JP2018045385A - Vehicle control apparatus, vehicle control method, and vehicle control program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a vehicle control apparatus configured to prevent a warning from being issued at a time not intended by a driver and accurately recognize an object, a vehicle control method, and a vehicle control program.SOLUTION: A vehicle control system includes: an object detection unit which detects an object near a vehicle; a recognition unit which determines a positional relationship between the vehicle and the object detected by the object detection unit, and identifies a type of the object detected by the object detection unit, to modify the content of processing for identifying the type of the object, on the basis of the determined positional relationship; and a travel support unit which supports travel of the vehicle, on the basis of an identification result of the recognition unit.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a vehicle control device, a vehicle control method, and a vehicle control program.

  Conventionally, there is an apparatus that detects an object around a vehicle and recognizes whether or not the detected object is a pedestrian. For example, it is a device that calculates the degree of risk of contact between a vehicle and a pedestrian based on the position of the vehicle, the vehicle speed, etc., and gives an alarm, and recognizes an object detected according to the degree of risk as a pedestrian An apparatus for switching and recognizing pedestrians around a vehicle and alarming is disclosed (for example, see Patent Document 1).

JP 2009-64274 A

  However, depending on the positional relationship between the vehicle and the detected object, the device of Patent Document 1 recognizes noise or an object that is not a pedestrian as a pedestrian, or recognizes a pedestrian as an object that is not noise or a pedestrian. Thus, an alarm may be issued at a timing unintended by the driver, which may impair the merchantability.

  The present invention has been made in view of such circumstances, and suppresses an alarm or the like at a timing not intended by the driver, and can accurately recognize an object, a vehicle control device, a vehicle control method, Another object is to provide a vehicle control program.

  According to the first aspect of the present invention, an object detection unit (12) that detects an object around the vehicle and a positional relationship between the vehicle and the object detected by the object detection unit are obtained and detected by the object detection unit. A recognition unit for identifying the type of the detected object, the recognition unit (32) for changing the content of the process for identifying the type of the object based on the obtained positional relationship, and the identification result by the recognition unit Based on this, it is a vehicle control system (1) provided with the driving assistance part (70) which assists driving | running | working of the said vehicle.

  Invention of Claim 2 is the vehicle control system of Claim 1, Comprising: The said recognition part exists on the advancing direction locus | trajectory which is the locus | trajectory which the said vehicle is going to advance in the area | region of the object detected object. The threshold used for identifying the type of the object is set higher than the threshold used for identifying the type of the object that is not on the trajectory in the traveling direction.

  Invention of Claim 3 is the vehicle control system of Claim 2, Comprising: The said recognition part determines that the said object is going to cross the advancing direction locus | trajectory which is a locus | trajectory which the vehicle is going to advance, The threshold value is set low with respect to the object to be crossed.

  Invention of Claim 4 is a vehicle control system of any one of Claim 1 to 3, Comprising: The said recognition part is the said threshold value, so that the area | region far from the said vehicle regarding the position of the horizontal direction of the said vehicle. Is set low.

  A fifth aspect of the present invention is the vehicle control system according to any one of the first to fourth aspects, wherein the recognition unit has a first distance greater than or equal to a first distance from the vehicle with respect to a traveling direction of the vehicle. For the region, the threshold value is set higher than that of the second region before the vehicle reaches the first distance.

  A sixth aspect of the present invention is the vehicle control system according to any one of the first to fifth aspects, wherein the recognition unit has a third distance less than a second distance from the vehicle with respect to a traveling direction of the vehicle. For the area, a threshold value used for identifying the type of the object is set higher than that of the fourth area that is equal to or greater than the second distance from the vehicle.

  A seventh aspect of the present invention is the vehicle control system according to any one of the second to sixth aspects, wherein the object detection unit is an imaging unit that captures an image, and an object edge distribution is A storage unit configured to store the stored object model information, wherein the recognition unit extracts edges from a time-series image captured by the imaging unit, and the distribution of the extracted edges and the storage unit store An index of similarity with the edge distribution stored in the object model information is derived for each of the time-series images, and based on the variation of the similarity index derived for each of the time-series images, the object Is to recognize.

  According to an eighth aspect of the present invention, the computer detects an object around the vehicle, obtains a positional relationship between the vehicle and the detected object, and executes a process of identifying the type of the detected object. The vehicle control method is configured to change the content of the process for identifying the type of the object based on the obtained positional relationship, and to support the traveling of the vehicle based on the identification result.

  The invention according to claim 9 causes a computer to detect an object around the vehicle, obtain a positional relationship between the vehicle and the detected object, and execute a process of identifying the type of the detected object. The vehicle control program changes the content of the process for identifying the type of the object based on the obtained positional relationship, and assists the vehicle on the basis of the identification result.

  According to the first to ninth aspects of the present invention, the driver intends to change the content of the process for identifying the type of the object based on the positional relationship between the vehicle and the object detected by the object detection unit. It is possible to suppress an alarm or the like at a timing when it is not performed and accurately recognize an object. In addition, the travel control unit can realize vehicle control according to the surrounding situation by assisting the vehicle travel based on the identification result of the recognition unit.

2 is a diagram illustrating an example of a functional configuration of a vehicle control system 1 including an object recognition device 20. FIG. It is the figure which showed the content of the pedestrian model information 64 typically. 4 is a flowchart showing a flow of processing executed by the object recognition device 20. 4 is a flowchart illustrating a flow of threshold setting processing executed by a processing unit. It is a figure for demonstrating the threshold value set when a candidate area | region exists in the advancing direction locus | trajectory. It is a figure for demonstrating the threshold set based on the position of the horizontal direction of the own vehicle. It is a figure for demonstrating the threshold value set according to the distance regarding the advancing direction. It is a figure which shows an example in case the process which sets a threshold value based on 2nd distance is abbreviate | omitted. It is a figure which shows an example when the process which sets a threshold value based on 1st distance is abbreviate | omitted. It is a figure for demonstrating the amount of laps. It is a figure for demonstrating the relationship between the attitude | position of a crossing pedestrian and the fluctuation | variation of the parameter | index of a similarity degree. It is a figure for demonstrating the fluctuation | variation of the parameter | index of the similarity identified as not being a crossing pedestrian. It is a flowchart which shows the flow of the process performed by the object recognition apparatus 20 of 2nd Embodiment. It is a figure which shows an example of a function structure of 1 A of vehicle control systems of 2nd Embodiment.

