JP3036301B2 - Obstacle detection device for unmanned vehicles - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an obstacle detection device for an unmanned vehicle, and more particularly to an obstacle detection system for an unmanned vehicle which is suitable for identifying whether or not an object existing around the unmanned vehicle interferes with traveling. The present invention relates to an object detection device.

[0002]

2. Description of the Related Art One type of autonomous unmanned vehicle used for various kinds of work is, for example, an unmanned vehicle in which a predetermined work is performed by running on rough terrain outdoors. With this type of conventional autonomous unmanned vehicle, the properties of the objects (two-dimensional shape, three-dimensional shape, color of the object, relative position of the object with respect to the object, etc.) existing around the vehicle during traveling or work are detected. However, the following detection methods were used.

That is, a method in which one video camera is mounted on an unmanned vehicle to detect the two-dimensional shape and color of an object around the vehicle,
A method of detecting the three-dimensional shape of an object and the relative position of the vehicle with respect to the object by stereoscopically viewing objects around the vehicle by mounting two video cameras on the unmanned vehicle, a laser range finder (laser mounted on the unmanned vehicle) Radar) to detect the three-dimensional shape of the object and the relative position of the vehicle to the object.
A method of detecting a three-dimensional shape or a relative position of a vehicle with respect to an object, a method of detecting the presence of an object with a tactile sensor such as a limit switch mounted on an unmanned vehicle, and the like have been used.

[0004]

Among the above-mentioned conventional detection methods, the detection method using a video camera or a laser radar is used when an object of a certain size or more exists on the path of an unmanned vehicle. Recognized all of the objects as obstacles. Therefore, when an unmanned vehicle equipped with a video camera or a laser radar is operated in a limited and organized place such as a factory, an object of a certain size or more existing on the course of the unmanned vehicle is recognized as an obstacle. There is no problem at all. However, when the unmanned vehicle is operated on uneven terrain outdoors, structures, walls, large trees, etc. may be recognized as obstacles. There is a problem that it is erroneously recognized as an obstacle similar to the above. For this reason, in places such as grasslands, there was a problem that unmanned vehicles were stuck and could not be operated. Further, in a detection method using an ultrasonic sensor or a tactile sensor, there is a possibility that the sensor may contact an obstacle or the like to impair the sensor function.

[0005]

SUMMARY OF THE INVENTION The object of the present invention is to improve the disadvantages of the prior art described above, and in particular, to determine whether or not an object existing around an unmanned vehicle interferes with traveling, thereby enabling smooth operation of the unmanned vehicle. It is an object of the present invention to provide an unmanned vehicle obstacle detection device that is enabled.

[0006]

According to the present invention, there is provided an object detecting means for detecting the presence / absence and position of an object outside an unmanned vehicle by an ultrasonic sensor or the like; a photographing means for photographing a situation outside the unmanned vehicle; Blower means for sending wind toward the outside of the unmanned vehicle, and control means for controlling the respective means, the control means actuating the blower means in accordance with the detection of the object by the object detection means and directed to the detected object A start-up timing control function to start air blowing, a fluctuation judgment function to compare the photographed image of the detected object before the start of air blowing and a photographed image after the start of the air blow to determine whether the detected object has fluctuated, and a detection object When the object fluctuates, the detection object is determined to be a non-obstacle, and when the detection object does not fluctuate, the detection object is determined to be an obstacle. This aims to achieve the above-mentioned object.

[0007]

According to the present invention, not only the position of the object existing outside the unmanned vehicle but also the obstacle to the running of the unmanned vehicle is determined based on whether or not the object swings by blowing to the object existing outside the unmanned vehicle. Since it is detected whether the vehicle is an obstacle or a non-obstacle that does not hinder traveling, it is possible to accurately grasp the situation outside the unmanned vehicle. Therefore, if the object does not hinder the traveling of the unmanned vehicle (for example, plants and trees), the unmanned vehicle is caused to travel normally, and the object becomes the obstacle to the traveling of the unmanned vehicle (eg, a structure or a large tree). In such a case, operation according to the external situation of the unmanned vehicle, such as temporarily stopping the running of the unmanned vehicle, becomes possible. As a result, it is possible to eliminate such a problem that plants and the like that do not hinder the running of the unmanned vehicle as an obstacle are erroneously recognized as obstacles and the unmanned vehicle is stuck, so that the unmanned vehicle can be operated without trouble even in a grassland or the like. It becomes possible.

