TWI704437B - Self-driving vehicle - Google Patents

Abstract

The present invention provides an automatic driving vehicle that can accurately identify the current position and the driving path ahead and automatically travel along a predetermined driving path.

The autonomous vehicle is configured to be able to travel autonomously along a predetermined travel path, and includes: a memory unit that memorizes trajectory-related information related to the travel trajectory of an autonomous vehicle that has traveled along the predetermined travel path in advance, and plural numbers on the predetermined travel path Distance-related information about the distance traveled from the starting point of the predetermined driving route at each measurement point; the travel distance measurement unit, which measures information about the travel distance from the start point to the current location; and the travel area estimation unit, which will measure The derived distance-related information is compared with the distance-related information to grasp the location of the current location, and based on the trajectory-related information, the travel area forward from the current location is estimated by calculation.

Description

Self-driving vehicle

The present invention relates to an autonomous vehicle configured to automatically travel along a predetermined travel path.

Previously, autonomous vehicles have been developed that detect electromagnetic induction lines embedded in the driving path by sensors and drive automatically along the induction lines. Such an autonomous vehicle is used as, for example, a golf cart that carries items such as a golf course or a player and drives (for example, refer to Patent Document 1 below). Furthermore, golf carts are also called "golf carts."

In addition, as a vehicle using electromagnetic induction wires, unmanned vehicles used in orchards and the like are proposed. For example, Patent Document 2 below discloses an unmanned vehicle that automatically travels on an induction line buried in the ground along a travel path between trees in an orchard. The vehicle is equipped with obstacle sensors on the front. Moreover, the vehicle is equipped with the function of detecting obstacles by using the sensor, and stopping when the distance to the detected obstacle is less than fixed.

[Prior Technical Literature] [Patent Literature]

[Patent Document 1] Japanese Patent Laid-Open No. 2000-181540

[Patent Document 2] Patent No. 2944814

In addition, there may be obstacles to the golf cart such as players or golf clubs in the area where the golf cart is scheduled to travel. Therefore, it is considered that the golf cart disclosed in Patent Document 1 that runs on a golf course is equipped with an obstacle sensor of the unmanned vehicle disclosed in Patent Document 2 used in orchards and the like.

However, the golf cart is automatically driven on a driving path set in accordance with the terrain of the golf course. Therefore, golf carts may also travel with a smaller radius of rotation in places with trees. For example, there is a situation in which a golf cart approaches a tree in the front while traveling along a straight portion in front of a portion of a travel path that revolves with a smaller radius of rotation. In the situation as described above, if the technique of Patent Document 2 described above is used to detect obstacles, trees existing in front of the vehicle and close to the vehicle are detected as obstacles. As a result, the vehicle may stop even though there is no obstacle during driving.

Therefore, it is not easy for the golf cart disclosed in Patent Document 1 to drive on a golf course to be equipped with the obstacle sensor of the unmanned vehicle disclosed in Patent Document 2 used in orchards and the like.

Therefore, the inventors of this case have conducted diligent research on the conditions necessary for the installation of obstacle sensors. As described above, a vehicle like a golf cart turns with a small radius of rotation when traveling along a predetermined travel path. Therefore, if the vehicle is equipped with a function of judging whether there is an obstacle on a predetermined travel path, it can perform stop control only when there is an obstacle on the travel path. As a result, it is possible to suppress unnecessary automatic stopping of the vehicle when there is no obstacle during driving. The inventor of the present case found that in order to achieve this, the vehicle only needs to grasp the current position of the vehicle and the driving area that is more forward than the current position. In addition, it is found that if the vehicle knows the current position of the vehicle and the driving area farther ahead than the current position, the above information can be used not only for obstacle detection, but also for vehicle information Speed control and other vehicle driving control.

The purpose of the present invention is to provide a method that can accurately identify the current position and its front An autonomous vehicle in the driving area that is automatically driving along a predetermined driving path.

The present invention is an automatic driving vehicle characterized in that it is configured to automatically travel along a predetermined driving path, and is provided with: a memory unit that memorizes trajectory-related information related to the driving trajectory of an automatic driving vehicle that has traveled along the predetermined driving path in advance , And the distance-related information related to the driving distance from the starting point of the above-mentioned predetermined driving path at a plurality of measurement points on the above-mentioned predetermined driving path; the driving distance measurement unit, which measures the distance-related information from the above-mentioned starting point to the current location Information; and a travel area estimation unit, which compares the travel distance-related information measured by the travel distance measurement unit with the distance-related information read from the memory unit, and detects the location of the current location based on The trajectory-related information read from the memory unit is used to calculate the travel area forward from the current location.