  Hereinafter, embodiments of a vehicle control system, a vehicle control method, and a vehicle control program of the present invention will be described with reference to the drawings.

<First Embodiment>
FIG. 1 is a diagram illustrating an example of a functional configuration of a vehicle control system 1 including an object recognition device 20. The vehicle control system 1 includes, for example, a radar device 10, a camera 12, a vehicle sensor 14, an object recognition device 20, a vehicle control device 70, a travel drive device 72, and a brake device 74.

  The radar apparatus 10 is provided, for example, in the vicinity of a bumper of a vehicle (hereinafter referred to as the host vehicle) on which the vehicle control system 1 is mounted, a front grille, and the like. For example, the radar apparatus 10 radiates a millimeter wave in front of the host vehicle, receives a reflected wave reflected by the radiated millimeter wave hitting the object, and analyzes the received reflected wave, thereby specifying the position of the object. The position of the object includes, for example, at least a distance from the own vehicle to the object, and may include an azimuth or a lateral position of the object with respect to the own vehicle. The radar apparatus 10 detects the position of an object by, for example, FM-CW (Frequency-Modulated Continuous Wave) method, and outputs the detection result to the object recognition apparatus 20. The radar apparatus 10 is an example of an “object detection unit”.

  The camera 12 is a digital camera using a solid-state imaging device such as a charge coupled device (CCD) or a complementary metal oxide semiconductor (CMOS). The camera 12 is attached to the upper part of the front windshield, the rear surface of the rearview mirror, etc. For example, the camera 12 periodically images the front of the host vehicle. For example, the camera 12 captures an image and outputs the captured image to the object recognition device 20. The camera 12 is not limited to one, and a plurality of cameras 12 may be provided in the host vehicle, or a stereo camera including a plurality of cameras may be used. The camera 12 is an example of an “object detection unit”.

  The vehicle sensor 14 includes a vehicle speed sensor that detects a vehicle speed, an acceleration sensor that detects acceleration, a yaw rate sensor that detects an angular velocity around a vertical axis, a direction sensor that detects the direction of the host vehicle, and the like. The vehicle sensor 14 outputs the detection result of each sensor to the vehicle control device 70.

  The object recognition device 20 includes, for example, an image acquisition unit 30, an image recognition unit 32, a surrounding situation recognition unit 50, and a storage unit 60. Some or all of the image acquisition unit 30, the image recognition unit 32, and the surrounding situation recognition unit 50 are realized by a processor executing a program (software). Some or all of these may be realized by hardware such as LSI (Large Scale Integration) and ASIC (Application Specific Integrated Circuit), or may be realized by a combination of software and hardware. Further, each functional unit included in the object recognition device 20 may be distributed by a plurality of computer devices. The storage unit 60 is realized by a ROM (Read Only Memory), a RAM (Random Access Memory), an HDD (Hard Disk Drive), a flash memory, or the like.

  The image acquisition unit 30 acquires an image captured by the camera 12.

  The image recognition unit 32 includes, for example, a feature amount extraction unit 34, a processing unit 36, and an identification unit 38. The processing unit 36 obtains the positional relationship between the vehicle and the region to be processed (region including the object) in the image acquired by the image acquisition unit 30. For example, the processing unit 36 derives a position corresponding to the region to be processed based on parameters such as the position of the camera 12, the focal length of the camera 12, and an infinite point (for example, a road vanishing point) in the image.

  The feature amount extraction unit 34 acquires a feature amount from the image acquired by the image acquisition unit 30, and derives an edge from the acquired feature amount. An edge is a pixel or a pixel group in which a luminance difference or a color parameter difference from a peripheral pixel or a pixel group obtained by grouping pixels changes more than a reference. For example, the edge is derived by obtaining a change in the feature amount between neighboring pixels or pixel groups using a SOBEL filter or the like.

The processing unit 36 changes the content of the process for identifying the type of the object based on the obtained positional relationship. In the following description, it is assumed that the type of the object to be identified is a pedestrian.
The change in the content of the process for identifying the type of the object is to change the threshold value used for the process in which the identification unit 38 identifies the object as a pedestrian. The processing unit 36 derives the position of the pedestrian in the real space identified by the identifying unit 38. The processing unit 36 derives the moving direction, moving speed, and the like of the object based on the change in the position of the object (for example, a pedestrian) in the time-series image.

  The identification unit 38 identifies whether the object is a pedestrian based on the feature amount extracted by the feature amount extraction unit 34 and the pedestrian model information 64 stored in the storage unit 60.

  The surrounding situation recognition unit 50 acquires the detection result of the radar device 10, and recognizes the position of the object based on the acquired result. Moreover, the surrounding situation recognition part 50 acquires the position of the pedestrian in real space as a process result of the image recognition part 32. FIG. The surrounding situation recognition unit 50 places importance on the distance from the host vehicle among the positions of the objects specified by the radar apparatus 10 and also places importance on the azimuth or lateral position among the positions of the pedestrians acquired from the image recognition unit 32. Therefore, these positions may be integrated to recognize the position of the pedestrian. Further, instead of (or in addition to) the radar apparatus 10, a sensor such as a laser radar or an ultrasonic sensor may be provided.

  Moreover, the surrounding state recognition part 50 may detect the position, speed, etc. of a pedestrian based on the information acquired from the vehicle-to-vehicle communication or the sensor which detects the pedestrian who walks or crosses a road. In this case, the host vehicle includes a communication unit that communicates with another vehicle, a sensor that detects a vehicle traveling on the road, and the like.

