JP7567738B2 - Obstacle Detection Device - Google Patents

Description

The present invention relates to an obstacle detection device.

Moving bodies such as vehicles may be equipped with an obstacle detection device for detecting obstacles that impede the progress of the moving body. For example, obstacle detection devices are known that use sensors such as a laser range finder, a stereo camera, or a millimeter wave radar to derive the distance and direction to a point that represents part of an object, and detect obstacles based on the derived results (for example, Patent Document 1).

JP 2014-106604 A

The obstacle detection device disclosed in Patent Document 1 detects objects such as raindrops, insects, and dust that do not impede the progress of a moving object as obstacles, which can sometimes impede the movement of the moving object.

An obstacle detection device that achieves the above-mentioned objective is an obstacle detection device mounted on a moving body, and is characterized by comprising: a sensor that detects the distance and direction to a point that represents a part of an object; an acquisition unit that acquires the detection result from the sensor; a detection unit that detects the position of the object from the detection result; a prediction unit that predicts a predicted future position of the object based on the detection result of the detection unit; and a determination unit that determines whether the object is an obstacle based on the predicted position and the detection result of the detection unit.

With this configuration, the obstacle detection device can prevent the movement of the moving body from being impeded by determining, as an obstacle, an object that does not impede the movement of the moving body based on whether or not an object is present at the predicted position.

In the obstacle detection device, when predicting the destination of the stopped object, the prediction unit may predict the predicted position based on the movement of the moving body, and the determination unit may determine that the object is an obstacle when the detection result of the detection unit indicates that the object is present at the predicted position.

According to this configuration, if the object is not present at the predicted position according to the movement of the moving body, the obstacle detection device determines that there is no obstacle that would impede the progress of the moving body. As a result, the obstacle detection device can prevent the travel of the moving body from being impeded by determining that an object that does not impede the progress of the moving body is an obstacle.

In the obstacle detection device, when predicting the destination of the moving object, the prediction unit predicts the predicted position based on detection results from a plurality of the detection units, and the determination unit may determine that the object is an obstacle when the detection results from the detection units indicate that the object is present within a predicted range up to a predetermined distance from the predicted position.

According to this configuration, if no object is present within the prediction range, the obstacle detection device determines that no object that would impede the progress of the moving body is present. As a result, the obstacle detection device is able to prevent the movement of the moving body from being impeded by determining that an object that does not impede the progress of the moving body is an obstacle.

In the obstacle detection device, the detection unit may perform the process of detecting the object at a predetermined time interval, and the prediction unit may increase the width of the prediction range as the predetermined time interval increases.

In the obstacle detection device, the determination unit may perform a plurality of determinations as to whether the object is an obstacle, and may determine that the object is an obstacle if the object is determined to be an obstacle a predetermined number of times or more among the plurality of determinations.

With this configuration, the determination unit can accurately determine whether or not an obstacle is present.

According to the present invention, the obstacle detection device determines, based on whether an object is present at a predicted position, that an object that does not impede the progress of a moving body is an obstacle, thereby preventing the movement of the moving body from being impeded.

1 is a schematic plan view showing an example of a forklift equipped with an obstacle detection device. FIG. 2 is a diagram illustrating an example of a configuration of an obstacle detection device. 1A to 1C are diagrams used to explain a prediction process and a determination process according to the first embodiment. 1A to 1C are diagrams used to explain a prediction process and a determination process according to the first embodiment. 4 is a flowchart showing an example of an operation of the forklift 10 according to the first embodiment. 13A to 13C are diagrams used to explain a prediction process and a determination process according to a second embodiment. 10 is a flowchart showing an example of an operation of the forklift 10 according to the second embodiment. 13A to 13C are diagrams used to explain other examples of the prediction process and the determination process. 13A to 13C are diagrams used to explain other examples of the prediction process and the determination process.

First Embodiment
Hereinafter, a first embodiment of an obstacle detection device will be described with reference to the drawings. As shown in Fig. 1, a forklift 10 as a moving body includes a vehicle body 11, drive wheels 12 arranged at a front lower part of the vehicle body 11, a steering wheel 13 arranged at a rear lower part of the vehicle body 11, and a loading device 14. The forklift 10 is an industrial vehicle that transports a load supported by the loading device 14. The forklift 10 may be an automatic vehicle, may be a vehicle operated by a human, or may be a vehicle that can be switched between automatic operation and manual operation.

