JP2019046147A - Travel environment recognition device, travel environment recognition method, and program - Google Patents

Abstract

To provide a technology for more accurately specifying on a map a travelable region far from a vehicle than in the prior art.SOLUTION: The travel environment recognition device 10 includes an information acquisition unit, an attribute specification unit, and a map generation unit. The information acquisition unit acquires detection information from a detection device 20. The attribute specification unit specifies that a three-dimensional object detected by the detection device is a moving object based on the detection information. The map generation unit generates a map representing a travelable region. The map generation unit represents, in the map, a region including at least a part of a region representing a moving three-dimensional object which is a three-dimensional object specified as the moving object by the attribute specifying unit as a travelable region.SELECTED DRAWING: Figure 1

Description

  The present disclosure relates to a technology for recognizing a traveling environment of a vehicle.

  In Patent Document 1, in a grid-like plane map generated based on a captured image acquired by an on-vehicle camera, a road surface grid that is a grid specified as a road surface based on the captured image, and a road surface grid on the planar map There has been proposed a technique of detecting a grid between the two as a road surface area. Hereinafter, the technology described in Patent Document 1 is referred to as a conventional technology.

JP, 2016-146113, A

  However, in the prior art, since grids between road surface grids are also detected as road surface areas, in order to interpolate grids between road surface grids as road surface areas like the above road surface areas in the distant area where resolution of on-vehicle camera becomes coarse. A problem may arise in that it is difficult to accurately specify a travelable area, which is an area in which the vehicle can travel, on a plane map.

  One aspect of the present disclosure is a traveling environment recognition device that recognizes an environment in which a vehicle travels, and provides a technology for specifying a travelable area far from the vehicle on a map more accurately than the prior art.

  One aspect of the present disclosure is a traveling environment recognition device (10), which includes an information acquisition unit (S10), an attribute identification unit (S40), and a map generation unit (S50). The information acquisition unit acquires detection information representing information from a detection device (20) that detects a three-dimensional object representing an object that interferes with the travel of the vehicle around the vehicle. The attribute specifying unit specifies that the three-dimensional object is a moving object based on the detection information. The map generation unit generates a map representing a drivable area which is an area in which the vehicle can travel. The map generation unit represents, in the map, a region including at least a part of a region representing a moving three-dimensional object which is a three-dimensional object specified as a moving object by the attribute specifying unit as a travelable region.

  According to such a configuration, when a moving object is detected at a long distance based on the detection information, an area including at least a part of the area representing the moving object is identified as the travelable area, and thus the travel at a long distance The possible area can be specified on the map more accurately than the prior art.

  Another aspect of the present disclosure is a driving environment recognition method. The traveling environment recognition method acquires detection information representing information from a detection device (20) that detects a three-dimensional object that represents an object that interferes with the travel of the vehicle around the vehicle, and the three-dimensional object moves based on the detection information It specifies that it is a body, generates a map that represents a drivable area, which is an area in which the vehicle can travel, and represents a movable three-dimensional object that is a three-dimensional object identified as a moving object by the attribute specification unit An area including at least a part of the area is referred to as a travelable area.

Also in the recognition method configured as described above, the travelable area at a distance can be specified on the map more accurately than in the prior art.
Another aspect of the present disclosure is a program that causes a computer to function, and causes the computer to function as an information acquisition unit (S10), an attribute identification unit (S40), and a map generation unit (S50). The information acquisition unit acquires detection information representing information from a detection device (20) that detects a three-dimensional object representing an object that interferes with the travel of the vehicle around the vehicle. The attribute specifying unit specifies that the three-dimensional object is a moving object based on the detection information. The map generation unit generates a map representing a drivable area which is an area in which the vehicle can travel. The map generation unit represents, in the map, a region including at least a part of a region representing a moving three-dimensional object which is a three-dimensional object specified as a moving object by the attribute specifying unit as a travelable region.

Also in the program configured as described above, the travelable area at a distant place can be specified on the map more accurately than the prior art.
In addition, the reference numerals in parentheses described in this column and 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.

1 is a block diagram showing the configuration of a traveling environment recognition system 1. FIG. 6 is a flowchart of area determination processing. Explanatory drawing explaining the example which correct | amends the occupancy grid map of 1 period before. The flowchart of a map generation process. An explanatory view showing an example of a sensor model, an explanatory view explaining an example in which a stationary object is detected according to the sensor model, and an explanatory view explaining an example in which a moving three-dimensional object is detected according to the sensor model. Explanatory drawing explaining the example of the observation direction by LIDAR. Explanatory drawing explaining a production | generation of an occupancy grid map. Explanatory drawing explaining the update of an occupancy grid map. The figure which demonstrates a drivable area | region in an occupancy grid map. Explanatory drawing which shows the example of a drivable area | region when self-vehicles are going to perform a lane change. Explanatory drawing which shows the example of a drivable area | region when self-vehicles try to turn left.

  Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. Note that “identical” in the following is not limited to the same in a strict sense, and may not be exactly the same as long as the same effect can be obtained.

[1. Constitution]
The traveling environment recognition system 1 shown in FIG. 1 is a system mounted on a vehicle. Hereinafter, a vehicle on which the traveling environment recognition system 1 is mounted is also referred to as a host vehicle. The traveling environment recognition system 1 includes a signal processing ECU 10. The traveling environment recognition system 1 may include a periphery monitoring sensor 20, a vehicle speed sensor 30, a yaw rate sensor 40, a vehicle control device 50, and a display device 60. Note that, in the present embodiment, the traveling environment recognition system 1 may not include the wireless communication device 70 shown by a dotted line in FIG. 1. An example in which the traveling environment recognition system 1 includes the wireless communication device 70 will be described in another embodiment.

  The surrounding area monitoring sensor 20 includes a sensor that detects a three-dimensional object present around the vehicle. The three-dimensional object referred to here is an object that interferes with the traveling of the vehicle. The three-dimensional object may include, for example, a person, a car, a fixed object, and the like. The perimeter monitoring sensor 20 may be, for example, LIDAR (Laser Detection and Ranging), a radar, a monocular camera, a stereo camera or the like. In the present embodiment, the surrounding area monitoring sensor 20 includes LIDAR.

  The LIDAR is provided at the front of a vehicle, emits pulse-like laser light from the light emitting unit, and receives the laser light reflected by an object present in front of the vehicle at the light receiving unit. The objects present in front of the vehicle referred to here include not only three-dimensional objects but also objects that do not disturb the traveling of the vehicle, such as the road surface. In addition, the observation point said to the following is the position where the object was observed, and refers to the position where reflection of the laser beam was observed in this embodiment.

  The LIDAR outputs, to the signal processing ECU 10, measurement information indicating the time difference between the emission time of the laser light and the light reception time of the reflected light, the reflected light intensity, and the irradiation angle of the laser light from which the reflected light is obtained. Based on the measurement information, the signal processing ECU 10 represents the detection result of the object by three-dimensional coordinates (x, y, z) whose coordinate origin is the attachment position of the surrounding area monitoring sensor 20 (hereinafter, sensor position). Identify.

  The vehicle speed sensor 30 and the yaw rate sensor 40 are sensors for detecting the amount of movement of the vehicle. The vehicle speed sensor 30 detects the traveling speed of the vehicle, and the yaw rate sensor 40 detects a yaw rate acting on the vehicle. The signal processing ECU 10 calculates an amount of movement of the vehicle based on detection signals from the vehicle speed sensor 30 and the yaw rate sensor 40. Specifically, the traveling direction and displacement of the vehicle are calculated as momentum.

  The vehicle control device 50 is an electronic control device including a CPU, a ROM, a RAM, and the like, and the vehicle travels in the travelable region based on the recognition result of the travelable region output from the signal processing ECU 10 Control the engine, brakes, etc. The travelable area refers to an area where the vehicle can travel on the map. For example, the road surface is included in the drivable area.

  The display device 60 is a device that displays, for example, an image or the like representing the travelable area, based on the recognition result of the travelable area output from the signal processing ECU 10. The display device 60 may be, for example, a screen of a navigation device or a head-up display.

  The signal processing ECU 10 includes a microcomputer (hereinafter, microcomputer) having a CPU, and a semiconductor memory (hereinafter, memory) such as a ROM, a RAM, and a flash memory. Each function implemented by the signal processing ECU 10 is implemented by the CPU executing a program stored in the non-transitional tangible storage medium. In this example, the semiconductor memory corresponds to a non-transitional tangible storage medium storing a program. Also, by executing this program, a method corresponding to the program is executed. The number of microcomputers constituting the signal processing ECU 10 may be one or more.

  The signal processing ECU 10 recognizes the traveling environment around the vehicle based on the detection signals output from the surroundings monitoring sensor 20, the vehicle speed sensor 30, and the yaw rate sensor 40. In the present embodiment, the signal processing ECU 10 particularly recognizes the traveling environment in front of the vehicle. The traveling environment referred to herein is a road environment, and can be represented, for example, by the presence or absence of the three-dimensional object which is an obstacle to the traveling of the vehicle.