  The vehicle control device 70 controls the host vehicle based on the detection result of the vehicle sensor 14 and the position of the pedestrian recognized by the surrounding situation recognition unit 50. For example, the vehicle control device 70 performs an automatic brake control that causes the brake device 74 to output a braking force when a TTC (Time To Collision) with the recognized pedestrian falls below a certain level. Further, the vehicle control device 70 may cause a speaker (not shown) to output an alarm. Further, the vehicle control device 70 may control the travel drive device 72 or the brake device 74 so that the distance from the pedestrian does not approach a certain distance or less. The vehicle control device 70 is an example of a “running support unit”.

  The storage unit 60 stores, for example, simple model information 62, pedestrian model information 64, and other control programs for controlling the vehicle. The simple model information 62 is pedestrian model information that represents a pedestrian edge distribution generated based on a gentler standard than the pedestrian model information 64. The gradual reference means that the edge distribution is smaller than the pedestrian edge distribution included in the pedestrian model information 64. In the pedestrian model information 64, at least pedestrian model information representing the distribution of pedestrian edges is stored together with the identification information of the object.

  FIG. 2 is a diagram schematically showing the contents of the pedestrian model information 64. The pedestrian model information 64 indicates, for example, an edge distribution assumed to be extracted from an image in which a pedestrian is captured. Simple model information 62 is obtained by thinning the edges from the pedestrian model information 64 at a predetermined interval. Moreover, in the pedestrian model information 64, the pedestrian model information may be subdivided and stored for each part of the pedestrian.

  The travel drive device 72 outputs a travel drive force (torque) for traveling of the vehicle to the drive wheels. For example, when the host vehicle is an automobile using an internal combustion engine as a power source, the travel drive device 72 includes an engine, a transmission, and an engine ECU (Electronic Control Unit) that controls the engine, and the host vehicle uses the motor as a power source. When the vehicle is a hybrid vehicle, the vehicle includes an engine, a transmission, and an engine ECU, and a travel motor and a motor ECU. . When the travel drive device 72 includes only an engine, the engine ECU adjusts the throttle opening, shift stage, and the like of the engine according to information input from the vehicle control device 70 described later. When travel drive device 72 includes only the travel motor, motor ECU adjusts the duty ratio of the PWM signal applied to the travel motor in accordance with information input from vehicle control device 70. When travel drive device 72 includes an engine and a travel motor, engine ECU and motor ECU control the travel drive force in cooperation with each other in accordance with information input from vehicle control device 70.

  The brake device 74 is, for example, an electric servo brake device that includes a brake caliper, a cylinder that transmits hydraulic pressure to the brake caliper, an electric motor that generates hydraulic pressure in the cylinder, and a braking control unit. The braking control unit of the electric servo brake device controls the electric motor according to the information output by an automatic brake control unit (not shown) so that the brake torque corresponding to the braking operation is output to each wheel. The electric servo brake device may include, as a backup, a mechanism that transmits the hydraulic pressure generated by operating the brake pedal to the cylinder via the master cylinder. The brake device 74 is not limited to the electric servo brake device described above, but may be an electronically controlled hydraulic brake device. The electronically controlled hydraulic brake device controls the actuator according to the information output by the vehicle control device 70, and transmits the hydraulic pressure of the master cylinder to the cylinder. Further, when the host vehicle includes a travel motor, the brake device 74 may include a regenerative brake by the travel motor described in the travel drive device 72.

FIG. 3 is a flowchart showing a flow of processing executed by the object recognition device 20.
This process is executed each time the image acquisition unit 30 acquires an image captured by the camera 12, for example. Note that the process of this flowchart is expressed as t in the current routine on the premise that the process is repeatedly executed.

  First, the image acquisition unit 30 acquires an image captured by the camera 12 (step S100). Next, the feature amount extraction unit 34 extracts an edge from the image acquired in step S100 (step S102). Next, the processing unit 36 determines whether or not the pedestrian has recognized the crossing pedestrian at time t-1 (that is, in the previous routine) (step S104). A crossing pedestrian is a pedestrian who is moving laterally with respect to the traveling direction of the host vehicle. Moving in the horizontal direction means that the pedestrian's movement locus is virtually extended with respect to the movement direction, and the own vehicle's movement locus is virtually extended with respect to the movement direction. In addition to the case where the virtual lines are orthogonal, the case where these virtual lines intersect is also included.

  When the pedestrian does not recognize the crossing pedestrian at time t-1, the processing unit 36 extracts a candidate area (step S106). The candidate area in step S106 is an area where a pedestrian is likely to exist among areas obtained by subdividing the area included in the image into a plurality of areas. The processing unit 36 extracts, as candidate areas, areas including edge distributions whose similarity between the extracted edge distribution and the edge distribution of the simplified model information 62 is greater than or equal to a predetermined value. For example, the processing unit 36 may extract a region where an object is detected by the radar device 10 or a region including a predetermined number of edges or more as candidate regions.

  Next, the processing unit 36 is used for the process of identifying a pedestrian based on the positional relationship between the vehicle and the object in order to identify whether or not a pedestrian exists in the candidate area extracted in step S106. A threshold value is set (step S108). The positional relationship is, for example, the relationship between the position of the host vehicle and the position of the object on the real plane. Details of the threshold setting will be described later.

  Next, based on the edge distribution extracted in step S102, the edge distribution of the pedestrian model information 64, and the threshold set by the processing unit 36, the identification unit 38 in the candidate region extracted in step S106. Whether or not the object is a pedestrian is identified (step S110). For example, the identification unit 38 corresponds to the extracted edge distribution when the similarity between the extracted edge distribution and the edge distribution of the pedestrian model information 64 is equal to or larger than the threshold set in step S108. The object is identified as a pedestrian.