As shown in FIG. 2, the forklift 10 includes a drive mechanism 21, a control device 22, a sensor 31, and an obstacle detection device 40. The drive mechanism 21 is a component for causing the forklift 10 to travel, and includes a travel motor for driving the drive wheels 12 and a steering mechanism for steering the steering wheels 13.

The control device 22 controls the traveling of the forklift 10. The control device 22 controls the driving mechanism 21 to make the forklift 10 travel. The control device 22 includes a processor and a storage unit as hardware. The processor may be, for example, a central processing unit (CPU), a graphics processing unit (GPU), or a digital signal processor (DSP). The storage unit includes a random access memory (RAM) and a read only memory (ROM). The storage unit stores program code or instructions configured to cause the processor to execute processing. The storage unit, i.e., the computer-readable medium, includes any available medium that can be accessed by a general-purpose or dedicated computer. The control device 22 may be configured by a hardware circuit such as an application specific integrated circuit (ASIC) or a field programmable gate array (FPGA). The control device 22, which is a processing circuit, may include one or more processors that operate according to a computer program, one or more hardware circuits such as an ASIC or an FPGA, or a combination thereof.

The sensor 31 is for detecting objects present around the forklift 10. The sensor 31 is capable of measuring the distance and direction to a point. In this embodiment, a laser range finder is used as the sensor 31. The laser range finder is also called LIDAR (Laser Imaging Detection and Ranging) or a laser range finder. The sensor 31 is a range finder that can recognize the surrounding environment by emitting a laser and receiving the reflected light from the point where the laser hits. Since the laser is reflected by hitting an object, it can be said that the point represents a part of the surface of the object. Note that the object here includes an obstacle that impedes the progress of the forklift 10 and an object that does not impede the progress of the forklift 10. Objects that do not impede the progress of the forklift 10 include raindrops, insects, dust, and snowflakes. In this embodiment, raindrops are used as an example of an object that does not impede the progress of the forklift 10.

As shown in FIG. 1, the sensor 31 irradiates the laser while changing the irradiation angle of the laser within the detectable range TA. The detectable range TA is, for example, a sector-shaped range determined by the irradiation angle θ at which the laser can be irradiated and the reachable distance L1 of the laser. The sensor 31 of this embodiment irradiates the laser while changing the irradiation angle in one direction within the irradiation angle θ by a constant angle θ1 according to the angular resolution. The irradiation angle is, for example, an angle when the center of the irradiation angle θ is set as the reference, i.e., 0 degrees. The sensor 31 is attached so that the direction in which the irradiation angle is changed is the horizontal direction. As a result, the sensor 31 irradiates the laser while changing the irradiation angle in the horizontal direction.

The sensor 31 can measure the distance to the point where the laser hits from the time required from emitting the laser to receiving the reflected light. The irradiation angle at which the laser is irradiated can be said to indicate the direction from the sensor 31 to the point. Since the sensor 31 irradiates the laser while changing the irradiation angle in the horizontal direction, the irradiation angle can be said to indicate the direction of the point relative to the horizontal direction. If the reflected light cannot be received despite irradiating the laser, it can be said that there is no object that reflects the laser at the irradiation angle of the laser. The sensor 31 detects information that associates the distance to the point with the irradiation angle. By using the distance to the point and the irradiation angle, it is possible to derive the coordinates of the point in a coordinate system with the sensor 31 as the origin. The coordinates can be said to represent the coordinates on a coordinate plane that represents the horizontal direction. The detection result of the sensor 31 may be any one that represents the positional relationship between the sensor 31 and the point in the horizontal direction. In this embodiment, the detection result of the sensor 31 includes information indicating the coordinates of the point and information indicating the irradiation angle when the point is detected.