  Specifically, the signal processing ECU 10 determines the relative position of the observation point with respect to the vehicle, that is, the distance and direction of the observation point with respect to the vehicle, based on the detection information which is information from the surrounding area monitoring sensor 20. It has a function to generate point group data by converting it to a position in an absolute coordinate system.

Here, the point cloud data indicates a set of data of three-dimensional observation points in each irradiation area obtained by LIDAR as the periphery monitoring sensor 20. The data of the observation point is represented by (x, y, z) coordinates with the sensor position as an origin.

  The absolute coordinate system is a unique coordinate system in which the position of the vehicle, that is, the sensor position, is set as an origin and the X and Y axes are set in an arbitrary direction, and does not directly correspond to the latitude and longitude coordinates. The absolute coordinate system used in the present embodiment has the sensor position as the origin, but is not limited to this, and an arbitrary point may be the origin.

  In the case where a monocular camera or a stereo camera is used as the periphery monitoring sensor 20, point cloud data may be generated based on the detection information, using the image captured by the monocular camera or the stereo camera as detection information. .

Further, the signal processing ECU 10 has a function of recognizing the traveling environment around the vehicle by executing the area determination processing based on the detection information, that is, the point cloud data.
[2. processing]
[2-1. Overview]
Next, an outline of processing executed in the traveling environment recognition system 1 of the present embodiment will be described.

  The travel environment recognition system 1 generates an occupancy grid map (hereinafter referred to as OGM) as a map for recognizing a travel environment around a vehicle. The OGM is a map that represents the traveling environment with probability, and is a map that represents the travelable area. The drivable area represents an area on which the vehicle can travel on the OGM.

  Specifically, the signal processing ECU 10 meshes an absolute coordinate system in which the position of the vehicle at a certain time t = 0, that is, the sensor position, is the origin, the vehicle horizontal direction is the X axis, and the vehicle longitudinal direction is the Y axis. The OGM is generated by dividing into a shape, that is, a grid shape, and storing the existence probability of a solid object for each grid.

  The existence probability of the three-dimensional object which is an obstacle to the traveling of the vehicle is obtained based on the information from the surrounding area monitoring sensor 20. The existence probability is the probability that a three-dimensional object is present, and is set to be larger in the presence of a three-dimensional object than in the absence of a three-dimensional object. In OGM, a region in which the probability of existence of three-dimensional objects is low, that is, a grid in which the probability of existence of three-dimensional objects is low corresponds to the travelable region.

The grid size is determined to an optimal size according to the detection accuracy of the surrounding area monitoring sensor 20 arbitrarily or. The grid is assumed to be a 1 m square grid in this embodiment.
The generated OGM and the recognition result of the drivable area based on the OGM are used by the vehicle control apparatus 50 by a collision system application for realizing forward monitoring control such as adaptive cruise control and pre-crash safety control, for example. Also good.

[2-2. Processing procedure]
The signal processing ECU 10 performs the area determination process shown in FIG. The area determination process is a process that is started, for example, when the power of the vehicle is turned on, and is a process that is repeatedly performed after each fixed cycle.

As shown in FIG. 2, the signal processing ECU 10 first acquires sensor data from various sensors at S10.
Specifically, the signal processing ECU 10 acquires detection information from the surrounding area monitoring sensor 20. Then, the signal processing ECU 10 generates the point cloud data based on the detection information by processing separate from the area determination processing, and acquires the generated point cloud data.

Further, the signal processing ECU 10 acquires the traveling speed of the vehicle from the vehicle speed sensor 30, and acquires the yaw rate acting on the vehicle from the yaw rate sensor 40.
Subsequently, in S20, the signal processing ECU 10 performs determination of a three-dimensional object using point cloud data generated based on the detection information. Specifically, the signal processing ECU 10 determines "a point on a three-dimensional object" and "a point on the road surface" based on the height information of the point cloud data, that is, the z coordinate. Here, the signal processing ECU 10 determines that the observation point corresponding to the point group data is “a point on a three-dimensional object” when the height information of the point group data is equal to or more than a predetermined road surface height threshold. . On the other hand, when the height information of the point cloud data is less than the road surface height threshold, the signal processing ECU 10 determines that the observation point corresponding to the point cloud data is “a point on the road surface”. The road surface height threshold is previously determined based on an experiment or the like, and is stored in advance in the memory. The threshold value of the road surface height may be, for example, about several cm to several tens cm.

Next, in S30, the signal processing ECU 10 acquires an occupancy grid map (hereinafter referred to as OGM- ₁ ) which is an occupancy grid map generated in S70 to be described later, which is generated one cycle earlier, and performs OGM- ₁ coordinate origin of such a sensor position of the current period, in consideration of the amount of movement of the coordinate origin of one period to the OGM _-1, corrects the OGM _-1.