  On the other hand, when the similarity between the extracted edge distribution and the edge distribution of the pedestrian model information 64 is less than the threshold set by the processing unit 36, the identification unit 38 corresponds to the extracted edge distribution. The object to be identified is not a pedestrian but noise, a depression on the road, or an object other than a pedestrian. In this case, the identification unit 38 may identify the object based on the similarity between the extracted edge distribution and the distribution of other model information. The other model information is information stored by associating identification information of an object different from the pedestrian and information indicating the distribution of the edge of the object. An object different from a pedestrian is a type of object to be watched when the host vehicle travels, such as a vehicle or an animal. Other model information is stored in the storage unit 60.

  When the crossing pedestrian is recognized at time t−1, the processing unit 36 extracts a candidate area based on the previous image (step S112). The candidate area in step S112 is an area in which there is a high possibility that a pedestrian being tracked exists among areas obtained by subdividing the image into a plurality of areas. For example, the processing unit 36 may have a pedestrian in the image acquired in the current routine based on the moving direction and moving speed of the pedestrian tracked derived based on the previous image. A candidate region having a high probability is extracted.

  Next, the process part 36 sets the threshold value used for the process which identifies a pedestrian (step S114). For example, the processing unit 36 sets a threshold based on at least one element among the relative position of the candidate area with respect to the host vehicle, the moving direction of the pedestrian being tracked, and the moving speed of the pedestrian being tracked. To do. The relative position is a relative position on the real plane and is a position mapped from a position on the image plane. Details of the processing in step S114 will be described later.

  Next, based on the edge distribution extracted in step S102, the distribution of the pedestrian model information 64, and the threshold set by the processing unit 36, the identification unit 38 selects objects in the candidate area extracted in step S112. Identify (step S116). For example, if the similarity between the extracted edge distribution and the distribution of the pedestrian model information 64 is equal to or greater than the threshold set in step S114, the identification unit 38 determines that the object corresponding to the extracted edge distribution is Identify as a pedestrian. Thereby, the crossing pedestrian is detected by the image recognition unit 32.

  Next, the image recognition unit 32 performs a tracking process for the detected pedestrian (step S118). For example, the image recognition unit 32 performs the following processing as tracking. For example, the image recognition unit 32 derives a relative position of the identified pedestrian with respect to the host vehicle. In addition, for example, when the pedestrian has been detected in the routine before the previous time, the image recognition unit 32, based on the relative position of the pedestrian derived in the current and previous routines, Deriving movement speed and the like. The moving direction and moving speed derived here are used for the process of estimating the position of the pedestrian in the image acquired by the process of the next routine. It is stored in the storage unit 60.

  Further, the surrounding situation recognition unit 50 acquires the processing result of the image recognition unit 32 and the information of the object specified by the radar apparatus 10, and grasps the pedestrians existing around the own vehicle based on the acquired information. To do. Then, the vehicle control device 70 controls the travel drive device 72 or the brake device 74 based on the position of the pedestrian, the moving direction, and the like ascertained by the surrounding situation recognition unit 50. Thereby, the process of this flowchart is complete | finished.

[Details of threshold setting (part 1)]
Here, the threshold value setting process in step S108 will be described. FIG. 4 is a flowchart showing a flow of threshold setting processing executed by the processing unit 36. First, the processing unit 36 determines whether or not the candidate area extracted in step S106 of FIG. 3 exists in the traveling direction locus (step S200). When the candidate area exists in the traveling direction trajectory, the threshold used for the process of identifying the pedestrian in the candidate area is set high (step S202). When the candidate area does not exist in the traveling direction locus, the process of step S204 is performed. move on. The traveling direction trajectory may be a region surrounded by two lines extending from the left and right ends of the host vehicle in the direction of the central axis of the host vehicle M, and has a bend in consideration of the steering angle and the like. It may be set. Setting the threshold used for identifying a pedestrian in a candidate area to be high means that the threshold is set to be relatively high compared to other areas (the same applies hereinafter).

  The processing unit 36 calculates the distance from the own vehicle to the candidate area. The position of the candidate area for which the distance has been calculated is represented in a two-dimensional space when the host vehicle is viewed from above, for example, as shown in FIG.

  FIG. 5 is a diagram for explaining a threshold value that is set when a candidate region is present in the traveling direction locus. In the figure, the X direction represents the lateral direction of the host vehicle M, and the Y direction represents the traveling direction of the host vehicle M. In the figure, the area A1 represents an area captured by the camera 12, and the area A2 represents an area corresponding to the traveling direction locus. In the drawing, a region CA1 and a region CA2 represent candidate regions to be processed.

  The processing unit 36 increases the threshold used for the process of identifying a pedestrian as compared with the area CA2 not included in the area A2 with respect to the area CA1 included in the area A2 corresponding to the traveling direction locus of the host vehicle M. Set. Here, if the threshold value is not set high in the area A2, there may occur a misrecognition that recognizes noise or the like as a pedestrian. In this case, the vehicle control device 70 is troublesome because it controls the host vehicle M based on the erroneously recognized result. On the other hand, if the threshold value is set high in the area A2 as described above, the occurrence of erroneous recognition is suppressed, and unnecessary control is not performed, so that the host vehicle M can be used according to the surrounding situation of the host vehicle M. Can be realized.

  Next, the process part 36 determines the threshold value used for the process which identifies a pedestrian based on the position of the horizontal direction of the own vehicle M (step S204). FIG. 6 is a diagram for explaining a threshold set based on the position of the host vehicle M in the lateral direction. In the figure, an area CA11 and an area CA12 represent candidate areas to be processed. The virtual line LA is a virtual line that is virtually extended so as to be orthogonal to the virtual line LB. The virtual line LB is a virtual line that virtually extends from the host vehicle M in the direction of the central axis of the host vehicle M. In addition, the description which overlaps with FIG.