A plurality of sensors 31 are provided. The plurality of sensors 31 include three sensors 31A, 31B, and 31C provided to detect objects in front of the forklift 10. The sensor 31A is attached to the vehicle body 11 so that the front of the forklift 10 is the detectable range TA. The sensor 31A is attached, for example, to the front of the vehicle body 11. The sensor 31B is attached to the vehicle body 11 so that the front of the forklift 10 and the left side of the forklift 10 are the detectable range TA. The sensor 31C is attached to the vehicle body 11 so that the front of the forklift 10 and the right side of the forklift 10 are the detectable range TA. The sensor 31A is attached, for example, to the front of the vehicle body 11. The two sensors 31B and 31C are attached, for example, separately to the left and right. Note that in FIG. 1, only the detectable range TA of the sensor 31A is shown, but the two sensors 31B and 31C also have the same detectable range TA as the sensor 31A.

By arranging the sensors 31 as described above, at least the front of the periphery of the forklift 10 in the horizontal direction is included in the detectable range TA of any of the sensors 31. Regardless of whether an object is present in front of the forklift 10 or to the left or right, it can be said that the object is located within the detectable range TA of any of the sensors 31. The three sensors 31 may be arranged so that their detectable ranges TA partially overlap each other, or so that their detectable ranges TA do not overlap each other.

2, the obstacle detection device 40 includes a control unit 100 and a storage unit 200 as hardware. The control unit 100 may be, for example, a CPU, a GPU, or a DSP. The storage unit 200 includes a RAM and a ROM. The storage unit 200 stores program code or instructions configured to cause the control unit 100 to execute processing. The storage unit 200, i.e., the computer-readable medium, includes any available medium accessible by a general-purpose or dedicated computer. The control unit 100 may be configured by a hardware circuit such as an ASIC or an FPGA. The control unit 100, which is a processing circuit, may include one or more processors that operate according to a computer program, one or more hardware circuits such as an ASIC or an FPGA, or a combination thereof.

The control unit 100 includes an acquisition unit 101, a detection unit 102, a prediction unit 103, and a determination unit 104 as functional elements that function by executing a program stored in the memory unit 200.

The acquisition unit 101 acquires information indicating the detection result from the sensor 31 .
The detection unit 102 detects the position of an object based on the detection result acquired by the acquisition unit 101. The detection result of the sensor 31 is information indicating a plurality of points. As described above, the points represent coordinates on a coordinate plane representing the horizontal direction. When a plurality of sensors 31 are provided, the positions of the origins of the respective sensors 31 are different. When a plurality of sensors 31 are provided, the orientation of the coordinate axes of the respective sensors 31 may be different. The detection unit 102 merges the detection results of the three sensors 31, and sets the coordinates of the points obtained by the three sensors 31 to coordinates of the same coordinate system, and then detects the position of the object using these coordinates. The detection unit 102 performs detection processing, for example, at predetermined time intervals.

For ease of explanation, this embodiment will be described assuming that the forklift 10 is positioned on a horizontal plane and that coordinates in a coordinate system defined on the horizontal plane with the forklift 10 as the reference are acquired from the three sensors 31. In this coordinate system, the axis extending in the width direction of the forklift 10 in the horizontal direction is the X-axis, and the axis perpendicular to the X-axis in the horizontal direction is the Y-axis. It can be said that the coordinates in this coordinate system indicate the position of a point in the horizontal direction.

The prediction unit 103 predicts a future position of the object based on the detection result of the detection unit 102. Hereinafter, the future position of the object predicted by the prediction unit 103 will be referred to as a "predicted position".
The determination unit 104 determines whether or not an object exists based on the predicted position predicted by the prediction unit 103 and the detection result of the detection unit 102. In detail, the determination unit 104 determines that an object exists when the position of the object detected by the detection unit 102 is the predicted position predicted by the prediction unit 103. Furthermore, the determination unit 104 determines whether or not the object determined to exist is an obstacle based on the predicted position and the detection result of the detection unit 102.

[Details of prediction process and determination process]
3 and 4, the process of the prediction unit 103 and the process of the determination unit 104 will be described in detail below. The prediction unit 103 in this embodiment predicts the predicted position of a stopped object based on the movement of the forklift 10.