FIG. 3 shows an actual traveling environment (hereinafter, actual environment), and in the actual environment one cycle before, a three-dimensional object (hereinafter, stationary object) stationary ahead of the vehicle 100 exists. An example is shown to correct OGM- ₁ when there are walls on the left and right of. The movement amount of the coordinate origin at one cycle, that is, the movement amount of the surrounding area monitoring sensor 20 is calculated based on the momentum of the vehicle 100 during one cycle. In FIG. 3, the arrow in the real environment one cycle before represents the momentum of the vehicle 100. Note that the signal processing ECU 10 estimates the amount of movement of the host vehicle 100 based on the vehicle speed and the yaw rate acquired in S10.

  Subsequently, at S40, the signal processing ECU 10 specifies that the three-dimensional object determined at S20 is a moving object based on the point cloud data. The moving body refers to an object moving at a speed higher than a predetermined speed. The predetermined speed referred to here may be, for example, the moving speed of a pedestrian or more. The moving body may include, for example, a vehicle, a pedestrian, a bicycle, and the like.

  Specifically, when detecting the moving object, the signal processing ECU 10 first clusters the observation point group on the three-dimensional object determined as the “point on the three-dimensional object” in S20, and detects each three-dimensional object. recognize. That is, the signal processing ECU 10 generates a cluster in which groups of observation points on the three-dimensional object whose distances are shorter than a predetermined threshold are combined, and recognizes the clusters as individual three-dimensional objects.

  Furthermore, the signal processing ECU 10 performs tracking. In the present embodiment, the signal processing ECU 10 tracks a cluster using the latest point cloud data, which is point cloud data of the current cycle, and point cloud data acquired one cycle earlier. However, the present invention is not limited to this, and tracking may be performed using the latest point cloud data and point cloud data acquired during the latest past plural cycles.

  Specifically, the signal processing ECU 10 monitors the movement state of clusters using, for example, a Kalman filter, a particle filter, etc., and tracks clusters having similar sizes and shapes before and after movement as those representing the same three-dimensional object. .

Then, the signal processing ECU 10 recognizes the size, moving speed, and the like of the tracked three-dimensional object, and specifies that the three-dimensional object is a moving object based on the size, moving speed, and the like. Identify the attribute. For example, attributes of a three-dimensional object such as a vehicle, a pedestrian, a bicycle, etc. are specified. Thus, the signal processing ECU 10 determines that the three-dimensional object is a moving object, specifies the size and moving speed of the moving object, and further specifies its attributes.

  When the single-eye camera or the stereo camera is used as the surroundings monitoring sensor 20, the signal processing ECU 10 uses the classifier learned based on the teacher data based on the captured image captured by the single-eye camera or the stereo camera. It may be determined that the object is a moving object, the size and moving speed of the moving object may be specified, and further the attribute thereof may be specified.

Next, in S50, the signal processing ECU 10 executes map generation processing. In the map generation process, an occupied grid map (hereinafter referred to as OGM ₀ ) in the current cycle is generated. In the present embodiment, a two-dimensional map that is an XY plane is generated. Details of the map generation process will be described based on the flowchart shown in FIG.

  In S110, the signal processing ECU 10 uses the point group data generated based on the detection information in S10 to calculate the existence probability of the solid object in the grid including the observation point. FIG. 5 is a sensor model which defines the correspondence between the position on a straight line passing through the light emitting part of the LIDAR as the periphery monitoring sensor 20 and the observation point, and the existence probability of a three-dimensional object. As shown in FIG. 6, the signal processing ECU 10 calculates the existence probability for each grid according to the sensor model for each of the plurality of orientations to which the laser light is irradiated.

  In the sensor model, the existence probability at an observation point observed as “a point on a three-dimensional object” is set to a value close to 1 (hereinafter, existence probability α). In the present embodiment, the existence probability α is set to one. Further, in the sensor model, the existence probability (hereinafter, existence probability β) at an observation point observed as “a point on the road surface” is set to a value close to zero. In the present embodiment, the existence probability β is set to zero.

  In addition, the existence probability (hereinafter referred to as “presence area”) of a region (hereinafter, unknown region) which is at a position before the observation point observed as “a point on a three-dimensional object” and whether or not a three-dimensional object exists. The probability ε) is a value close to 0 and is set to a value larger than the existence probability β and smaller than the existence probability M described later. Further, the existence probability of the position behind the observation point (hereinafter, the existence probability M) is set to an intermediate value between α and β. In the present embodiment, the existence probability M is set to 0.5.