It is assumed that the above-mentioned autonomous vehicle automatically travels along a predetermined travel path prescribed in advance. Therefore, the information about the trajectory (trajectory-related information) obtained when the autonomous vehicle travels on the predetermined travel path in advance is substantially identical to the trajectory of the vehicle when it is traveling. Similarly, the information (distance-related information) related to the distance traveled from the starting point to the plurality of measurement points on the predetermined travel path when the autonomous vehicle travels on the predetermined travel path in advance is related to the self-driving vehicle. The actual distance traveled from the starting point to each measuring point is almost the same.

Therefore, in the above-mentioned autonomous vehicle, by comparing the actual travel distance measured by the travel distance measurement unit with the travel distance recorded in the distance-related information read from the memory unit, the position of the vehicle at the current location can be grasped . Therefore, based on the travel trajectory recorded in the trajectory-related information read from the memory unit, it is possible to estimate the travel area that the autonomous vehicle will pass through when it is advancing from the current location.

Here, as the vehicle used to obtain trajectory-related information and distance-related information, it may be the same vehicle as the autonomous vehicle targeted for the estimated travel area, or other autonomous vehicles as long as the shape and size are the same.

The travel distance measurement unit may include, for example, a rotation angle sensor mounted on a wheel. The travel distance can be measured by using a rotation angle sensor to measure the rotation angle of the wheel from the starting point to the measuring point, and multiply it by the diameter of the wheel. Furthermore, the above product can be multiplied by a specific error coefficient as needed. In addition, information related to the rotation angle not related to the travel distance may be recorded in the distance-related information.

As an autonomous vehicle equipped with a function for acquiring distance-related information, for example, the following configuration can be adopted. That is, the vehicle may be the following, which includes: an imaging unit that photographs a specific direction as viewed from the autonomous vehicle at the plurality of measurement points while traveling along the predetermined travel path; and The derivation unit is configured to derive a traveling trajectory based on a plurality of camera data captured by the imaging unit using a visual ranging method; and the trajectory related information includes information based on the traveling trajectory derived by the trajectory derivation unit.

According to the method of visual distance measurement, the trajectory of the moving body can be derived by detecting the displacement of the coordinate of the feature point on each image based on the continuous plural image information taken by one side moving and the other side. As a result, it is possible to create trajectory-related information related to the driving trajectory on the predetermined driving path by driving the autonomous vehicle having the above-mentioned configuration on the predetermined driving path in advance.

The autonomous vehicle may also include a parallax image creating unit that creates a parallax image based on the camera data, and the driving area estimating unit includes a function of identifying an area on the parallax image corresponding to the derived travel trajectory, and The trajectory related information includes the above-mentioned parallax image of each of the above-mentioned plural measurement points It is composed of information related to the area corresponding to the above-mentioned driving track.

According to the above-mentioned structure, the information indicating the area corresponding to the trajectory of the autonomous vehicle traveling in advance on the parallax image of each of the plurality of measurement points is described as trajectory-related information. Thereby, it is possible to record information related to the area that the autonomous vehicle will pass through when traveling on a predetermined driving path with a smaller amount of information.

More specifically, the trajectory-related information may also include information obtained by establishing a correspondence relationship between the range of the coordinates of the area corresponding to the driving trajectory on the parallax image of each of the plurality of measurement points according to the parallax value.

Furthermore, the autonomous vehicle may include: an obstacle detection unit that detects an obstacle existing in front of the autonomous vehicle; and a determination unit that determines whether the detected obstacle is present in the travel area In the above-mentioned travel area estimated by the estimation unit.

According to the above configuration, the autonomous vehicle can estimate the travel area from the current location forward by the travel area estimation unit, and therefore, the determination unit can determine whether an obstacle exists in the travel area. Thereby, by including, for example, a control unit that controls the automatic driving vehicle to stop only when an obstacle is determined by the determination unit to be present in the driving area, it is possible to suppress unnecessary automatic stop when there is no obstacle during driving. The launch.

The autonomous vehicle may be configured to be capable of autonomously traveling on an electromagnetic induction line embedded in the predetermined traveling path.

The plurality of measurement points may include an installation position of a fixed-point member that is embedded in the predetermined travel path and can be read by a sensor mounted on the autonomous vehicle.

The aforementioned autonomous vehicle can be used as, for example, a golf cart.

According to the present invention, the self-driving vehicle that automatically travels along the predetermined travel path can accurately recognize the current position and the travel area in front of it.

