JP7144271B2 - Road shape recognition device - Google Patents

Description

The present disclosure relates to a road shape recognition device.

As a road shape recognition device, there is known one that calculates a road model representing the road shape with a line from the edge points of the division lines of the road recognized by a camera and updates it as needed (Patent Document 1).

JP-A-2002-109695

However, when the lane markings far away from the vehicle cannot be recognized due to obstacles such as preceding vehicles, the road model is calculated using only the edge points of the lane markings near the vehicle. There is a risk. Therefore, a technology capable of appropriately updating the road model has been desired.

The present disclosure has been made to solve the above problems, and can be implemented as the following modes.

According to a first aspect of the present disclosure, there is provided a road shape recognition device (110) mounted on a vehicle (10) having a perimeter sensor (120). The road shape recognition device includes a peripheral environment recognition unit (111) that recognizes at least one or more of the shape of the road division line detected by the peripheral sensor, the shape of the roadside object, and the movement history of other vehicles as peripheral information. and a point sequence estimating unit (112) for estimating a point sequence representing the road shape with points using the peripheral information, and determining the distance from the vehicle to the furthest point in the point sequence as the recognition distance. a recognition distance determination unit (113), a road model calculation unit (114) that calculates a road model representing the road shape with lines using the peripheral information, and an output unit (115) that outputs the road model The road model calculation unit updates the previously calculated road model to the newly calculated road model when the difference between the recognition distance and the previously determined recognition distance is equal to or less than a predetermined threshold distance. , when the difference between the recognition distance and the previously determined recognition distance is longer than a predetermined threshold distance, the previously calculated road model is not updated, and the recognition distance determination unit determines the recognition distance and the past If the difference from the previously determined recognition distance is longer than a predetermined threshold distance, the recognition distance is updated to a value obtained by subtracting the movement distance of the vehicle from the previously determined recognition distance .

According to a second aspect of the present disclosure, there is provided a vehicle-mounted road shape recognition device having a surrounding sensor and a vehicle condition sensor (126). The road shape recognition device includes a surrounding environment recognition unit that recognizes as surrounding information at least one or more of a shape of a road division line detected by the surrounding sensor, a shape of a roadside object, and a movement history of another vehicle; A point sequence estimating unit for estimating a point sequence representing a road shape with points using surrounding information; and a recognition distance determining unit for determining a distance from the vehicle to a point farthest from the vehicle in the point sequence as a recognition distance. , a road model calculation unit that calculates a road model representing a road shape with a line using the peripheral information, and an output unit that outputs the road model, wherein the road model calculation unit calculates the recognition distance and the past If the difference from the determined recognition distance is equal to or less than a predetermined threshold distance, the road model calculated in the past is updated to the road model, and the difference between the recognition distance and the previously determined recognition distance is a predetermined threshold. If the distance is longer than the distance, the previously calculated road model is corrected using at least the speed and yaw rate of the vehicle detected by the vehicle condition sensor.

According to a third aspect of the present disclosure, there is provided a road shape recognition device mounted on a vehicle having a peripheral sensor and a vehicle condition sensor. The road shape recognition device includes a surrounding environment recognition unit that recognizes as surrounding information at least one or more of a shape of a road division line detected by the surrounding sensor, a shape of a roadside object, and a movement history of another vehicle; A point sequence estimating unit for estimating a point sequence representing a road shape with points using surrounding information; and a recognition distance determining unit for determining a distance from the vehicle to a point farthest from the vehicle in the point sequence as a recognition distance. , a road model calculation unit that calculates a road model representing a road shape with a line using the peripheral information, and an output unit that outputs the road model, wherein the road model calculation unit calculates the recognition distance and the past If the difference from the determined recognition distance is equal to or less than a predetermined threshold distance, the road model calculated in the past is updated to the road model, and the difference between the recognition distance and the previously determined recognition distance is a predetermined threshold. distance and the speed of the vehicle detected by the vehicle condition sensor is lower than a predetermined speed, the road model calculated in the past is not updated, and the recognized distance and the past is longer than a predetermined threshold distance and the speed of the vehicle detected by the vehicle condition sensor is equal to or higher than a predetermined speed, the vehicle condition At least the speed and yaw rate of the vehicle detected by the sensors are used to correct the previously calculated road model.

