represented.

(3d) In the above embodiment, the control unit 15 specifies the estimated line L _{n in the} captured image of the current period T _n based on the sequence of the past detection positions C, but is not limited to this. For example, the control unit 15 may specify the estimated line L _n using the position of the white line based on the map data 141.

  (3e) A plurality of functions of one constituent element in the above embodiment may be realized by a plurality of constituent elements, or a single function of one constituent element may be realized by a plurality of constituent elements. . Further, a plurality of functions possessed by a plurality of constituent elements may be realized by one constituent element, or one function realized by a plurality of constituent elements may be realized by one constituent element. Moreover, you may abbreviate | omit a part of structure of the said embodiment. In addition, at least a part of the configuration of the above embodiment may be added to or replaced with the configuration of the other embodiment. In addition, all the aspects included in the technical idea specified from the wording described in the claims are embodiments of the present disclosure.

  (3f) In addition to the above-described in-vehicle system 10, the control unit 15, and the CPU 51, various programs such as a program for causing the control unit 15 to function, a non-transition actual recording medium such as a semiconductor memory in which the program is recorded, an object detection method, etc. The present disclosure can also be realized in the form of.

[4. Correspondence between Embodiment and Claims]
The control unit 15 corresponds to an object detection device, the positioning device 12 corresponds to a position estimation device, the map data recording unit 14 corresponds to a map storage unit, and the memory 52 corresponds to a recording unit. The control unit 15 corresponds to a vehicle position unit, a target position unit, an image acquisition unit, a basic unit, a reserve unit, a reference setting unit, a determination unit, a recording instruction unit, an estimation line unit, and a correspondence specifying unit.

  Further, S115 corresponds to the processing as the vehicle position portion, S120 corresponds to the processing as the target position portion, S130 corresponds to the processing as the image acquisition portion, and S135 corresponds to the processing as the basic portion. S145 corresponds to processing as a spare unit, S155 corresponds to processing as a reference acquisition unit, S160 corresponds to processing as a determination unit, and S165 corresponds to processing as a recording instruction unit. Further, S215 corresponds to the process as the estimated line part, and S220 corresponds to the process as the correspondence specifying part.

  DESCRIPTION OF SYMBOLS 1 Vehicle, 11 Camera, 12 Positioning device, 14 Map data recording part, 15 Control part, 52 Memory.

Claims (6)

An object detection device (15) mounted on a vehicle (1),
A vehicle position unit (S115) for acquiring an estimated position, which is the position of the vehicle estimated by the position estimating device, from a position estimating device (12) that estimates the position of the vehicle;
A target position part (S120) for acquiring a target position which is the position of the target object from a map storage part (13) for storing map data including the position of the predetermined target object;
An image acquisition unit (S130) for acquiring a captured image captured by a camera (11) mounted on the vehicle;
At least one neighboring position is set with respect to the target position, and each of the target position and the neighboring position in the captured image is based on a relative position of the target position with reference to the estimated position. A basic part (S135) for identifying a corresponding position as a candidate position;
A determination unit (S160) for determining whether or not the candidate position represents the position of the object in the captured image;
A recording instruction unit (S165) that causes the recording unit (52) to record the candidate position as a detection position when the candidate position represents the position of the object in the captured image;
With
The target object is an object arranged along a direction in which a road on which the vehicle travels extends,
The vicinity position is a position within a predetermined range including the target position,
The candidate position is a position where the object can exist in the captured image. The object detection device according to claim 1,
The recording instruction unit repeatedly records the detection position in the recording unit at a predetermined processing cycle,
Correspondence specifying unit (S220) for specifying a current image position that is a position corresponding to the past detection position recorded in the recording unit in the captured image acquired by the image acquisition unit.
Further comprising
The determination unit determines the candidate position using each of the candidate position specified by the basic unit and the current image position as the candidate position. The object detection device according to claim 2,
In the captured image acquired by the image acquisition unit, a preliminary unit that sets at least one preliminary position for the current image position (S145)
Further comprising
The determination unit determines the candidate position with the candidate position specified by the basic unit, the current image position, and the spare position as the candidate position,
The preliminary position is a position within a predetermined range including the current image position.
Object detection device. It is an object detection apparatus of Claim 2 or Claim 3, Comprising: The estimation line part which specifies the estimation line which is a line represented by the arrangement | sequence of the said past detection position (S215)
Further comprising
The correspondence specifying unit specifies a position on the estimated line as a position in the captured image as the current image position. The object detection device according to any one of claims 1 to 4,
A reference acquisition unit (S155) for acquiring a reference luminance distribution representing a predetermined luminance distribution for the object;
The determination unit acquires a luminance distribution of pixels within a predetermined range including the candidate position, and the candidate position is determined based on a degree of coincidence between the acquired luminance distribution and the reference luminance distribution. An object detection device that determines whether or not it represents the position of an object. The object detection device according to claim 5,
The reference luminance distribution is set for each type of the object.