1‧‧‧Autonomous vehicles

3‧‧‧Camera Department

3a‧‧‧Left Image Sensor

3b‧‧‧Right image sensor

4‧‧‧Steering wheel

5‧‧‧Right front wheel

6‧‧‧Left front wheel

7‧‧‧Reading section

7a‧‧‧Fixed point sensor

7b‧‧‧Induction line sensor

9‧‧‧Rotation Angle Sensor

11‧‧‧Automatic operation control department

13‧‧‧Travelling distance measurement department

15‧‧‧Driving Area Estimation Department

17‧‧‧Memory Department

19‧‧‧Trajectory export section

21‧‧‧Driving path

23‧‧‧Pointed component

24‧‧‧Electromagnetic induction wire

31‧‧‧Parallax Image Production Department

33‧‧‧Obstacle Detection Department

35‧‧‧Judgment Department

50‧‧‧Parallax image

50a‧‧‧ Parallax image

51‧‧‧ Obstacles on parallax images

52‧‧‧ Obstacles on parallax images

53‧‧‧ Obstacles on parallax images

54‧‧‧ Obstacles on parallax images

C0‧‧‧Starting point

C1‧‧‧Location

C2‧‧‧Location

C3‧‧‧Location

d1~d7‧‧‧disparity value

X1~X10‧‧‧X coordinate area

Figure 1 is a schematic diagram of an autonomous vehicle when viewed from the front.

Fig. 2 is a block diagram functionally showing the configuration of the first embodiment of the autonomous vehicle.

Fig. 3 is a schematic diagram showing an example of a traveling path of an automatic traveling vehicle.

Figure 4 (a) ~ (d) are diagrams used to illustrate the information related to the trajectory.

Fig. 5 is a block diagram functionally showing the structure of the second embodiment of the autonomous vehicle.

Fig. 6 is a diagram showing an example of a parallax image.

Fig. 7 is a block diagram functionally showing the configuration of the third embodiment of the autonomous vehicle.

FIG. 8 is a diagram showing an example of a parallax image when there is an obstacle on the traveling path.

[First Embodiment]

The structure of the first embodiment of the autonomous vehicle of the present invention will be described with reference to the drawings. Furthermore, in the following drawings, the actual size ratio and the size ratio on the drawing may not be the same.

In this embodiment, a golf cart is exemplified as an autonomous vehicle for description. However, as an autonomous vehicle, it is not limited to golf carts, but also includes unmanned transport vehicles that run in factories or orchards. In addition, the autonomous vehicle in the present invention is not limited to a four-wheeled vehicle, and may be a tricycle or a monorail type. The same applies to the content after the second embodiment described below.

(The composition of the vehicle)

Fig. 1 is a schematic diagram of the automatic driving vehicle in this embodiment when viewed from the front. The autonomous vehicle 1 shown in FIG. 1 is a golf cart that automatically drives in a golf course. again In addition, FIG. 2 is a block diagram functionally showing the structure of the autonomous vehicle 1.

The autonomous vehicle 1 shown in FIG. 1 is equipped with an imaging unit 3 at the center of the front. The imaging unit 3 includes, for example, a stereo camera, and includes a left image sensor 3a and a right image sensor 3b. The image sensors (3a, 3b) include general visible light sensors such as CCD (Charge-Coupled Device) or CMOS (Complementary MOS (Metal Oxide Semiconductor)). In addition, in this manual, descriptions such as "front and rear" or "left and right" are based on the direction in which the autonomous vehicle 1 advances.

The autonomous vehicle 1 includes a steering wheel 4 and a right front wheel 5 and a left front wheel 6 that are steered by the rotation of the steering wheel 4. In addition, the autonomous vehicle 1 is equipped with a reading unit 7 in the lower part of the vehicle body. The reading unit 7 includes a fixed-point sensor 7a and an induction line sensor 7b (refer to FIG. 2).

The right front wheel 5 of the autonomous vehicle 1 is provided with a rotation angle sensor 9 for detecting the rotation angle of the right front wheel 5. The rotation angle sensor 9 detects the rotation angle of the wheel, and includes, for example, a rotary encoder. Furthermore, the rotation angle sensor 9 can also replace the right front wheel 5 or be provided on the left front wheel 6 or the rear wheel in addition to the right front wheel 5.

FIG. 2 is a functional block diagram showing the structure of the autonomous vehicle 1. The automatic driving vehicle 1 includes an automatic operation control unit 11, a travel distance measurement unit 13, a travel area estimation unit 15, a storage unit 17, and a trajectory derivation unit 19. The travel distance measurement unit 13, the travel area estimation unit 15, and the trajectory derivation unit 19 include, for example, arithmetic devices such as a CPU (Central Processing Unit). In addition, the storage unit 17 includes, for example, a memory or a hard disk.