According to the road shape recognition devices of these forms, the road model calculation unit converts the previously calculated road model into the surrounding information when the difference between the recognition distance and the previously determined recognition distance is equal to or less than the predetermined threshold distance. , the road model can be updated appropriately.

1 is a schematic diagram showing the configuration of an automatic driving system; FIG. It is a flow chart showing road shape recognition processing. It is a figure which shows an example of a dot sequence. FIG. 10 is a diagram showing another example of a dot sequence; It is a figure which shows the example of a road model. It is a flow chart showing road shape recognition processing in a 2nd embodiment. It is a figure which shows the comparative example which held the past road model. It is a figure which shows the example which corrected the road model. It is a flow chart showing road shape recognition processing in a 3rd embodiment. FIG. 10 is a diagram showing another example of a dot sequence; FIG. 10 is a diagram showing still another example of a sequence of points;

A. First embodiment:
As shown in FIG. 1 , the vehicle 10 has an automatic driving control system 100 . In this embodiment, the automatic driving control system 100 executes automatic driving of the vehicle 10 . The driver can drive manually. In the present embodiment, the automatic driving control system 100 includes a road shape recognition device 110, a peripheral sensor 120, an own vehicle condition sensor 126, a driving control unit 210, a driving force control ECU (Electronic Control Unit) 220, and a controller. A power control ECU 230 and a steering control ECU 240 are provided. Road shape recognition device 110 , driving control unit 210 , driving force control ECU 220 , braking force control ECU 230 and steering control ECU 240 are connected via in-vehicle network 250 .

The road shape recognition device 110 includes a surrounding environment recognition section 111 , a point sequence estimation section 112 , a recognition distance determination section 113 , a road model calculation section 114 and an output section 115 . The road shape recognition device 110 is composed of a central processing unit (CPU), a microcomputer configured with a RAM and a ROM, etc. The microcomputer executes pre-installed programs to realize the functions of these units. . However, some or all of the functions of these units may be realized by hardware circuits.

The peripheral environment recognition unit 111 recognizes peripheral information using the detection signal of the peripheral sensor 120 . More specifically, the surrounding environment recognition unit 111 recognizes at least one or more of the shape of road markings detected by the surrounding sensor 120, the shape of roadside objects, and the movement history of other vehicles as surrounding information. . The sequence of points estimating unit 112 estimates a sequence of points representing the road shape of the travel route of the vehicle 10 using the recognition information. The recognition distance determining unit 113 determines the distance from the vehicle 10 to the farthest point from the vehicle 10 in the point sequence estimated by the point sequence estimating unit 112 as the recognition distance. The road model calculation unit 114 calculates a road model representing the road shape with lines using the surrounding information. The road model calculation unit 114 calculates a road model using, for example, a Kalman filter or a least-squares method. The output unit 115 outputs the road model calculated by the road model calculation unit 114 to the driving control unit 210 and the like through the in-vehicle network 250 .

Perimeter sensor 120 includes camera 122 and peripheral object sensor 124 . The camera 122 acquires an image by imaging the surroundings of the own vehicle. The surrounding object sensor 124 detects the surrounding conditions of the own vehicle. Examples of the peripheral object sensor 124 include object sensors using reflected waves such as laser radar, millimeter wave radar, and ultrasonic sensors. In this embodiment, the surrounding environment recognition unit 111 recognizes the presence and position of left and right lane markings on the road on which the vehicle is traveling, and the existence of roadside objects, based on the image captured by the camera 122 and the detection result of the surrounding object sensor 124. Its position, presence of other vehicles, position, size, distance, traveling direction, speed, yaw angular velocity, etc. are detected. The surrounding environment recognition unit 111 may detect part or all of these information through inter-vehicle communication with other vehicles.