[0008]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment in which an unmanned vehicle obstacle detection device of the present invention is applied to an autonomous unmanned vehicle will be described below with reference to the drawings.

First, an overall configuration of an autonomous unmanned vehicle (hereinafter abbreviated as an unmanned vehicle) equipped with the obstacle detection device of the present embodiment will be described with reference to FIG. Is configured to include a three-dimensional object recognition device 2, an obstacle detection device 3, and an integrated control device 4. The unmanned vehicle 1 sends wind toward a work area around the vehicle or an object existing on the path of the vehicle, detects the fluctuation of the object due to the air blowing, and determines whether the object is obstructed to travel or work based on the presence or absence of the fluctuation. It has a function to determine whether or not it will be. Reference numerals 5 and 6 in the figure indicate running wheels.

The three-dimensional object recognizing device 2 constantly detects, for example, the shape and position of an object existing in the work area around the unmanned vehicle or on the course of the unmanned vehicle by the ultrasonic sensor 2A,
The detection data is output to the integrated control device 4. In this case, use a video camera or laser range finder (laser laser) instead of using an ultrasonic sensor.
May be used to detect the shape and position of the object.

When a detection direction command value indicating the direction of an object to be detected is input from the integrated control device 4 to the obstacle detection device 3, the obstacle detection device 3 detects a work area around the unmanned vehicle detected by the three-dimensional object recognition device 2 or Detects whether or not the object fluctuates by blowing on the object existing on the path of the unmanned vehicle, in other words, whether the object is a soft object that does not hinder the traveling of the unmanned vehicle or a robust object that hinders the traveling. The obstacle data or the non-obstacle data is output to the integrated control device 4 according to the detection result. In this case, when the place where the unmanned vehicle 1 is operated is on uneven terrain, for example, structures and large trees are obstacles, and plants and the like are non-obstacles.

The integrated control device 4 includes the three-dimensional object recognition device 2
When detection data indicating the shape or position of an object existing in the work area of the unmanned vehicle or on the path of the unmanned vehicle is input from the device, the azimuth where the object is located is detected based on the detection data, and the obstacle is detected. A detection azimuth command value is output to the object detection device 3 to detect whether or not the object becomes an obstacle to the unmanned vehicle.

Further, the integrated control device 4 associates the object detected by the three-dimensional object recognition device 2 with the object detected as a non-obstacle by the obstacle detection device 3,
The objects detected by the obstacle detection device 3 as non-obstacles are excluded from the objects detected by the three-dimensional object recognition device 2. When the integrated control device 4 determines that the object existing in the work area of the unmanned vehicle or the path of the unmanned vehicle is a non-obstacle (for example, a plant or the like), the integrated control device 4 controls a driving mechanism (not shown). An unmanned vehicle is normally driven.

Next, the detailed configuration of the obstacle detection device 3 is shown in FIG.
The obstacle detection device 3 includes a blower 7
, A video camera unit 8, an attitude control unit 9, an image processing unit 10, and a main control unit 11. The blower unit 7 and the video camera unit 8 are fixed to a camera platform 12. I have. The head 12 is provided with a servo mechanism 13, and the attitude control unit 9 is provided with a servo mechanism 14.

The blower 7 comprises a blower for blowing pulsating wind to an object existing in the working area of the unmanned vehicle or the path of the unmanned vehicle, a nozzle for emitting wind, and the like. When a blow start command is input via the line 15, the blower intermittently generates wind and blows wind (pulsating wind) from the nozzle toward the object. In addition, when a blow stop command is input from the main control unit 11 via the signal line 15, the blower 7 stops the intermittent generation of wind by the blower. In this case, for example, if air is continuously blown to a soft object such as a plant or the like, the object is in an unsteady state and it cannot be checked whether the object is fluctuating due to the air blow. By doing so, it is easy to confirm the presence or absence of the fluctuation of the object.