The automatic operation control unit 11 controls the automatic operation of the automatic driving vehicle 1 for automatic operation along an electromagnetic induction line provided on a predetermined traveling path. FIG. 3 is an example of a travel route scheduled for the autonomous vehicle 1 to travel. As shown in FIG. 3, an electromagnetic induction wire 24 is embedded in the travel path 21. The induction line sensor 7b receives the electromagnetic wave emitted from the electromagnetic induction line 24 and outputs a detection signal to the automatic operation control unit 11. The automatic operation control unit 11 controls a steering mechanism (not shown) based on the detection signal. With this, the vehicle 1 is automatically operated at It runs automatically on the driving route 21.

In addition, as shown in FIG. 3, a fixed point member 23 is embedded in a plurality of predetermined positions including the starting point C0 on the travel path 21. The pointing member 23 is constituted by, for example, a combination of a plurality of magnets. The fixed-point sensor 7a can read the magnetic field information from the fixed-point member 23, and includes, for example, a magnetic sensor. These fixed-point members 23 send instruction signals for instructing running, stopping, decelerating, etc., for example. When the autonomous vehicle 1 passes on the fixed-point member 23, the fixed-point sensor 7a receives the instruction signal from the passing fixed-point member 23, and outputs the instruction signal to the automatic operation control unit 11. The automatic operation control unit 11 controls the automatic driving vehicle 1 based on the instruction signal. Thereby, the autonomous vehicle 1 automatically performs control such as running, stopping, decelerating, etc. based on the information designated by the pointing member 23.

In addition, the fixed-point sensor 7a outputs the information of the content to the travel distance measurement unit 13 at the time when the autonomous vehicle 1 passes the fixed-point member 23. The travel distance measuring unit 13 measures the distance traveled after passing the fixed point member 23 based on the information related to the rotation angle of the wheel output from the spin angle sensor 9 based on the time point when the self-fixed point sensor 7a passes through the fixed point member 23 . The travel distance measuring unit 13 can memorize information related to the diameter of the right front wheel 5 in advance. In this way, based on the rotation angle (number of revolutions) of the right front wheel 5 from a specific time point and the diameter of the right front wheel 5, the travel distance of the autonomous vehicle 1 from the above specific time point can be calculated by calculation .

Therefore, by using the time point at which the starting point C0 is passed as a reference, the traveling distance measuring unit 13 can measure the traveling distance from the starting point C0 to the current location.

The following track-related information and distance-related information are stored in the memory unit 17. The information is generated and memorized in the memory unit 17 when the autonomous vehicle 1 travels on the travel path 21 in advance. The traveling area estimating unit 15 has a function of automatically detecting the traveling distance from the starting point C0 to the current location measured by the traveling distance measuring unit 13 based on the above-mentioned information stored in the memory unit 17 in advance. The current location of the driving vehicle 1. Furthermore, the travel area estimating unit 15 has an estimate of travel forward from the current location The function of the area. The various information stored in the storage unit 17 and the specific calculation content in the travel area estimation unit 15 will be described below.

(Various information memorized in memory section 17)

As described above, the trajectory-related information and distance-related information are stored in the storage unit 17 in advance. The trajectory-related information is information created by the trajectory derivation unit 19 when the autonomous vehicle 1 travels on the travel route 21 in advance.

When creating trajectory-related information, first, while the autonomous vehicle 1 is traveling on the travel path 21, the camera unit 3 continuously photographs the front of the autonomous vehicle 1 at a specific frame rate. In this way, the plural points (C0, C1, C2,...) captured by the imaging unit 3 correspond to "a plurality of measurement points".

Next, the trajectory derivation unit 19 specifies the position of the autonomous vehicle 1 and the direction of the vehicle body based on the continuous images. FIG. 4 schematically shows four consecutive photos taken by the imaging unit 3, and the position and direction of the imaging unit 3 at the time when each photo was taken. In addition, in the following description, the position of FIG. 4(a) is assumed to be the starting point C0 in FIG. 3.