The own vehicle condition sensor 126 includes a vehicle sensor and a yaw rate sensor. The own vehicle condition sensor 126 detects the speed of the vehicle 10 (hereinafter also referred to as “vehicle speed”) and the yaw rate as the condition of the vehicle 10 .

The operation control unit 210 is composed of a central processing unit (CPU), a microcomputer composed of a RAM and a ROM, and the like, and realizes an automatic operation function by having the microcomputer execute pre-installed programs. The driving control unit 210 controls the driving force control ECU 220, the braking force control ECU 230, and the steering control ECU 240 using the road model calculated by the road model calculation unit 114, for example.

The driving force control ECU 220 is an electronic control unit that controls actuators such as the engine that generate the driving force of the vehicle. When the driver manually drives the vehicle, the driving force control ECU 220 controls the power source such as the engine or the electric motor according to the amount of operation of the accelerator pedal. On the other hand, when performing automatic operation, the driving force control ECU 220 controls the power source according to the required driving force calculated by the operation control section 210 .

The braking force control ECU 230 is an electronic control unit that controls the brake actuator that generates the braking force of the vehicle. When the driver manually drives the vehicle, the braking force control ECU 230 controls the brake actuator according to the amount of operation of the brake pedal. On the other hand, when performing automatic operation, the braking force control ECU 230 controls the brake actuators according to the required braking force calculated by the operation control section 210 .

The steering control ECU 240 is an electronic control unit that controls a motor that generates steering torque for the vehicle. When the driver manually drives the vehicle, the steering control ECU 240 controls the motor according to the operation of the steering wheel to generate an assist torque for the steering operation. As a result, the driver can operate the steering wheel with a small amount of force, thereby achieving steering of the vehicle. On the other hand, when performing automatic driving, the steering control ECU 240 performs steering by controlling the motor according to the required steering angle calculated by the driving control section 210 .

The road shape recognition process shown in FIG. 2 is a series of processes in which the road model calculation unit 114 determines the road model of the travel route of the vehicle 10 . This process is a process repeatedly executed by the road shape recognition device 110 while the vehicle 10 is running, for example, a process repeatedly executed every 100 ms.

First, the sequence of points estimation unit 112 acquires peripheral information in step S100. More specifically, peripheral information is acquired from the peripheral image of the vehicle 10 captured by the camera 122 and the peripheral situation of the vehicle 10 detected by the peripheral object sensor 124 .

Next, in step S110, the sequence of points estimating unit 112 estimates a sequence of points representing the road shape using the surrounding information acquired in step S100. More specifically, the coordinate position of the point sequence is estimated using the shape of the road division line and the shape of the roadside object detected by the peripheral sensor 120 and the movement history of the other vehicle. In this embodiment, the sequence of points estimating unit 112 estimates a sequence of points from the shape of the lane marking of the road.

Subsequently, in step S120, the recognition distance determining unit 113 determines the distance from the vehicle 10 to the furthest point in the point sequence estimated in step S110 as the recognition distance.

As shown in FIGS. 3 and 4, the sequence of points estimated by the sequence of points estimation unit 112 in step S110 (FIG. 2) is represented by x. In FIG. 3, the farthest point from the vehicle 10 in the point sequence is point P1, and the recognition distance is distance D1. As shown in FIG. 4, for example, when a part of the road division line cannot be detected due to the lane change of the preceding vehicle 20, the farthest point from the vehicle 10 in the point sequence is point P2, and the recognition distance is greater than the distance D1. The short distance is D2.