The attitude of the video camera section 8 is controlled by the attitude control section 9 together with the blower section 7 via the camera platform 12, and the video camera section 8 emits the wind before and from the blower section 7. An image of the object to be detected later is photographed, and the image processing unit 1
The original image data of the detected object before the air blowing and the original image data of the detected object after the air blowing are output to signal line 0 through signal line 16.

The attitude control unit 9 adjusts the angle of the camera platform 12 via the servo mechanisms 14 and 13 when the detected azimuth command value is input from the main control unit 11 via the signal line 17. The azimuths of the blower unit 7 and the video camera unit 8 are controlled so that the wind emitted from the blower unit 7 hits the detected object and the detected object enters the imaging range of the video camera unit 8. I have. In addition, the posture control unit 8
When the actual azimuths of the blower unit 7 and the video camera unit 8 match the detected azimuth command value, an attitude control end response is output to the main control unit 11 via the signal line 17 to report that the attitude control has ended. It is supposed to.

As shown in FIG. 3, the image processing unit 10 has an image input unit 21 connected to the video camera unit 8 by a signal line 16, and an image processing unit 10 connected to the main control unit 11 by a signal line 18, and an integrated control unit 4. And a memory 23 having an area for storing an image of the detected object before blowing, an area for storing an image of the detected object after blowing, a work area, a program area, and the like. And an image output unit 24 connected to the integrated control device 4 by a signal line 20
And a CPU (Central Processing Unit) 25. In the figure, reference numeral 26 indicates a bus for transmitting and receiving signals.

When an image capture command SA or SB (described later) is input from the main control unit 11 via a signal line 18 to the image processing unit 10, the image processing unit 10 transmits the image capture command SA or SB via the signal line 16 from the video camera unit 8 before the air is blown. The image data of the detected object is blown after the air is blown, and when the image data is completely captured, an image capture completion response is output to the main control unit 11 via the signal line 18. The image processing unit 10 outputs non-obstacle data or obstacle data based on the image processing result to the integrated control device 4 via the signal line 20, and outputs the signal line 1.
9, an input completion response is output.
Details of the image processing will be described in detail with reference to FIG.

When a detected direction command value indicating the direction of the object to be detected is input from the integrated control device 4 via the signal line 27 from the integrated control device 4, the main control unit 11 sends the detected direction command value to the attitude control unit 9 via the signal line 17. When a detection direction command value is output and a posture control end response is input from the posture control unit 9 via the signal line 17,
An image capture command SA for capturing an image of the detected object before blowing is output to the image processing unit 10 via a signal line 18.

When a response to the completion of image capture based on the image capture command SA is input from the image processor 10 via the signal line 18, the main controller 11 blows air to the blower 7 via the signal line 15. After outputting a start command to start blowing air to the detected object, a sufficient time is required until the wind of the blowing unit 7 hits the detected object, and then the image processing unit 10 is transmitted to the image processing unit 10 via the signal line 18. An image capture command SB for capturing an image of the detected object after blowing (the wind is blown) is output. Further, when an image capture end response based on the image capture command SB is input from the image processor 10 via the signal line 18, the main controller 11 sends a blow stop command to the blower 7 via the signal line 15. The air is output to stop blowing air to the detected object.

Next, an obstacle detection operation in the embodiment constructed as described above will be described mainly with reference to FIGS.

When the unmanned vehicle 1 is traveling, the three-dimensional object recognizing device 2 always detects the shape and position of the object existing in the working area around the unmanned vehicle or on the course of the unmanned vehicle by the ultrasonic sensor. The detection data is output to the integrated control device 4.

On the other hand, the main controller 11 of the obstacle detection device 3
When the detected azimuth command value is input from the integrated control device 4 via the signal line 27 (step S1), the detected azimuth command value is output to the attitude control unit 9 via the signal line 17 (step S2). . Thereby, the attitude control unit 11 adjusts the angle of the camera platform 12 via the servo mechanisms 14 and 13 so that the wind emitted from the blower unit 7 hits the detection target object and the video camera unit 8 The azimuth of the blower unit 7 and the video camera unit 8 is controlled so that the detected object enters the imaging range. The video camera unit 8 in which the detection direction is controlled,
An image of the detected object before being blown from the blower 7 is captured.