  The processing unit 36 sets a lower threshold for the lateral position of the host vehicle M in a region farther from the host vehicle M. In the example shown in the figure, the processing unit 36 compares the length of the virtual line LA2 indicating the distance between the area CA12 and the virtual line LB with respect to the length of the virtual line LA1 indicating the distance between the area CA11 and the virtual line LB. Since it is long, the threshold used for the process for identifying the pedestrian in the area CA12 is set lower than the threshold used for the process for identifying the pedestrian in the area CA11. As for the lateral position of the host vehicle M, a pedestrian who is in a region far from the host vehicle M is not a target to be watched at this point, so there is no problem even if the threshold value is set low as described above. Rather, by setting the threshold value low, the object can be easily recognized as a pedestrian, and the recognized pedestrian can be set as a tracking target. Thereby, it is possible to more reliably recognize the recognized object type in the subsequent processing.

  Next, the processing unit 36 determines whether or not the candidate area extracted in step S106 is included in an area equal to or greater than the first distance from the host vehicle M with respect to the traveling direction of the host vehicle M (step S206). When there is a candidate area in the area that is equal to or greater than the first distance, the processing unit 36 sets a high threshold to be used for the process of identifying the pedestrian in the candidate area (step S208), and the candidate is set in the area that is equal to or greater than the first distance. If the area does not exist, the process proceeds to step S210.

  Next, the processing unit 36 determines whether or not the candidate area extracted in Step S106 exists in an area less than the second distance from the own vehicle M with respect to the traveling direction of the own vehicle M (Step S210). When there is a candidate area in the area less than the second distance, the processing unit 36 sets a high threshold to be used for the process of identifying the pedestrian in the candidate area (step S212), and the candidate is set in the area below the second distance. If the area does not exist, the process of this flowchart ends.

  FIG. 7 is a diagram for explaining a threshold value set according to the distance in the traveling direction. In the figure, a region A11 is a region equal to or greater than a first distance from the reference point (for example, the center of gravity) of the host vehicle M with respect to the traveling direction of the host vehicle M. The area A11 is an example of a “first area”. In an area greater than or equal to the first distance, since an object is recognized as being small, it may be difficult to identify whether the target included in the candidate area is a pedestrian or noise. For this reason, the processing unit 36 suppresses misrecognition by setting a high threshold to be used for the process of identifying a pedestrian (OB1 in the figure) in the candidate region of the first distance or more.

  In the figure, a region A12 is a candidate region that is less than the second distance from the reference point of the host vehicle M with respect to the traveling direction of the host vehicle M. The area A12 is an example of a “third area”. In an area less than the second distance, if noise or the like is identified as a pedestrian due to misrecognition, the host vehicle M may perform an operation to avoid the pedestrian, sudden braking, or the like. Becomes unstable. For this reason, the processing unit 36 suppresses misrecognition by setting a high threshold value for use in the process of identifying a pedestrian (OB2 in the figure) in an area less than the second distance.

  Note that the processing unit 36 does not change the threshold value used for the process of identifying the object OB3 existing in the region A13 that is in the area closer to the first distance than the second distance in the traveling direction of the host vehicle M. The region A13 is an example of “a second region before the first distance from the vehicle” and “a fourth region that is equal to or more than the second distance from the vehicle”.

  In the illustrated example, the first distance and the second distance are described as different distances, but the same distance may be used.

  In addition, one of a process for setting a threshold based on the first distance (steps S206 and S208) and a process for setting a threshold based on the second distance (steps S210 and S212). Processing may be omitted.

  FIG. 8 is a diagram illustrating an example of a case where the process of setting a threshold based on the second distance is omitted. The description overlapping with FIG. 7 is omitted. In the drawing, a region A13 # is a region less than the first distance from the reference point of the host vehicle M with respect to the traveling direction of the host vehicle M. The area A13 # is an example of “a second area in front of reaching the first distance from the vehicle”. In the area equal to or greater than the first distance, the processing unit 36 sets the threshold used for the process for identifying the pedestrian (OB1 in the figure) higher than the area A13 # in front of the vehicle from the vehicle to the first distance. By setting, erroneous recognition is suppressed.

  FIG. 9 is a diagram illustrating an example of a case where the process of setting the threshold based on the first distance is omitted. The description overlapping with FIG. 7 is omitted. In the figure, a region A13 ## is a candidate region that is at least a second distance from the reference point of the host vehicle M with respect to the traveling direction of the host vehicle M. Area A13 ## is an example of “fourth area that is equal to or greater than the second distance from the vehicle”. In the area less than the second distance, the processing unit 36 sets the threshold value used for the process of identifying the pedestrian (OB2 in the figure) as compared to the area A13 ## that is equal to or greater than the second distance from the vehicle. Setting it higher suppresses misrecognition.

  Further, the processing unit 36 may change the threshold value based on the size of the object instead of the processing based on the distance between the host vehicle M and the candidate area. For example, the processing unit 36 may set the threshold value high when the size of the object is equal to or smaller than a predetermined size.

[Details of threshold setting (part 2)]
Here, the process in which the processing unit 36 sets a threshold value used for the object identifying process in step S114 in FIG. 3 will be described. The processing unit 36 identifies a pedestrian in the candidate area when the candidate area is in the traveling direction locus with respect to the candidate area that is highly likely to be the pedestrian extracted in step S106. The threshold value used for the process is set high, but this is not the case when it is determined that there is a crossing pedestrian in the traveling direction of the host vehicle M, and the threshold value of the candidate region in the traveling direction locus is set low. Thereby, the detection stability of a crossing pedestrian can be improved. Further, a condition is further added to set the threshold value low, and if the candidate area exists in the traveling direction trajectory and the possibility of collision between the own vehicle M and the crossing pedestrian is high, walking is performed in the candidate area. The threshold used for the process of identifying a person may be set low. The possibility of a collision is high when the own vehicle M passes through the area where the crossing pedestrian is present or in the vicinity of the area where the crossing pedestrian is predicted to exist in the traveling direction trajectory. It is possible to predict approaching.