Subsequently, in step S130 (FIG. 2), the road model calculation unit 114 determines that the difference between the recognition distance determined in step S120 and the recognition distance determined in the past (hereinafter referred to as "determination distance") is a predetermined threshold distance. Determine whether or not: The "determination distance" is positive when the previously determined recognition distance is longer than the recognition distance determined in step S120, and is negative when the previously determined recognition distance is shorter than the recognition distance determined in step S120. there is The threshold distance can be determined experimentally in advance. When the determination distance is equal to or less than the threshold distance, the process proceeds to step S140, and the road model calculation unit 114 calculates a road model M1 as shown in FIG. 5 and updates the previously calculated road model. On the other hand, if the determination distance is longer than the threshold distance, that is, if the recognition distance is shorter than the recognition distance determined in the past by the threshold distance, the process proceeds to step S145 (FIG. 2), and the road model calculation unit 114 calculates The road model is held without being updated, and the process proceeds to step S150. The "road model calculated in the past" is the road model calculated in step S140 or the road model held in step S145 in the previous road shape recognition process. In other words, in the present embodiment, the road model calculator 114 retains previously calculated road models until the determination distance becomes longer than the threshold distance.

In step S150, the recognition distance determination unit 113 updates the previously determined recognition distance. More specifically, the recognition distance determining unit 113 subtracts the movement distance of the vehicle 10 from the previously determined recognition distance. The travel distance of the vehicle 10 can be obtained using, for example, the following formula (1) or (2).

D=v×CT (1)
D = v x CT + 1/2 x a x CT2 (2)
where D is the distance traveled by the vehicle 10 and v is the speed of the vehicle 10 . CT is the cycle (seconds) for repeating the road shape recognition process, and a is the acceleration of the vehicle 10 .

According to the road shape recognition device 110 of the present embodiment described above, the road model calculation unit 114 detects the previously calculated The road model calculated in the past is updated to a newly calculated road model M1 using the surrounding information. Do not update the model. Specifically, for example, if part of the road markings cannot be detected due to the movement of the preceding vehicle, and information about the distant surroundings of the vehicle cannot be obtained, the previously calculated road model will not be updated, and the road markings will not be updated. can be detected, and the road model calculated in the past is updated to the newly calculated road model only when the information on the surrounding area far from the vehicle can be acquired. Therefore, the road model can be appropriately updated.

Further, when the difference between the recognition distance and the previously determined recognition distance is longer than a predetermined threshold distance, the recognition distance determination unit 113 updates the previously determined recognition distance to a value obtained by subtracting the travel distance of the vehicle. Therefore, the relationship between the determination distance and the threshold distance can be accurately determined in the next road shape recognition process, so the road model can be updated more appropriately. Note that this process (FIG. 2, step S150) may be omitted.

B. Second embodiment:
The road shape recognition process of the second embodiment shown in FIG. 6 is different from the first embodiment in that step S145 is replaced with step S147, and other steps are the same as those of the first embodiment. Since the configuration of the automatic operation control system of the second embodiment is the same as the configuration of the automatic operation control system of the first embodiment, the explanation of the configuration of the automatic operation control system is omitted.

In the second embodiment, the road model calculator 114 corrects the previously calculated road model in step S147. More specifically, the road model calculator 114 corrects the previously calculated road model using at least the vehicle speed and the yaw rate detected by the own vehicle condition sensor 126 . In FIG. 7, the road model M2 calculated in the past is indicated by a dashed line, and the road model M3 holding the road model calculated in the past is indicated by a solid line. In FIG. 8, the road model M2 calculated in the past is indicated by a dashed line, and the corrected road model M4 is indicated by a solid line. If the road model calculated in the past is simply stored, for example, as shown in FIG. A road model M3 is represented by a function obtained by shifting the coordinates (X, Y) indicating the center of the tip of the vehicle 10 to the coordinates (X3, Y3) indicating the center of the tip of the vehicle 10 at the current position of the vehicle 10 . When correction is performed using the vehicle speed and the yaw rate, as shown in FIG. 8, a road model M4 corrected in consideration of the movement of the vehicle 10 is obtained. become. More specifically, for example, the road model calculation unit 114 uses the vehicle speed and the yaw rate to obtain the amount of parallel movement and the amount of rotational movement per cycle of road shape recognition processing of the vehicle 10, and calculates the road model M2. A road model M4 is calculated by applying parallel movement and rotational movement to .