In FIG. 3, the detection unit 102 detects a stopped object P1. As described above, the detection unit 102 performs a detection process at a predetermined time interval. The prediction unit 103 predicts a predicted position FP of the object P1 based on the amount of movement of the forklift 10 from when the detection unit 102 detects the object P1 until the timing of the next detection process. In detail, when the displacement of the forklift 10 in the +X-axis direction is the amount of movement Δx, the prediction unit 103 sets the coordinate obtained by subtracting the amount of movement Δx from the X-coordinate of the object P1 detected by the detection unit 102 as the X-coordinate of the predicted position FP. When the displacement of the forklift 10 in the -X-axis direction is the amount of movement Δx, the prediction unit 103 sets the coordinate obtained by adding the amount of movement Δx to the X-coordinate of the object P1 detected by the detection unit 102 as the X-coordinate of the predicted position FP. Furthermore, when the displacement of the forklift 10 in the +Y axis direction is a movement amount Δy, the prediction unit 103 determines the coordinate obtained by subtracting the movement amount Δy from the Y coordinate of the object P1 detected by the detection unit 102 as the Y coordinate of the predicted position FP.

At the next detection timing, the detection unit 102 detects object P2. The determination unit 104 determines that the position of object P2 detected by the detection unit 102 matches the predicted position FP predicted by the prediction unit 103. In this case, the determination unit 104 determines that object P1 has been displaced to the position of object P2 as the forklift 10 moves, and determines that the object is an obstacle.

In FIG. 4, the determination unit 104 determines that the position of object P2 detected by the detection unit 102 does not match the predicted position FP predicted by the prediction unit 103. In this case, the determination unit 104 determines that the stopped object P1 has not been displaced to the position of object P2 as the forklift 10 moves, but that objects P1 and P2 are different objects. In other words, the determination unit 104 determines that objects P1 and P2 are objects that do not impede progress, and determines that no obstacles exist.

Here, if object P1 and object P2 are different objects, object P2 should be detected at the timing when detection unit 102 first detects object P1. However, object P2 is not detected at the timing when detection unit 102 first detects object P1. Furthermore, object P1 is not detected at the timing when detection unit 102 later detects object P2. Therefore, object P1 and object P2 detected by detection unit 102 are likely to be raindrops or the like that do not impede the progress of forklift 10, and determination unit 104 determines that no obstacle is present.

As shown in Figures 3 and 4, a monitoring range SA is provided within the detectable range TA. The monitoring range SA is, for example, a range on a horizontal plane designated in advance, and is a rectangular range provided in front of the forklift 10. Information indicating the monitoring range SA is, for example, stored in the memory unit 200. The information indicating the monitoring range SA is, for example, information indicating the coordinates of the vertices of the monitoring range SA in the detectable range TA. If an object other than an object that does not hinder the progress of a moving object, such as raindrops, insects, or dust, exists within the monitoring range SA, the object is an obstacle that hinders the travel of the forklift 10. The determination unit 104 refers to the information indicating the monitoring range SA, and if the position of the object determined to exist is within the monitoring range SA, outputs a detection result including information indicating that the object is an obstacle in the monitoring range SA to the control device 22.

The control device 22 controls the forklift 10 according to the detection results obtained from the obstacle detection device 40. For example, if an obstacle is present in the monitoring range SA, the control device 22 stops or decelerates the forklift 10 by controlling the drive mechanism 21. The forklift 10 is decelerated, for example, by setting an upper vehicle speed limit. The control device 22 drives the forklift 10 so that the vehicle speed of the forklift 10 does not exceed the upper vehicle speed limit. Therefore, if the vehicle speed of the forklift 10 is below the upper vehicle speed limit, the forklift 10 may not be decelerated even if an obstacle is present in the monitoring range SA. The upper vehicle speed limit may be a predetermined constant value, or may be a variable value that decreases as the obstacle approaches the forklift 10.

[Operation flow]
Hereinafter, the details of the operation of the forklift 10 according to the first embodiment will be described with reference to FIG. 5. The process shown in the flowchart of FIG. 5 is repeatedly executed at predetermined time intervals. First, the acquisition unit 101 acquires detection results from each sensor 31 (step S100). Next, the detection unit 102 detects an object present in the detectable range TA based on the detection results acquired in step S100, and specifies the position of the object (step S102). Next, the prediction unit 103 predicts a predicted position FP of the object based on the position of the object specified by the detection unit 102 in step S102 (step S104). Next, the acquisition unit 101 acquires detection results from each sensor 31 (step S106). Next, the detection unit 102 detects an object present in the detectable range TA based on the detection results acquired in step S106, and specifies the position of the object (step S108).