  For example, the processing unit 36 determines a time zone in which the crossing pedestrian is predicted to exist in the traveling direction trajectory based on the position and moving speed of the crossing pedestrian in time series, or the crossing pedestrian travels in the traveling direction trajectory. The time predicted to move out of the trajectory in the traveling direction is derived. The processing unit 36 multiplies the speed of the host vehicle M by the time between the current time and the time when the crossing pedestrian is predicted to move out of the traveling direction trajectory from the traveling direction trajectory. The future position of the vehicle M is predicted. In this case, the processing unit 36 determines that the possibility of a collision is low if the future position of the host vehicle M is a predetermined distance or more before the position of the crossing pedestrian in the traveling direction of the host vehicle M. If the distance is not more than the distance or the predicted position of the crossing pedestrian is passing, it is determined that the possibility of the collision is high.

  Further, the processing unit 36 evaluates the amount of overlap between the traveling direction trajectory of the host vehicle M and the moving trajectory that is the trajectory in the moving direction of the crossing pedestrian when the crossing pedestrian has not substantially passed the collision point. You may determine the possibility of a collision using the lap amount and lap ratio which are index values. The collision point is a portion where the traveling direction trajectory of the host vehicle M and the movement trajectory of the pedestrian overlap.

  FIG. 10 is a diagram for explaining the lap amount. In the figure, the crossing pedestrian OB1 # is the predicted position of the pedestrian at the predicted time when the host vehicle M reaches the collision point. Further, y1 is a virtual line extending from the left end portion of the host vehicle M in the traveling direction of the host vehicle M, and y2 is a virtual line extending from the right end portion of the host vehicle M in the traveling direction of the host vehicle M.

  The processing unit 36 obtains the predicted position of the crossing pedestrian, and the second virtual straight line y2 that has already passed through the front end from the front end (in this case, the left end H1L) considering the moving direction of the crossing pedestrian OB1 #. Is calculated as a lap amount. Here, when the crossing pedestrian OB1 # exists outside the trajectory in the traveling direction that is an area between the first virtual line y1 and the second virtual line y2, the lap amount is calculated as zero. Further, the processing unit 36 divides the wrap amount by the width W between the first virtual line y1 and the second virtual line y2, and calculates a value obtained by multiplying by 100 as a wrap rate [%]. And the process part 36 determines with the possibility of a collision being high, when a lap | wrap rate is 70 [%] or more. In addition, even when the lap rate is less than 70 [%], the processing unit 36 tends to have a higher possibility of collision as the lap rate becomes larger than 0 [%] when the lap rate exceeds 0 [%]. Judge that there is. In addition, as described above, the processing unit 36 determines that the possibility of a collision is low when a crossing pedestrian exists outside the traveling direction trajectory. When determining that the possibility of a collision is high, the reference value of the lap rate is not limited to 70 [%], and it is sufficient that the crossing pedestrian OB1 # wraps to some extent, for example, 50 [%]. And so on.

  Further, the processing unit 36 may set the threshold value used for the process of identifying the pedestrian moving in the lateral direction to be low instead of (in addition to) the above [details of setting the threshold value (part 2)]. The threshold used for the process of identifying an object that is highly likely to be the same as the object identified as a crossing pedestrian in the previous process may be set low. The high possibility of being the same is that the position predicted from the position and moving speed of the object in the previous process and the position of the object detected in the current process match.

  Further, the processing unit 36 replaces (in addition to) the above [details of threshold value setting (part 2)], the vehicle speed of the host vehicle M is equal to or lower than a predetermined speed, and is a predetermined distance from the position of the host vehicle M. When an object exists in the above position, the threshold used for the process for identifying a pedestrian may be increased.

  In the flowchart of FIG. 3 described above, the processing unit 36 sets a threshold value for each image captured at a certain time and identifies whether or not the object is a crossing pedestrian, but instead ( In addition, it may be determined whether or not the object is a crossing pedestrian based on a plurality of images captured within a predetermined time.

  Here, when it is identified whether or not the object included in the image is a pedestrian, using the similarity between the edge distribution extracted from the pedestrian and the edge distribution included in the pedestrian model information, It is identified whether the object is a crossing pedestrian. However, since crossing pedestrians change postures by crossing movements, it is necessary to prepare edge distribution information for each crossing pedestrian posture in order to identify pedestrians with high accuracy. Also, even if you prepare edge distribution information for each pedestrian's posture, it is necessary to compare the edge distribution for each posture with the edge distribution extracted from the image, so the processing becomes complicated There is.

  FIG. 11 is a diagram for explaining the relationship between the posture of the crossing pedestrian and the variation of the similarity index. The posture of the crossing pedestrian changes from (a) to (d) in the figure according to the walking motion. In this case, the index of similarity between the edge distribution extracted from the image in which the crossing pedestrian is imaged and the edge distribution of the pedestrian model information 64 varies for each pedestrian movement (time). The vertical axis of the graph in the figure indicates the similarity index, and the horizontal axis indicates time. In addition, since the crossing pedestrian repeats the walking motion, the variation of the similarity index in different periods matches. As shown in the figure, for example, the variation in the similarity index in the cycle t-x in FIG. 11 matches the variation in the similarity index in the cycle t in FIG. When the variation of the similarity index continuously matches in different periods as described above, the processing unit 36 identifies that the target object in the image is a crossing pedestrian. In addition, when the variation of the similarity index continuously matches in different cycles and the similarity index exceeds the threshold in one cycle, the processing unit 36 determines that the target object in the image is You may identify as a crossing pedestrian. On the other hand, when the change of the similarity index does not match continuously in different cycles, the processing unit 36 identifies that the object in the image in which the crossing pedestrian is captured is not a crossing pedestrian.