  This is because it is inferred that no three-dimensional object exists in front of the observation point observed as "a point on a three-dimensional object", and it is unclear whether a three-dimensional object exists behind the observation point. . In addition, a sensor model is not limited to what is shown in FIG. 5, It may be set arbitrarily.

Subsequently, in S120, the signal processing ECU 10 calculates the existence probability in all of the plurality of azimuths to which the laser light is irradiated, stores the existence probability for each grid, and becomes a base of OGM _0. Generate a basemap. FIG. 7 shows an example of a base map corresponding to an actual environment in which a moving vehicle exists ahead of the host vehicle 100 and walls exist on the left and right of the host vehicle 100. In FIG. 7, observation points on a three-dimensional object are indicated by triangles. In particular, the observation points on the moving solid are indicated by black triangles.

Then, the signal processing ECU10 is in S130, the base map, the existence probability to be stored in the grid represents a moving three-dimensional object is updated to a value below the traveling area threshold value, generates the OGM _0. The moving three-dimensional object represents the three-dimensional object specified as the moving object at S40. The travel area threshold indicates a threshold used in threshold determination performed in S70 described later. The travel area threshold may be, for example, a value less than the existence probability M. Also, the traveling region threshold may be, for example, a value less than the existence probability M and less than the existence probability ε.

In the present embodiment, a region representing a moving solid on OGM ₀ (hereinafter referred to as a moving region), that is, a grid representing a moving solid consists of at least one grid. For example, on OGM _0, the vehicle is moving as the moving three-dimensional object (hereinafter, moving vehicles) region representing a (hereinafter, the mobile vehicle area), i.e. a grid representing the moving vehicle, a plurality of grids, OGM ₀ It is determined in advance so as to be the same as the shape and size of the vehicle 100 at the top. The signal processing ECU 10 specifies the moving vehicle area such that a grid representing an observation point on the moving vehicle is included. FIG. 7 shows an example of OGM ₀ generated based on the above base map. In OGM ₀ shown in FIG. 7, a region P including six grids represents a moving vehicle region.

Note that the shape and size of the mobile object area may be set for each attribute of the mobile object. Alternatively, the shape and size of the mobile region may be set to be constant regardless of the attribute of the mobile. Moreover, in the present embodiment, as described above, the moving vehicle area is set to the same size as the shape and size of the host vehicle 100 on the OGM ₀ , but the present invention is not limited to this.

The moving vehicle region may be a region larger than the host vehicle 100 on the OGM ₀ in consideration of the detection error, or may be a small region in consideration of the detection error. Also, for example, the moving vehicle area is set to the same shape and size as a part of the own vehicle 100 on the OGM ₀ in the front-rear direction, such as the rear half of the own vehicle 100 and the rear quarter of the own vehicle 100. It may be done. Thus, the moving object region comprises at least a portion of a region representing a moving three-dimensional object itself on OGM _0.

In the base map, the presence probability α or the presence probability M is stored as a presence probability in the grid representing the moving solid. In S130, the signal processing ECU 10 updates the presence probability stored in the grid representing the moving three-dimensional object to the presence probability β, that is, 0. Thus, OGM ₀ is generated. The generated OGM ₀ is stored in the memory. Above, the signal processing ECU 10 ends the map generation processing, and shifts the processing to S60.

Return to FIG. 2 and continue the description. The signal processing ECU 10 updates the OGM in S60. Specifically, as shown in FIG. 8, the signal processing ECU 10 stores, as a new OGM ₀ , a combination of OGM- ₁ and OGM ₀ in the memory. In the real environment of FIG. 8, a moving vehicle is present ahead of the host vehicle 100, the mobile vehicle is moving between the time t _-1 time t _0. For this reason, in the corrected OGM- ₁ , the existence probability β is stored in the grid corresponding to the area according to the trajectory of the moving vehicle. The corrected OGM- ₁ and the OGM ₀ corresponding to the real environment at time t ₀ are superimposed to generate a new updated OGM ₀ .

Subsequently, at S70, the signal processing ECU 10 specifies, in OGM ₀ , a grid whose presence probability is less than the travel area threshold as the travelable area.
Thereby, as shown in FIG. 9, the grid including the observation point observed as "a point on the road surface" is specified as the travelable area. Furthermore, an area including at least a part of the area representing the moving solid, specifically, the moving vehicle area is specified as the drivable area.

Note that the signal processing ECU 10 may specify a stationary object in OGM ₀ . Specifically, the signal processing ECU 10 specifies a grid in which the presence probability that is equal to or more than a predetermined stationary object threshold is stored as a region representing the stationary object. The stationary object threshold is a value larger than the presence probability M and can be set to a value equal to or more than the presence probability α. Further, the signal processing ECU 10 may specify, in the OGM ₀ , an area other than the travelable area and the area representing the stationary object as the unknown area. Above, signal processing ECU10 ends field judging processing.