The trajectory derivation unit 19 estimates the position of the autonomous vehicle 1 and the direction of the vehicle body based on the imaging data captured by the imaging unit 3 at a specific frame rate. As this estimation method, a method such as visual odometry can be used. As a specific example, the trajectory derivation unit 19 extracts a plurality of feature points on the imaging data and detects the displacement of each feature point on two consecutive imaging data. With this, the amount of change in the position and direction of the autonomous vehicle 1 between the two camera data is calculated.

i，ψi)。軌跡導出部19係以此方式遍及整個行駛路徑21製作自動行駛車輛1之行駛軌跡，並將其記憶於記憶部17。該資訊與「軌跡相關資訊」對應。 Then, taking the starting point C0 as the origin, the calculated changes are sequentially accumulated from the starting point C0, thereby, as shown in Fig. 4, a driving trajectory composed of a total of 6 components of position and direction of the autonomous vehicle 1 is obtained (xi, yi, zi, θi,

i, ψi). In this way, the trajectory derivation unit 19 creates the traveling trajectory of the autonomous vehicle 1 over the entire traveling path 21 and stores it in the memory unit 17. This information corresponds to "track related information".

In addition, the trajectory derivation unit 19 captures images before the autonomous vehicle 1 by the imaging unit 3 The coordinate information of the 6-axis of the autonomous vehicle 1 at the time of the party and the information related to the travel distance of the autonomous vehicle 1 from the starting point C0 to each point are associated and stored in the memory unit 17. The information obtained by associating the coordinates with the driving distance corresponds to the "distance related information". Furthermore, the distance-related information can be information about the driving distance of the autonomous vehicle 1 from the starting point C0, or information related to the rotation angle of the right wheel 5 of the autonomous vehicle 1 from the starting point C0. It can also be information obtained by multiplying the equivalent value by a specific coefficient such as error.

(Contents of processing by the driving area estimation unit 15)

In the memory portion 17 of the autonomous vehicle 1, as described above, the trajectory-related information and distance-related information obtained by the autonomous vehicle 1 traveling on the driving path 21 are stored in advance. The travel area estimating unit 15 is provided with information about the travel distance from the starting point C0 to the current location from the travel distance measuring unit 13 while the autonomous vehicle 1 is traveling on the travel route 21. The travel area estimation unit 15 compares the information related to the travel distance with the distance-related information read from the memory unit 17 to detect the coordinates of the current location of the autonomous vehicle 1. Furthermore, the travel area estimating unit 15 reads the trajectory-related information from the memory unit 17, and estimates the travel area of the autonomous vehicle 1 forward from the current location detected just now.

In this way, the autonomous vehicle 1 can recognize which route it is scheduled to travel on during the process of autonomously traveling on the travel route 21. Thus, for example, by outputting information related to the travel area estimated by the travel distance estimating unit 15 to the automatic operation control unit 11, the information can be used for speed control or travel control of the automatic driving vehicle 1, thereby contributing Improve the safety of automatic operation. In FIG. 2, the content of the information output by the self-driving area estimation unit 15 to the automatic operation control unit 11 is indicated by a broken line with an arrow. However, in the present embodiment, the automatic operation control unit 11 performs automatic operation control based on the information about the travel area estimated by the travel distance estimating unit 15 arbitrarily, and this function need not be provided.

In addition, when the autonomous vehicle 1 has an obstacle detection function, by The information related to the travel area estimated by the travel distance estimating unit 15 is used for the obstacle detection, and the obstacle detection with high accuracy corresponding to the characteristics of the travel path 21 can be performed. This content will be described in the third embodiment below.

(Another configuration of the first embodiment)

Information about the distance along the travel path 21 from the starting point C0 to each of the fixed-point members 23 may also be stored in the memory portion 17. The travel distance measurement unit 13 outputs the travel distance calculated from the starting point C0 at that time to the travel area estimation unit 15 when the fixed-point sensor 7a detects that the autonomous vehicle 1 has passed the fixed-point member 23. The travel area estimating unit 15 reads out the information related to the travel distance from the starting point C0 to each fixed-point member 23 from the memory unit 17, and compares it with the information related to the travel distance output from the travel distance measurement unit 13, and specifies the travel distance The value of the nearest fixed-point component 23. Then, the travel area estimating unit 15 reads out the trajectory related information on the position of the specified pointing member 23 from the memory unit 17, and estimates the travel area of the autonomous vehicle 1 forward from the specified pointing member 23.

Furthermore, the travel area estimating unit 15 replaces the distance measured by the travel distance measuring unit 13 as the distance traveled by the autonomous vehicle 1 from the starting point C0 to passing the specified fixed-point member 23 with the distance read from the memory unit 17 The distance. In FIG. 2, the content of the information output from the traveling area estimating unit 15 to the traveling distance measuring unit 13 is indicated by a broken line with an arrow. Thereby, the traveling distance measuring unit 13 can eliminate the measurement error from the starting point C0 to passing the fixed point member 23. That is, according to this configuration, the measurement error of the travel distance measurement unit 13 can be eliminated every time the fixed-point member 23 passes, and therefore, the measurement accuracy of the travel distance by the travel distance measurement unit 13 can be improved.