  FIG. 12 is a diagram for explaining changes in the similarity index that is identified as not being a crossing pedestrian. The vertical axis of the graph in the figure indicates the similarity index, and the horizontal axis indicates time. As shown in the figure, for example, when the variation of the similarity index in the cycle t-x # in FIG. 12 and the variation in the similarity index in the cycle t # in FIG. The target object in the image is identified as not a crossing pedestrian.

  In this case, the processing unit 36 determines that the target object in the image is a crossing pedestrian when the similarity index exceeds a threshold value within a predetermined ratio among the similarity indices included in one cycle. You may identify. In addition, the process of identifying a pedestrian based on the above-described change in the similarity index may be performed only on the candidate area existing in the traveling direction locus.

  In the above-described processing, it is possible to identify whether an object is a crossing pedestrian based on a plurality of images captured within a predetermined time without preparing edge distribution information for each pedestrian posture. However, no consideration is given to accurately recognizing a pedestrian in an image taken at a certain time. In the present embodiment, the processing unit 36 changes a threshold used to identify a pedestrian based on the positional relationship between the host vehicle M and an object with respect to an image captured at a certain time. Thereby, it is possible to suppress erroneous detection and accurately recognize a pedestrian without preparing edge distribution information for each posture of the pedestrian.

  In addition, since the object recognition device 20 of the present embodiment further identifies the pedestrian in consideration of the detection result of the radar device 10 or the like, the content of the processing for identifying the pedestrian based on the positional relationship is changed. Even if it exists, the identification performance with respect to a pedestrian is ensured.

  In the first embodiment described above, the image recognition unit 32 is an image recognition unit that obtains the positional relationship between the vehicle and the object and identifies the type of the detected object, and based on the obtained positional relationship. By changing the content of the process for identifying the type of object, it is possible to suppress an alarm or the like at a timing not intended by the driver and to recognize the object with high accuracy. Moreover, the vehicle control apparatus 70 can implement | achieve control of the vehicle according to the surrounding condition by assisting driving | running | working of a vehicle based on the identification result of the image recognition part 32. FIG.

<Second Embodiment>
Hereinafter, the second embodiment will be described. In the first embodiment, the object recognizing device 20 selects either a candidate area extracted when a crossing pedestrian is recognized or a candidate area extracted when a crossing pedestrian is not recognized. A process for setting a threshold for the candidate area is executed. In contrast, the object recognition device 20 of the second embodiment includes a candidate area extracted when a crossing pedestrian is recognized, and a candidate area extracted when a crossing pedestrian is not recognized. In parallel, processing for setting a threshold value in parallel is executed. Here, the difference from the first embodiment will be mainly described, and description of functions and the like common to the first embodiment will be omitted.

  FIG. 13 is a flowchart illustrating a flow of processing executed by the object recognition device 20 according to the second embodiment. This process is, for example, a process executed on an image captured by the camera 12 at time t.

  First, the image acquisition unit 30 acquires an image captured by the camera 12 (step S300). Next, the feature quantity extraction unit 34 extracts an edge from the image acquired in step S300 (step S302). Next, the processing unit 36 extracts candidate areas (step S304). The processing unit 36 extracts, as candidate areas, areas including edge distributions whose similarity between the extracted edge distribution and the edge distribution of the simplified model information 62 is greater than or equal to a predetermined value.

  Next, the processing unit 36 is used for the process of identifying a pedestrian based on the positional relationship between the vehicle and the object in order to identify whether or not a pedestrian exists in the candidate area extracted in step S304. A threshold is set (step S306). The threshold used for the process for identifying a pedestrian in step S306 is set according to the process executed in steps S200 to S212 in FIG. 4, for example.

  Next, the processing unit 36 determines whether or not the pedestrian has recognized a crossing pedestrian at time t-1 (that is, in the routine of the previous flowchart) (step S308). When the crossing pedestrian is not recognized at time t−1, the process proceeds to step S314. When the crossing pedestrian is recognized at the time t−1, the processing unit 36 selects a candidate area that is likely to have a pedestrian in the image acquired in the current routine based on the previous image. Extract (step S310). Next, the process part 36 sets the threshold value used for the process which identifies a pedestrian with respect to the candidate area | region extracted by step S310 (step S312). The threshold used for the process for identifying the pedestrian in step S312 is, for example, the process executed in steps S200 to S212 in FIG. 4 and the above-described [Details of threshold setting (part 2)]. Set by. For example, when the candidate area extracted in step S310 exists in the traveling direction trajectory, the threshold value used for the process for identifying a pedestrian in the candidate area is reduced. It is a process to set.

  Next, the identification unit 38 extracts in step S304 and / or step S310 based on the edge distribution extracted in step S302, the edge distribution of the pedestrian model information 64, and the threshold set by the processing unit 36. It is identified whether or not the object in the candidate area is a pedestrian (step S314). Next, the image recognition unit 32 executes tracking processing for the detected pedestrian (step S316). Thereby, the process of this flowchart is complete | finished. And the object recognition apparatus 20 performs the process mentioned above with respect to the image imaged with the camera 12 at the time t + 1.

  In the above-described process, the process of extracting candidate regions that are likely to have a pedestrian in step S310 may be omitted. In this case, in step S312, the processing unit 36 further executes a process of [Details of threshold setting (part 2)] on the candidate area extracted in step S304 and subjected to the process of step S306. Set the threshold.

  According to the second embodiment described above, the object recognition device 20 includes a candidate area extracted when a crossing pedestrian is recognized and a candidate area extracted when a crossing pedestrian is not recognized. In contrast, by executing the process of setting the threshold value in parallel, the same effect as that of the first embodiment can be obtained, and the situation around the vehicle can be recognized more quickly.

<Third Embodiment>
Hereinafter, a third embodiment will be described. In the first embodiment, the vehicle control device 70 executes automatic brake control, alarm output, or inter-vehicle distance control based on the processing result of the object recognition device 20, but in the third embodiment, The automatic driving control device 100 performs automatic driving control based on the processing result of the object recognition device 20. Here, the difference from the first embodiment will be mainly described, and description of functions and the like common to the first embodiment will be omitted.