[3. effect]
According to the embodiment described above, the following effects can be obtained.
(3a) The signal processing ECU 10 acquires detection information in S10. In S40, the signal processing ECU 10 specifies that the three-dimensional object is a moving three-dimensional object based on the detection information. In S50, the signal processing ECU 10 generates the OGM as a map representing the travelable area, and represents an area including at least a part of the area representing the moving three-dimensional object in the map as the travelable area.

  As a result, when a moving solid is detected at a distance based on the detection information, a region including at least a part of the region representing the moving solid is identified as the travelable region, so , And can be specified more accurately on the map representing the travelable area than the prior art.

  Here, it is an apparatus for comparing with signal processing ECU10, Comprising: The comparison which was not comprised so that a solid thing was not distinguished to a stationary object and a moving solid thing, and the area | region where a solid thing existed was not specified as a traveling possible area | region Assume a device. In the comparison device, for example, as shown in FIG. 5, when a moving vehicle as a moving three-dimensional object is present on a slope ahead of the own vehicle, the moving vehicle is simply identified as a three-dimensional object. That is, the area on the slope can not be identified as the travelable area on the map.

  On the other hand, the signal processing ECU 10 distinguishes a three-dimensional object into a stationary object and a moving three-dimensional object, and specifies an area in which a moving three-dimensional object is present as a travelable area. For this reason, as shown in FIG. 5, for example, when there is a moving vehicle as a moving solid object on the slope ahead of the own vehicle as shown in FIG. 5, the area representing the moving vehicle, ie, the area on the slope Can be identified as the drivable area in

Thus, the signal processing ECU 10 can specify a slope ahead of the host vehicle with high accuracy.
(3b) The map representing the travelable area includes a plurality of grids. In S120, the signal processing ECU 10 calculates the presence probability for each grid, and generates an OGM that is the above map and stores the presence probability for each grid. In S130, the signal processing ECU 10 updates, in the OGM, the existing probability stored in the grid representing the moving three-dimensional object to a value close to the existing probability when no three-dimensional object is present and less than a predetermined traveling area threshold. . In S70, the signal processing ECU 10 specifies, in the OGM, a grid whose existence probability is less than the travel area threshold as the travelable area.

As a result, it is possible to specify a drivable area on the OGM that represents the traveling environment of the host vehicle by probability.
(3c) In the signal processing ECU 10, the traveling area threshold value is a value smaller than an intermediate value M between the existence probability α when there is a three-dimensional object and the existence probability β when there is no three-dimensional object.

As a result, it is possible to specify an area representing an existence probability closer to the existence probability in the absence of a three-dimensional object by threshold determination using a traveling area threshold value.
(3d) In S130, the signal processing ECU 10 sets the presence probability stored in the grid representing the moving solid in the occupancy grid map to the presence probability β, specifically 0, indicating that there is no solid. It is configured. As a result, the grid representing the moving three-dimensional object can be reliably identified as the travelable area.

In the above embodiment, the signal processing ECU 10 corresponds to a traveling environment recognition device, an information acquisition unit, an attribute identification unit, a map generation unit, a grid map unit, an update unit, and an area identification unit. Equivalent to. Also, S10 corresponds to processing as an information acquisition unit, S40 corresponds to processing as an attribute identification unit, S50 corresponds to processing as a map generation unit, and S70 corresponds to processing as an area identification unit. S120 corresponds to processing as a grid map unit, and S130 corresponds to processing as an updating unit. Also, the exclusive grid map corresponds to the map and the grid map.

[4. Modified example]
(4a) Modified Example 1
In the signal processing ECU 10, the updating unit may be configured to obtain the identification accuracy and calculate the presence probability stored in the grid representing the moving three-dimensional object smaller as the identification accuracy is larger. The identification accuracy is a value that represents the degree of accuracy with which the attribute identification unit identifies a moving solid object, and is a value represented by a larger value as the accuracy is higher.

As a result, the larger the identification accuracy, the easier the area representing the moving three-dimensional object can be identified as the travelable area.
Specifically, in the signal processing ECU 10, the information acquisition unit may be configured to repeatedly acquire the detection information. The updating unit is configured to specify the number of times the moving solid is continuously detected based on the detection information, and set the specific accuracy to a larger value as the number of times the moving solid is continuously detected is larger. It may be done. Then, the updating unit may be configured to calculate the presence probability stored in the grid representing the moving three-dimensional object based on the specific accuracy.