However, in this embodiment, it is arbitrary to adjust the value of the traveling distance measured by the traveling distance measuring part 13 every time the fixed-point member 23 passes, and this function does not need to be provided.

[Second Embodiment]

The configuration of the second embodiment of the autonomous vehicle will be described with reference to the drawings. In addition, in the following embodiments, the same reference numerals are given to the components common to the first embodiment, and the description is omitted as appropriate.

FIG. 5 is a block diagram functionally showing the structure of the autonomous vehicle 1 in this embodiment. The autonomous vehicle 1 of the present embodiment is different in that it includes a parallax image creation unit 31 in addition to the configuration of the first embodiment.

In this embodiment, the image sensors (3a, 3b) included in the imaging unit 3 are installed at a fixed distance in the horizontal direction. That is, the left image sensor 3a and the right image sensor 3b are respectively arranged in a parallel three-dimensional positional relationship. The left image sensor 3a and the right image sensor 3b are arranged in such a way that the positions of the respective rows of the images respectively captured are the same, that is, arranged in a way that the epipolar lines are identical.

Furthermore, in the following, the direction in which the left image sensor 3a and the right image sensor 3b are connected, that is, the left and right direction is set as the X axis, and the direction orthogonal to the plane of the travel path 21, that is, the vertical direction is set Is the Y axis. In addition, the front and rear direction of the autonomous vehicle 1 is the Z axis.

As described in the first embodiment above, when creating trajectory-related information, in advance, the autonomous vehicle 1 is driven on the travel path 21 while the camera unit 3 continuously controls the autonomous vehicle 1 at a specific frame rate. Before shooting. The images captured at this time, more specifically, the images captured by each of the left image sensor 3a and the right image sensor 3b are temporarily stored in a buffer not shown. The stored image can also be one that appropriately corrects lens distortion, focus deviation, etc.

The parallax image creation unit 31 creates a parallax image based on the stored image data. An example of a method of creating a parallax image will be described. The image data obtained from the left image sensor 3a is set as a reference image, and the image data obtained from the right image sensor 3b is set as a reference image. Then, a pixel of the reference image is set as a pixel of interest, and a pixel on the reference image corresponding to the pixel of interest (hereinafter referred to as "corresponding pixel") is searched.

In this search, methods such as stereo matching can be used. As a stereo match, there are regions Basic matching or feature-based matching, etc. For example, in the case of area-based matching, set an area centered on the pixel of interest (hereinafter referred to as "reference area"). The reference area is compared with the reference image, and the area on the reference image that is most similar to the reference area is identified. Then, the pixel located at the center of the area on the specified reference image is determined as the corresponding pixel.

When searching for corresponding pixels, calculate the lateral (X-axis direction) offset between the noticeable pixel on the reference image and the corresponding pixel on the reference image. This offset is equivalent to the parallax in the pixel of interest.

After obtaining the parallax for one pixel, continue to reset other pixels on the reference image as the notable pixels in the same way, and repeat the same processing. In this way, the parallax is also calculated for other pixels. A parallax image can be created by establishing a corresponding relationship between the obtained parallax and each pixel. The created parallax image is stored in the storage unit 17. An example of parallax images is shown in FIG. 6.

FIG. 6 shows a parallax image 50 of a certain point Ci on the travel route 21. The parallax image 50 shown in FIG. 6 is composed of regions representing 7 types of parallax values (d1 to d7). The relationship between the disparity values d1~d7 is d1>d2>d3>d4>d5>d6>d7.

As described above in the first embodiment, the trajectory derivation unit 19 creates the travel trajectory of the autonomous vehicle 1 covering the entire travel route 21 and stores it in the memory unit 17. In the autonomous vehicle 1 of the present embodiment, the travel area estimating unit 15 reads out the created information about the travel track and the parallax image from the memory unit 17, and obtains the area of the travel path 21 on the parallax image. Specifically, the X coordinate of the travel path on the parallax image is associated with the parallax value. For example, in the case of the parallax image 50 shown in FIG. 6, the driving path 21 is specified as the X coordinate area X1~X10 when the parallax value d1, the X coordinate area X2~X9 when the parallax value d2, and the parallax The X-coordinate area X3~X8 at the value d3, the X-coordinate area X4~X7 at the parallax value d4, and the X-coordinate area X5~X6 at the parallax value d5.