The main control unit 11 receives an attitude control end response from the attitude control unit 9 via the signal line 17 when the actual directions of the blower unit 7 and the video camera unit 8 match the detected azimuth command value. When it comes (step S3), an image capture command SA for capturing an image of the detected object before blowing is output to the image processing unit 10 via the signal line 18 (step S3).
4). As a result, the image processing unit 10 transmits the image A of the detected object before blowing from the video camera unit 8 via the signal line 16.
Take in.

When the main controller 11 receives an image capture end response based on the image capture command SA from the image processor 10 via the signal line 18 (step S5), the main controller 11 sends the signal to the blower 7 via the signal line 15. To output a blow start command (step S6). As a result, the blower 7 generates intermittent wind by the blower and emits the wind from the nozzle toward the detected object. At this time, the video camera unit 8 captures an image of the detected object blown by the blower 7.

After the main control unit 11 outputs the blow start command,
After a sufficient time until the wind of the blower 7 hits the detected object, an image capture command SB is output to the image processor 10 via the signal line 18 to capture the image of the detected object after blowing. (Step S7). Thereby, the image processing unit 1
0 captures an image B of the detected object to which the air has been blown (to which the wind has been applied) from the video camera unit 8 via the signal line 16.

The image processing unit 10 determines the edge of the image A of the detected object before the air blown in from the video camera unit 8 based on the image capture command SA in step S4 and the image capture command SB in step S7. Video camera section 8
The edge of the image B of the object to be detected after blown (taken by the wind) taken from the image B is extracted, the image A ′ after the extracted edge is compared with the image B ′, and the image A ′ and the image are compared. Image C that does not fluctuate even when blown by taking the logical product with B '
(A large tree in the illustrated example) is detected.

Next, the image processing section 10 removes noises N and N '(trace images of plants and the like after image processing) from the image C which does not fluctuate even when the air is blown, using a noise filter (not shown).
The unstabilized image C ′ from which noise has been removed is output as obstacle data to the integrated control device 4 via the signal line 20. Further, the image processing unit 10 subtracts the image C ′ having no fluctuation after noise removal from the image A ′ after the edge extraction, and outputs the obtained image D (plants in the illustrated example) to the integrated control device 4 as non-obstacle data. Output via line 20.

Thus, the integrated control device 4 can control the object detected by the three-dimensional object recognition device 2 and the obstacle detection device 3
And the object detected by the three-dimensional object recognition device 2 is excluded from the objects detected by the obstacle detection device 3 as non-obstacles (for example, plants and the like). That is, when the object existing in the work area or the course of the unmanned vehicle is a non-obstacle (for example, a plant or the like), the integrated control device 4 controls the driving mechanism (not shown) to cause the unmanned vehicle to travel normally.

Thereafter, the main control unit 11 controls the image processing unit 10
When an image capture end response is input via the signal line 18 (step S8), a blow stop command is output to the blower unit 7 via the signal line 15 (step S9). Thus, the blower 7 stops blowing air to the detected object.

As described above, according to the present embodiment, not only the shape and position of the object existing around the unmanned vehicle but also whether the object swings by blowing air to the object existing around the unmanned vehicle is determined. Detects whether the vehicle is an obstacle that does not interfere with the traveling of the unmanned vehicle or is a non-obstruction that does not interfere with the traveling of the unmanned vehicle, so that the situation around the unmanned vehicle can be accurately grasped. If the vehicle does not hinder the running of the vehicle (for example, plants and trees), the unmanned vehicle is allowed to travel normally. The unmanned vehicle can be smoothly operated according to the surrounding situation, for example, by temporarily stopping running.

Therefore, it is possible to solve the problem that plants and the like that do not hinder the operation of the unmanned vehicle are erroneously recognized as obstacles, and that the unmanned vehicle is stuck and cannot be operated as in the related art. In addition, the unmanned vehicle can be operated without trouble even in a grassland, for example.

[0034] Further, according to the above description, it is possible to select and drive the unmanned vehicle on a path in which an object (for example, a plant or the like) that does not hinder the traveling is performed. Therefore, the number of obstacles on the path of the unmanned vehicle is substantially reduced. As a result, the unmanned vehicle can reach the target point earlier than before.

Further, the above makes it possible to avoid unnecessary travel of the unmanned vehicle, so that it is possible to reduce the consumption of energy sources such as electric power and fuel for driving the unmanned vehicle, and as a result, In addition, the operation efficiency of the unmanned vehicle can be improved as compared with the related art.