The determination unit 104 determines whether or not an object is present at the predicted position FP based on the predicted position FP predicted by the prediction unit 103 and the position of the object identified in step S108 (step S110). If the position of the object identified in step S108 matches the predicted position FP, the determination unit 104 determines that the object is an obstacle, and if the position of the object identified in step S108 does not match the predicted position FP, the determination unit 104 determines that the object is not an obstacle.

If it is determined that there is an obstacle, the determination unit 104 determines whether or not an obstacle is present in the monitoring range SA (step S112). If the determination unit 104 determines that there is an obstacle in the monitoring range SA, the determination unit 104 outputs a detection result to the control device 22. The detection result includes information indicating that an obstacle is present in the monitoring range SA. If the control device 22 obtains a detection result including information indicating that an obstacle is present in the monitoring range SA from the obstacle detection device 40, it stops or decelerates the forklift 10 (step S114).

When it is determined that no obstacle is present, or that the object is an obstacle but is not present in the monitoring range SA, the determination unit 104 outputs a detection result to the control device 22. The detection result includes information indicating that no obstacle is present, or that the object is an obstacle but is not present in the monitoring range SA. When the control device 22 obtains a detection result including information indicating that no obstacle is present, or that the object is an obstacle but is not present in the monitoring range SA from the obstacle detection device 40, the control device 22 allows the forklift 10 to continue traveling (step S116). Next, the prediction unit 103 predicts a predicted position FP of the object based on the position of the object identified by the detection unit 102 in step S108 (step S118), and the process proceeds to step S106.

[Effects of the obstacle detection device 40 of the first embodiment]
(1-1) In this embodiment, the determination unit 104 determines whether or not an object is an obstacle based on the predicted position FP predicted by the prediction unit 103 and the detection result of the detection unit 102. In detail, when predicting the movement destination of a stopped object, the prediction unit 103 predicts the predicted position FP based on the movement of the forklift 10. When the detection result of the detection unit 102 indicates that an object is present at the predicted position FP, the determination unit 104 determines that the object is an obstacle.

Here, the detection unit 102 may detect objects such as raindrops, insects, dust, and snowflakes that do not impede the progress of the forklift 10. On the other hand, even if the detection unit 102 detects objects such as raindrops, insects, dust, and snowflakes that do not impede the progress of the forklift 10, there is a high possibility that they will not be detected again at the predicted position FP. In this embodiment, the determination unit 104 determines that an object detected by the detection unit 102 is an obstacle that impedes the progress of the forklift 10 when the object is present at the predicted position FP. In other words, the determination unit 104 can determine whether or not an obstacle that impedes the progress of the forklift 10 is present.

In general, if the control device 22 stops or slows down the forklift 10 based on the presence of an object that does not impede the progress of the forklift 10, such as raindrops, insects, dust, and snowflakes, the forklift 10 may not be able to travel properly. The obstacle detection device 40 of this embodiment detects an object and determines whether the object is an obstacle that will impede the progress of the forklift 10. Therefore, the obstacle detection device 40 can prevent the travel of the forklift 10 from being impeded by determining that an object that does not impede the progress of the forklift 10 is an obstacle.

Second Embodiment
Hereinafter, a second embodiment of the obstacle detection device will be described with reference to the drawings. In the above embodiment, the prediction unit 103 predicts the predicted position of a stationary object. In the second embodiment, the prediction unit 103 predicts the predicted position of a moving object.

When predicting the destination of a moving object, the prediction unit 103 predicts the predicted position FP based on the detection results of the multiple detection units 102. The details of the processing of the prediction unit 103 in this embodiment are described below.

As shown in FIG. 6, the detection unit 102 detects an object P1 at a certain detection timing. Then, the detection unit 102 detects an object P2 at the next detection timing. The prediction unit 103 predicts the predicted position FP based on the detection result at the nth detection timing and the detection result at the n+1th detection timing. n is a natural number. In this case, the prediction unit 103 predicts the predicted position FP based on the detection result of the object P1 and the detection result of the object P2. In detail, the prediction unit 103 derives a value obtained by subtracting the X coordinate of the object P2 from the X coordinate of the object P1 as the movement amount Δx1 in the X-axis direction, and derives a value obtained by subtracting the Y coordinate of the object P2 from the Y coordinate of the object P1 as the movement amount Δy1 in the Y-axis direction. The prediction unit 103 predicts the predicted position FP based on the movement amount Δx1 and the movement amount Δy1. For example, the prediction unit 103 determines the coordinate obtained by adding the movement amount Δx1 to the X coordinate of the object P2 as the X coordinate of the predicted position FP, and determines the coordinate obtained by adding the movement amount Δy1 to the Y coordinate of the object P2 as the Y coordinate of the predicted position FP.