The driving area estimating unit 15 is based on the parallax images of each point stored in the memory unit 17 In the same way, the area of the driving route 21 on the parallax image is specified. In addition, the information of the area of the travel route 21 on the parallax image specifying each point on the travel route 21 is added to the trajectory-related information and stored in the memory unit 17.

According to the configuration of the present embodiment, the information of the trajectory that the autonomous vehicle 1 travels along the travel route 21 can be stored in the memory unit 17 as a state added to the parallax image. Therefore, according to the autonomous vehicle 1 of this embodiment, in addition to the functions described in the first embodiment above, for example, the following function can be added, that is, the parallax map created by the parallax image creation unit 31 is used during driving. The image is compared with the trajectory-related information memorized in the memory portion 17 to quickly detect that the autonomous vehicle 1 has deviated from the driving path 21.

[Third Embodiment]

The configuration of the third embodiment of the autonomous vehicle will be described with reference to the drawings. FIG. 7 is a block diagram functionally showing the structure of the autonomous vehicle 1 in this embodiment. The autonomous vehicle 1 of this embodiment is different in that it includes an obstacle detection unit 33 and a determination unit 35 in addition to the configuration of the second embodiment.

During the driving of the autonomous vehicle 1 on the driving path 21, the imaging unit 3 photographs the front according to a specific time sequence, and the parallax image creation unit 31 creates a parallax image based on the imaging data and outputs it to the obstacle detection unit 33 . The obstacle detection unit 33 determines whether there is an obstacle on the sent parallax image, and captures the area on the parallax image of the obstacle when there is an obstacle. As an example, the obstacle detection unit 33 detects the following area in the parallax image as an obstacle, where different parallax values in the X direction of the area are close, and areas with the same parallax value have a predetermined value or more The number of pixels in the Y direction, in other words, the parallax value is the same and has a predetermined height. For example, in the case of the parallax image 50 shown in FIG. 6, the obstacle detection unit 33 detects the areas 51, 52, and 53 as obstacles.

The determination unit 35 determines whether the detected obstacle exists on the travel path 21. As an example, the determination unit 35 reads out the track-related information from the storage unit 17 and detects the area of the X coordinate of the travel route 21 on the parallax image of the current location. Then, the judging section 35 only needs to be disabled The range of the X coordinate value at the lower end of the obstacle area output by the obstacle detection unit 33 is included in the range of the X coordinate of each parallax of the travel path 21, and it is determined that the obstacle exists on the travel path 21. For example, in the case of the parallax image 50 shown in FIG. 6, the determination unit 35 determines that there is no obstacle on the travel path 21.

On the other hand, the case where the parallax image output from the parallax image creation unit 31 to the obstacle detection unit 33 is an image as shown in FIG. 8 will be discussed. The parallax image 50a shown in FIG. 8 was created based on the data taken at the same place as the parallax image 50 shown in FIG. 6, and it is assumed that the parallax image 50a was created in the autonomous vehicle 1 at the time when the parallax image 50a was created. The scene with people was reflected before.

The obstacle detection unit 33 uses the same method as described above to detect the areas 51, 52, 53, and 54 as obstacles based on the information of the parallax image 50a. The determining unit 35 detects that the range of the X coordinate value at the lower end of the area 54 is within the range of X4 or more and X8 or less, and represents an area on the travel path 21 with a parallax value d3. As a result, the determination unit 35 determines that there is an obstacle on the travel route 21 at the current time.

If the determination unit 35 determines that an obstacle is present on the travel route 21, it outputs information of the content to the automatic operation control unit 11. The automatic operation control unit 11 performs deceleration or stop control of the automatic driving vehicle 1 accordingly.

According to the autonomous vehicle 1 of the present embodiment, the information related to the area of the travel path 21 on the parallax image is stored in the storage unit 17 in advance. Therefore, when an obstacle is detected, it can be determined whether the obstacle exists Travel path 21. Thereby, it may be configured such that even when an obstacle is detected, when the obstacle does not exist on the travel path 21, the automatic operation control unit 11 does not perform deceleration or stop control. As a result, according to the self-driving vehicle 1 of the present embodiment, it is possible to suppress unnecessary automatic stopping of the vehicle when there is no obstacle during driving.