  As a result, as the number of times in which the moving three-dimensional object is continuously detected increases, the identification accuracy increases, and the presence probability stored in the grid representing the moving object is calculated to be small. That is, as the number of times in which the moving three-dimensional object is detected continuously increases, the region representing the moving three-dimensional object can be easily identified as the travelable region.

(4b) Modification 2
In the signal processing ECU 10, the attribute specifying unit may be configured to specify that the moving three-dimensional object has the same size as the host vehicle. Then, the updating unit stores the existing probability stored in the grid representing the moving three-dimensional object specified as the same size as the own vehicle as the moving three-dimensional object not specified as the same size as the own vehicle. It may be configured to be smaller than the existence probability stored in the grid to be represented.

  As a result, it is possible to accurately specify the drivable area for the host vehicle in the OGM so that the host vehicle is neither too large nor too small for the host vehicle.

(4c) Modification 3
In the signal processing ECU 10, the area specifying unit is configured to generate an image representing the travel area on the occupied grid map, and to display the generated image on the display device 60. May be Further, in the signal processing ECU 10, the area specifying unit may be configured to output the recognition result of the OGM or the travelable area in the OGM to the vehicle control device 50.

(4d) Modification 4
In the signal processing ECU 10, even when there are a plurality of moving solid objects, for example, a plurality of moving vehicles around the own vehicle 100 as shown in FIG. 10 and FIG. The possible area may be specified. Specifically, the signal processing ECU 10 may generate a base map and update the presence probability stored in the grid representing each of the plurality of moving vehicles in the base map, as in the above embodiment.

  As a result, for example, when the host vehicle 100 as shown in FIG. 10 attempts to make a lane change, or when the host vehicle 100 turns left as shown in FIG. Also in the case, the drivable area on the OGM can be specified.

[5. Other embodiments]
As mentioned above, although embodiment of this indication was described, this indication can be variously deformed and implemented, without being limited to the above-mentioned embodiment.

  (5a) In the above-mentioned embodiment, although OGM was generated as a map showing a run possible area, a map showing a run possible area is not limited to this. For example, the signal processing ECU 10 may generate, as the map, a map in which an area representing a moving three-dimensional object is specified as a travelable area on a map created based on map data.

  (5b) Although LIDAR is used as the periphery monitoring sensor 20 in the above embodiment, the present invention is not limited to this. For example, a radar may be used as the perimeter monitoring sensor 20. Further, for example, a monocular camera or a stereo camera may be used as the periphery monitoring sensor 20.

  (5c) In the above embodiment, the signal processing ECU 10 is configured to identify the travelable area in front of the host vehicle. However, the present invention is not limited to this. The signal processing ECU 10 is configured to identify a movable three-dimensional object in an arbitrary direction with respect to the host vehicle, such as the host vehicle backward, the host vehicle right, the host vehicle left, etc. It is also good.

  (5d) In the above embodiment, the signal processing ECU 10 mounted on the host vehicle performs the traveling area detection process to identify the travelable area of the host vehicle 100 on the map, but the present invention is limited thereto. It is not a thing. As indicated by a dotted line in FIG. 1, the server 9 capable of wirelessly communicating with the wireless communication device 70 mounted on the host vehicle may be configured to specify the travelable area of the host vehicle 100 on the map. The server 9 is installed outside the host vehicle, and includes an electronic control unit including a CPU, a memory such as a ROM, a RAM, and the like. In addition, a communication device capable of wireless communication with the wireless communication device 70 is provided.

  In this case, the server 9 may be configured to perform the same process as the traveling area detection process shown in FIG. Here, the signal processing ECU 10 may be configured to transmit detection results of various sensors to the server 9 via the wireless communication device 70. The detection results of the various sensors include, for example, point cloud data based on detection information from the surrounding area monitoring sensor 20 and detection results of the vehicle speed sensor 30 and the yaw rate sensor 40.

  On the other hand, the server 9 may be configured to acquire the received detection information instead of S10 of FIG. In addition, the server 9 may be configured to execute the processing of S20 to S70 of FIG. 2 based on the received detection information, and specify a travelable area on the OGM. Then, the server 9 may be configured to transmit information representing the travelable area on the OGM to the wireless communication device 70.

  The signal processing ECU 10 acquires information representing a travelable area on the OGM from the server 9 via the wireless communication device 70, and outputs an instruction to the vehicle control device 50 or the display device 60 based on the information. May be configured.