According to the road shape recognition device 110 of the present embodiment described above, the road model calculation unit 114 determines that the difference between the recognition distance and the recognition distance determined in the past is longer than a predetermined threshold distance. The calculated road model M2 is corrected using the speed and yaw rate of the vehicle 10 detected by the own vehicle condition sensor 126 . Therefore, for example, even if part of the road division line cannot be detected due to the movement of the preceding vehicle, and peripheral information at a distance from the vehicle cannot be obtained, the road model calculated in the past can be used to reflect the movement of the vehicle 10. Since it corrects, the road model can be updated more appropriately.

C. Third embodiment:
The road shape recognition process of the third embodiment shown in FIG. 9 is characterized in that, when the determination distance is longer than the threshold distance, it is determined whether or not to hold the road model calculated in the past according to the speed of the vehicle 10. Unlike the first embodiment, other steps are the same as in the first and second embodiments. Since the configuration of the automatic operation control system of the third embodiment is the same as the configuration of the automatic operation control system of the first embodiment, the explanation of the configuration of the automatic operation control system is omitted.

In the third embodiment, when it is determined in step S130 that the determination distance is longer than the threshold distance, the road model calculator 114 determines in step S143 whether or not the vehicle speed is slower than a predetermined threshold speed. This threshold speed can be determined experimentally in advance and is a low speed (eg 5-10 km/h). If the vehicle speed is slower than the threshold speed, the road model calculator 114 proceeds to step S145 and holds the past road model. On the other hand, if the vehicle speed is equal to or higher than the threshold speed, the road model calculator 114 proceeds to step S147 and corrects the past road model as in the second embodiment.

According to the road shape recognition device 110 of the present embodiment described above, when the determination distance is determined to be equal to or less than the threshold distance and the speed of the vehicle 10 is equal to or greater than the threshold speed, the past road model is to correct. When the vehicle 10 is traveling at a low speed, the amount of change in the yaw rate is small, so it may not be possible to determine whether the peripheral sensor 120 is noise. Therefore, when the speed of the vehicle 10 is slower than the threshold speed, the past road model is held without being corrected, so that the road model can be updated more appropriately.

D. Other embodiments:
In the above-described embodiment, the road model calculation unit 114 calculates the road model using surrounding information. Alternatively, the road model calculation unit 114 may calculate the road model using not only the surrounding information but also the road model calculated by the other vehicle acquired through inter-vehicle communication with the other vehicle.

In the above-described embodiment, the sequence of points estimating unit 112 estimates the sequence of points from the shape of the division line of the road. Instead, the point sequence estimating unit 112 combines not only the shape of the road marking line, but also the shape of the road marking line and the shape of the roadside object 30 such as a wall or guardrail as shown in FIG. Alternatively, as shown in FIG. 11, the shape of the road division line and the movement history 21 of the other vehicle 20 may be combined for estimation. Since the shape of the roadside object 30 and the movement history 21 of the other vehicle 20 are easier to recognize farther than the road division line, by using them in combination with the road division line shape, a highly accurate road model can be calculated over a long distance. can. Note that the movement history 21 of the preceding vehicle 20 and the roadside object 30 tend to have a lower recognition rate than the shape of the road division line. Even if it becomes impossible to recognize the road model, the road model calculated in the past is retained and corrected, so it is possible to calculate a highly accurate road model for a long distance.

The present disclosure is not limited to the embodiments described above, and can be implemented in various configurations without departing from the scope of the present disclosure. For example, the technical features in the embodiments corresponding to the technical features in each form described in the outline of the invention are In addition, it is possible to perform replacement and combination as appropriate. Moreover, if the technical feature is not described as essential in this specification, it can be deleted as appropriate.