  FIG. 13 is a diagram illustrating an example of a functional configuration of a vehicle control system 1A according to the third embodiment. The vehicle control system 1A includes an automatic driving control device 100 instead of the object recognition device 20 of the first embodiment. The automatic driving control device 100 includes a storage unit 102, a target lane determining unit 104, and an automatic driving control unit 110. The storage unit 102 stores, for example, information such as high-precision map information, target lane information, and action plan information.

  The target lane determining unit 104 is realized by, for example, an MPU (Micro-Processing Unit). The target lane determination unit 104 divides the route provided from the navigation device into a plurality of blocks (for example, every 100 [m] with respect to the vehicle traveling direction), and refers to the high-accuracy map information for each block. Determine the lane. The target lane determining unit 104 determines, for example, what number lane from the left to travel. For example, the target lane determination unit 104 determines the target lane so that the host vehicle can travel on a reasonable travel route for proceeding to the branch destination when there is a branch point or a merge point in the route. The target lane determined by the target lane determination unit 104 is stored in the storage unit 102 as target lane information.

  The automatic driving control unit 110 includes, for example, an action plan generation unit 112, a track generation unit 114, a travel control unit 116, and a switching control unit 118.

  The action plan generation unit 112 sets a starting point of automatic driving and / or a destination of automatic driving. The action plan generation unit 112 generates an action plan in a section between the start point and the destination for automatic driving. The action plan is composed of, for example, a plurality of events that are sequentially executed. Events include, for example, a deceleration event for decelerating the host vehicle, an acceleration event for accelerating the host vehicle, a lane keeping event for driving the host vehicle so as not to deviate from the driving lane, a lane change event for changing the driving lane, Events that follow the vehicle are included. Information indicating the action plan generated by the action plan generation unit 112 is stored in the storage unit 102 as action plan information.

  The track generation unit 114 determines and determines one of the traveling modes such as constant speed traveling, following traveling, low speed following traveling, deceleration traveling, curve traveling, obstacle avoidance traveling, lane change traveling, merging traveling, branch traveling, and the like. A trajectory candidate is generated based on the travel mode. For example, the track generation unit 114 generates a track as a collection of target positions (track points) that the reference position (for example, the center of gravity and the center of the rear wheel axis) of the host vehicle should reach at predetermined future times.

  The travel control unit 116 controls the travel drive device 72 or the brake device 74 so that the host vehicle passes the track generated by the track generation unit 114 at a scheduled time. The switching control unit 118 switches between the automatic operation mode and the manual operation mode based on a signal input from the automatic operation switch 120.

  According to the third embodiment described above, the same effect as that of the first embodiment can be obtained, and the automatic operation can be performed with higher accuracy.

  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 ... Vehicle control system 1, 10 ... Radar apparatus, 12 ... Camera, 14 ... Vehicle sensor, 20 ... Object recognition apparatus, 30 ... Image acquisition part, 32 ... Image recognition part, 34 ... Feature-value extraction part, 36 ... Processing part , 38 ... Identification unit 50 ... Surrounding situation recognition unit, 60 ... Storage unit, 72 ... Travel drive device, 74 ... Brake device

Claims (9)

An object detection unit for detecting objects around the vehicle;
A recognition unit that obtains a positional relationship between the vehicle and the object detected by the object detection unit, and that identifies a type of the object detected by the object detection unit, based on the obtained positional relationship, A recognition unit that changes the content of the process for identifying the type of the object;
Based on the identification result by the recognition unit, a driving support unit that supports driving of the vehicle;
A vehicle control system comprising: The recognizing unit does not have a threshold value used for identifying the type of the object on the traveling direction trajectory that is the trajectory that the vehicle is scheduled to travel in the target region for detecting the object that is not on the traveling direction trajectory. Set higher than the threshold used to identify the type of object,
The vehicle control system according to claim 1. When the recognition unit determines that the object is about to cross a travel direction trajectory that is a trajectory that the vehicle is going to travel, the recognition unit sets the threshold to be low for the object about to cross.
The vehicle control system according to claim 2. The recognition unit sets a lower threshold to be used for identifying the type of the object in a region farther from the vehicle with respect to the lateral position of the vehicle.
The vehicle control system according to any one of claims 1 to 3. The recognizing unit determines the type of the object in the first region that is equal to or greater than the first distance from the vehicle with respect to the traveling direction of the vehicle, as compared to the second region in front of the vehicle reaching the first distance. Set a high threshold for identification;
The vehicle control system according to any one of claims 1 to 4. The recognizing unit identifies the type of the object in the third area that is less than the second distance from the vehicle with respect to the traveling direction of the vehicle, as compared to the fourth area that is greater than or equal to the second distance from the vehicle. Set a high threshold value for
The vehicle control system according to any one of claims 1 to 5. The object detection unit is an imaging unit that captures an image,
A storage unit for storing object model information in which the distribution of the edge of the object is stored;
The recognition unit extracts edges from a time-series image captured by the imaging unit, and the similarity between the extracted edge distribution and the edge distribution stored in the object model information stored in the storage unit An index is derived for each time-series image, and the object is recognized based on a variation of the similarity index derived for each time-series image.
The vehicle control system according to any one of claims 2 to 6. Computer
Detect objects around the vehicle,
Contents of processing for obtaining a positional relationship between the vehicle and the detected object, executing processing for identifying the type of the detected object, and identifying the type of the object based on the obtained positional relationship Change
Based on the identification result, assisting the traveling of the vehicle,
Vehicle control method. On the computer,
Detect objects around the vehicle,
Contents of processing for obtaining a positional relationship between the vehicle and the detected object, causing the type of the detected object to be identified, and identifying the type of the object based on the obtained positional relationship Change
Based on the identification result, driving of the vehicle is supported.
Vehicle control program.

Images