(5e) The multiple functions of one component in the above embodiment may be realized by multiple components, or one function of one component may be realized by multiple components. . Also, a plurality of functions possessed by a plurality of components may be realized by one component, or one function realized by a plurality of components may be realized by one component. In addition, part of the configuration of the above embodiment may be omitted. In addition, at least a part of the configuration of the above-described embodiment may be added to or replaced with the configuration of the other above-described embodiment. In addition, all the aspects contained in the technical thought specified from the wording described in the claim are an embodiment of this indication.

  (5f) A program for causing the signal processing ECU 10 to function in addition to the signal processing ECU 10, the traveling environment recognition system 1, and the server 9 described above, a program for causing the server 9 to function, non transition of semiconductor memory etc. The present disclosure can also be realized in various forms, such as physical recording media and driving environment recognition methods.

  10 signal processing ECU, 20 ambient environment sensors.

Claims (9)

A traveling environment recognition device (10),
An information acquisition unit (S10) for acquiring detection information representing information from a detection device (20) for detecting a three-dimensional object representing an object that obstructs the traveling of the vehicle around the vehicle;
An attribute specifying unit (S40) for specifying that the three-dimensional object is a moving object based on the detection information;
A map generation unit that generates a map representing a travelable area, which is an area in which the vehicle can travel, and in the map, a movable solid that is the solid object identified as the moving object by the attribute identification unit A map generation unit (S50) representing an area including at least a part of an area representing an object as the travelable area;
A traveling environment recognition device comprising: The traveling environment recognition device according to claim 1, wherein
The map comprises a plurality of grids,
The map generation unit
Based on the detection information, an existence probability which is a probability that the three-dimensional object exists and which is set larger when the three-dimensional object is present than when the three-dimensional object does not exist is calculated for each grid. A grid map unit (S120) for generating a grid map which stores the existence probability for each grid;
An updating unit (S130) configured to update the presence probability stored in the grid representing the moving three-dimensional object to a value close to the presence probability when the three-dimensional object does not exist in the grid map;
Equipped with
The driving environment recognition device
An area specifying unit (S70) for specifying the grid having the existence probability less than the travel area threshold in the grid map as the travelable area
A traveling environment recognition device further comprising: The traveling environment recognition device according to claim 2,
The traveling environment recognition device, wherein the traveling area threshold value is a value smaller than an intermediate value between the existence probability when the three-dimensional object exists and the existence probability when the three-dimensional object does not exist. The traveling environment recognition device according to claim 2 or 3, wherein
The update unit is a value representing a degree of accuracy in which the attribute specification unit specifies the moving solid object, and acquires a specification accuracy represented by a larger value as the accuracy is higher, and a grid representing the moving solid object A traveling environment recognition device configured to calculate an existing probability stored in the smaller as the specific accuracy is larger. The driving environment recognition apparatus according to claim 4, wherein
The information acquisition unit is configured to repeatedly acquire the detection information;
The updating unit specifies the number of times the moving solid object is continuously detected by the detection device based on the detection information, and the specification is performed as the number of times the moving solid object is continuously detected is larger. A traveling environment recognition device configured to calculate a presence probability stored in a grid representing the moving three-dimensional object based on the specified accuracy, with a large value of accuracy. The environment recognition device according to any one of claims 2 to 5, wherein
The attribute specifying unit is configured to specify that the moving solid object has the same size as the vehicle.
The mobile stereo not having the same size as the vehicle identified as the presence probability stored in the grid representing the mobile solid identified to be the same size as the vehicle. A traveling environment recognition device configured to be smaller than an existence probability stored in the grid representing an object. The traveling environment recognition device according to claim 2,
The traveling environment recognition device according to claim 1, wherein the updating unit is configured to set the presence probability stored in the grid representing the moving three-dimensional object in the grid map as the presence probability indicating that the three-dimensional object is absent. Acquiring detection information representing information from a detection device (20) for detecting a three-dimensional object representing an object that obstructs the traveling of the vehicle around the vehicle;
Identifying that the three-dimensional object is a moving object based on the detection information;
A map representing a drivable area, which is an area in which the vehicle can travel, is generated, and at least at least the area representing the movable three-dimensional object which is the three-dimensional object specified by the attribute specification unit in the map. A travel environment recognition method, wherein an area including a part is represented as the travelable area. An information acquisition unit (S10) for acquiring detection information representing information from a detection device (20) for detecting a three-dimensional object representing an object that obstructs the traveling of the vehicle around the vehicle;
An attribute specifying unit (S40) for specifying that the three-dimensional object is a moving object based on the detection information;
A map generation unit that generates a map representing a travelable area, which is an area in which the vehicle can travel, and in the map, a movable solid that is the solid object identified as the moving object by the attribute identification unit As a map generation unit (S50) representing an area including at least a part of an area representing an object as the travelable area,
A program that makes a computer function.