Furthermore, the determination unit 35 may also determine whether the rectangular area surrounding the obstacle is the same as the area of the travel path 21 after the obstacle detection unit 33 detects the presence of an obstacle. After that, only the obstacles having a rectangular area overlapping with the area of the travel path 21 are determined as described above. With this two-stage determination, it is possible to determine whether an obstacle exists on the travel path at a higher speed. Furthermore, when the judgment unit 35 is a case where the Y coordinate value (height) of the lower end of the rectangle is greater than the predetermined height, it is judged to be a part of a tree such as a bridge or branch erected above the travel path 21 instead of an obstacle. This can also improve the accuracy of obstacle determination.

[Another embodiment]

Hereinafter, another embodiment will be described.

<1> In each of the above embodiments, the distance-related information and trajectory-related information previously memorized by the memory unit 17 are set to be generated by the autonomous vehicle 1 provided with the memory unit 17 running on the driving path 21. The description. However, the above-mentioned information stored in the memory unit 17 may also be caused by the autonomous vehicle 1 having substantially the same shape and substantially the same size, which is different from the autonomous vehicle 1 equipped with the memory unit 17 and travels on the travel path 21 in advance. Producer. In this case, in the autonomous vehicle 1 of the first embodiment, the imaging unit 3 and the trajectory deriving unit 19 are not necessary. Similarly, in the autonomous vehicle 1 of the second embodiment, the imaging unit 3, the trajectory deriving unit 19, and the parallax image creating unit 31 are not necessary.

<2> The invention of the present application and the automatic driving vehicle (automatically driven vehicle) of this specification is a vehicle capable of driving automatically. An autonomous vehicle is a vehicle that can travel automatically without being steered by an operator. An autonomous vehicle is a vehicle that can travel automatically without being accelerated and decelerated by the operator. In addition, an autonomous vehicle includes an autonomously driven vehicle equipped with at least one sensor and capable of autonomously driving according to a signal from the sensor.

1‧‧‧Autonomous vehicles

3‧‧‧Camera Department

7‧‧‧Reading section

7a‧‧‧Fixed point sensor

7b‧‧‧Induction line sensor

9‧‧‧Rotation Angle Sensor

11‧‧‧Automatic operation control department

13‧‧‧Travelling distance measurement department

15‧‧‧Driving Area Estimation Department

17‧‧‧Memory Department

19‧‧‧Trajectory export section

Claims (5)

An autonomous vehicle characterized in that it is configured to automatically travel along a predetermined travel path, and is equipped with a memory unit that memorizes trajectory-related information related to the travel trajectory of an autonomous vehicle that has traveled along the predetermined travel path in advance, and the foregoing Distance-related information about the driving distance from the starting point of the above-mentioned predetermined driving path at a plurality of measurement points on the predetermined driving path; and the driving distance measurement unit, which measures information about the driving distance from the above-mentioned starting point to the current location; and A travel area estimating unit that compares the information related to the travel distance measured by the travel distance measurement unit with the distance-related information read from the memory unit to grasp the location of the current location, and based on the memory The trajectory-related information read by the unit reads out and calculates to estimate the travel area from the current location forward; the camera unit, in the process of traveling along the predetermined travel path, compares the automatic travel from the plurality of measurement points to the Vehicle observation is shooting in a specific direction; and a trajectory derivation unit, which derives the driving trajectory based on a plurality of camera data captured by the above-mentioned camera unit, using the method of visual ranging; and the above-mentioned trajectory-related information includes derived based on the use of the above-mentioned trajectory The above-mentioned automatic driving vehicle can be automatically driven on the electromagnetic induction line embedded in the predetermined driving path; the plurality of measurement points include embedded in the predetermined driving path and mounted on the automatic driving vehicle The position of the fixed-point component that can be read by the sensor. For example, the automatic driving vehicle of claim 1, which includes a parallax image creation unit that creates a parallax image based on the above-mentioned camera data, and the travel area estimating unit includes a device that specifies an area on the parallax image corresponding to the derived travel trajectory Function, and the trajectory-related information includes information about the area corresponding to the travel trajectory on the parallax image of each of the plurality of measurement points. For example, the automatic driving vehicle of claim 2, wherein the trajectory-related information includes the coordinate range of the region corresponding to the driving trajectory on the parallax image of each of the plurality of measurement points, and the corresponding relationship is established according to the parallax value Information. For example, the automatic driving vehicle of any one of claims 1 to 3, further comprising: an obstacle detection unit that detects obstacles in front of the automatic driving vehicle; and a determination unit that determines the detected obstacles Whether an obstacle exists in the travel area estimated by the travel area estimating unit. Such as the automatic driving vehicle of any one of claims 1 to 3, wherein the automatic driving vehicle is a golf cart.

Images