Furthermore, since an object existing around the unmanned vehicle is detected in a non-contact manner, there is an effect that the object and the unmanned vehicle are not damaged.

[0037]

As described above, according to the obstacle detection device for an unmanned vehicle according to the present invention, not only the position of the object existing outside the unmanned vehicle but also the air flowing to the object existing outside the unmanned vehicle is blown. It is possible to accurately grasp the situation outside the unmanned vehicle in order to detect whether the object is an obstacle that does not interfere with the traveling of the unmanned vehicle or a non-obstruction that does not interfere with the traveling based on whether the vehicle fluctuates Therefore, if the object is an object that does not hinder unmanned vehicle traveling (for example, vegetation), the unmanned vehicle is allowed to travel normally, and the object is an obstacle to unmanned vehicle traveling (for example, a structure or a large tree, etc.). ), It is possible to operate the unmanned vehicle smoothly according to the external situation, such as temporarily stopping the running of the unmanned vehicle,
As a result, it is possible to eliminate such a problem that plants and the like that do not hinder the running of the unmanned vehicle as an obstacle are erroneously recognized as obstacles and the unmanned vehicle is stuck, so that the unmanned vehicle can be operated without trouble even in a grassland or the like. This makes it possible to achieve the effect of being able to do so. In addition, the above allows the unmanned vehicle to travel while selecting a path on which an object that does not hinder the traveling (for example, a plant or the like) exists. Therefore, it is possible to substantially reduce the number of obstacles on the path of the unmanned vehicle. As a result, the unmanned vehicle can reach the target point earlier than before. Further, the above makes it possible to avoid unnecessary travel of the unmanned vehicle, so that it is possible to reduce the consumption of the energy source of the unmanned vehicle, and as a result, the operation efficiency of the unmanned vehicle is improved as compared with the conventional case. There is an effect that can be.

Further, the control means further comprises a second air blowing start timing control function for intermittently operating the air blowing means in response to the detection of the object by the object detecting means to start the intermittent air blowing toward the detected object. In the case of, for example, it is easy to confirm the presence or absence of fluctuation of a soft detection object such as a plant,
There is an effect that it is possible to more reliably determine whether the detected object is a non-obstacle or an obstacle.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of an obstacle detection device according to an embodiment of the present invention.

FIG. 2 is a schematic diagram illustrating a configuration of an unmanned vehicle according to the present embodiment.

FIG. 3 is a block diagram illustrating a configuration of an image processing unit according to the embodiment.

FIG. 4 is a flowchart showing a control sequence of a main control unit in the embodiment.

FIG. 5 is an explanatory diagram illustrating image processing of an image processing unit according to the present embodiment.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Unmanned vehicle 2 Three-dimensional object recognition device as an object detection means 3 Obstacle detection device 4 Integrated control device 7 Blower unit as a blower unit 8 Video camera unit as a photographing unit 9 Attitude control unit 10 Image processing unit 11 As a control unit Main control section

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int.Cl. ⁷ , DB name) G01S ⁷ /52-7/64 G01S 15/00-15/96 G05D 1/02

Claims (2)

(57) [Claims] An object detecting means for detecting the presence or absence and position of an object outside an unmanned vehicle by an ultrasonic sensor or the like,
A photographing means for photographing a situation outside the unmanned vehicle; a blowing means for blowing wind toward the outside of the unmanned vehicle; and a control means for controlling each of the means. The air blowing start timing control function for starting the air blowing toward the detection object by operating the air blowing means in accordance with the detection, and comparing the captured image of the detection object before the air blowing is started with the captured image after the air blowing is started, to detect the detected object. A fluctuation judgment function for judging whether or not the vibration has occurred;
An obstacle detection function for an unmanned vehicle, comprising: an obstacle determination function that determines the detection object as a non-obstacle when the detection object swings and determines the detection object as an obstacle when the detection object does not swing. apparatus. 2. The second blowing start timing control, wherein the control means further intermittently operates the blowing means in accordance with the detection of an object by the object detecting means to start the intermittent blowing toward the detected object. The obstacle detection device for an unmanned vehicle according to claim 1, further comprising a function.