10 vehicle, 20 preceding vehicle, 21 movement history, 30 roadside object, 100 automatic driving control system, 110 road shape recognition device, 111 surrounding environment recognition unit, 112 point sequence estimation unit, 113 recognition distance determination unit, 114 road model calculation unit , 115 output unit, 120 peripheral sensor, 122 camera, 124 peripheral object sensor, 126 own vehicle condition sensor, 210 driving control unit, 220 driving force control ECU, 230 braking force control ECU, 240 steering control ECU, 250 in-vehicle network

Claims (4)

A road shape recognition device (110) mounted on a vehicle (10) having a surrounding sensor (120),
a peripheral environment recognition unit (111) that recognizes at least one or more of the shape of a road marking detected by the peripheral sensor, the shape of a roadside object, and the movement history of other vehicles as peripheral information;
a point sequence estimating unit (112) for estimating a point sequence representing a road shape with points using the surrounding information;
a recognition distance determination unit (113) that determines the distance from the vehicle to the farthest point from the vehicle in the sequence of points as the recognition distance;
a road model calculation unit (114) for calculating a road model representing a road shape with lines using the peripheral information;
An output unit (115) that outputs the road model,
The road model calculation unit
updating the previously calculated road model to the newly calculated road model when the difference between the recognition distance and the previously determined recognition distance is equal to or less than a predetermined threshold distance;
when the difference between the recognition distance and the recognition distance determined in the past is longer than a predetermined threshold distance , without updating the road model calculated in the past,
When the difference between the recognition distance and the previously determined recognition distance is longer than a predetermined threshold distance, the recognition distance determination unit further subtracts the movement distance of the vehicle from the previously determined recognition distance. a road shape recognition device that updates the recognition distance to . A road shape recognition device mounted on a vehicle having a surrounding sensor and a vehicle condition sensor (126),
a peripheral environment recognition unit that recognizes at least one or more of the shape of road markings detected by the peripheral sensor, the shape of roadside objects, and the movement history of other vehicles as peripheral information;
a point sequence estimating unit for estimating a point sequence representing a road shape with points using the surrounding information;
a recognition distance determination unit that determines a distance from the vehicle to a farthest point from the vehicle in the sequence of points as a recognition distance;
a road model calculation unit that calculates a road model representing a road shape with a line using the surrounding information;
an output unit that outputs the road model,
The road model calculation unit
when the difference between the recognition distance and the recognition distance determined in the past is equal to or less than a predetermined threshold distance, updating the previously calculated road model to the road model;
A road model calculated in the past using at least the speed and yaw rate of the vehicle detected by the vehicle condition sensor when the difference between the recognition distance and the recognition distance determined in the past is longer than a predetermined threshold distance. A road shape recognition device that corrects the A road shape recognition device mounted on a vehicle having a surrounding sensor and a vehicle condition sensor,
a peripheral environment recognition unit that recognizes at least one or more of the shape of road markings detected by the peripheral sensor, the shape of roadside objects, and the movement history of other vehicles as peripheral information;
a point sequence estimating unit for estimating a point sequence representing a road shape with points using the surrounding information;
a recognition distance determination unit that determines a distance from the vehicle to a farthest point from the vehicle in the sequence of points as a recognition distance;
a road model calculation unit that calculates a road model representing a road shape with a line using the surrounding information;
an output unit that outputs the road model,
The road model calculation unit
when the difference between the recognition distance and the recognition distance determined in the past is equal to or less than a predetermined threshold distance, updating the previously calculated road model to the road model;
When the difference between the recognition distance and the recognition distance determined in the past is longer than a predetermined threshold distance, and when the speed of the vehicle detected by the vehicle condition sensor is lower than the predetermined speed , without updating the road model calculated in the past,
When the difference between the recognition distance and the recognition distance determined in the past is longer than a predetermined threshold distance, and when the speed of the vehicle detected by the vehicle condition sensor is equal to or higher than the predetermined speed 7. A road shape recognition device that corrects a previously calculated road model using at least the speed and yaw rate of the vehicle detected by the vehicle condition sensor. The road shape recognition device according to claim 2 or 3 ,
When the difference between the recognition distance and the previously determined recognition distance is longer than a predetermined threshold distance, the recognition distance determination unit further subtracts the movement distance of the vehicle from the previously determined recognition distance. a road shape recognition device that updates the recognition distance to .

Images