At the next detection timing, the detection unit 102 detects an object P3. The determination unit 104 determines whether the position of the object P3 detected by the detection unit 102 matches the predicted position FP predicted by the prediction unit 103, or is within a prediction range FA up to a position a predetermined distance D away from the predicted position FP. The predetermined distance D is, for example, a distance by which the object may displace from the predicted position FP as it moves. Therefore, the prediction range FA is, for example, a range within which the object may move against the object's tendency to move.

In FIG. 6, the determination unit 104 determines that the position of object P3 detected by the detection unit 102 does not match the predicted position FP predicted by the prediction unit 103. On the other hand, the determination unit 104 determines that the position of object P3 detected by the detection unit 102 is within the prediction range FA. In this case, the determination unit 104 determines that an object exists, assuming that object P2 has been displaced to the position of object P3 due to movement. Note that, if the position of object P3 detected by the detection unit 102 does not match the predicted position FP and is not within the prediction range FA, the determination unit 104 determines that objects P2 and P3 are different objects. In other words, the determination unit 104 determines that objects P1 and P2 are objects that do not impede progress, and determines that no obstacle exists.

[Operation flow]
Hereinafter, the operation of the forklift 10 according to the second embodiment will be described in detail with reference to Fig. 7. The process shown in the flowchart of Fig. 7 is repeatedly executed at predetermined time intervals. Moreover, among the processes shown in Fig. 7, the same processes as those shown in Fig. 5 are denoted by the same step reference numerals and the description thereof will be omitted.

In this embodiment, the acquisition unit 101 acquires the detection results from each sensor 31 after the process of step S102 (step S200). Next, the detection unit 102 detects an object present in the detectable range TA based on the detection result acquired in step S200, and identifies the position of the object (step S202). Next, the prediction unit 103 predicts the predicted position FP based on the position of the object identified by the detection unit 102 in step S102 and the position of the object identified by the detection unit 102 in step S202 (step S204), and proceeds to the process of step S106. For example, the prediction unit 103 derives a movement amount Δx1 in the X-axis direction and a movement amount Δy1 in the Y-axis direction from the coordinates of the object identified in step S102 to the coordinates of the object identified in step S202. The prediction unit 103 predicts the predicted position FP based on the movement amount Δx1 and the movement amount Δy1.

The determination unit 104 determines whether or not an object is present at the predicted position FP or within the predicted range FA based on the predicted position FP, the predicted range FA up to a position a predetermined distance D away from the predicted position FP, and the position of the object identified in step S108 (step S206). If the position of the object identified in step S108 matches the predicted position FP or is within the predicted range FA, the determination unit 104 determines that the object is an obstacle and proceeds to step S112. If the position of the object identified in step S108 does not match the predicted position FP and is not within the predicted range FA, the determination unit 104 determines that no obstacle is present and proceeds to step S116.

[Effects of the obstacle detection device 40 according to the second embodiment]
(2-1) The prediction unit 103 of this embodiment predicts a predicted position FP based on the detection results of the multiple detection units 102. In detail, the prediction unit 103 derives, for example, a movement amount Δx1 in the X-axis direction and a movement amount Δy1 in the Y-axis direction from the coordinates of the multiple objects identified by the detection unit 102. The prediction unit 103 predicts a predicted position FP based on the movement amount Δx1 and the movement amount Δy1. If the position of the object identified by the detection unit 102 matches the predicted position FP or is within the prediction range FA, the determination unit 104 determines that the object is an obstacle, and if the position does not match the predicted position FP and is not within the prediction range FA, the determination unit 104 determines that the object is not an obstacle.

The prediction unit 103 can predict the predicted position FP based on the tendency of the object's movement by predicting the predicted position FP based on the detection results of the multiple detection units 102. On the other hand, even if the prediction unit 103 predicts the predicted position FP based on the tendency of the object's movement, if the object is moving, the object may not move to the predicted position FP but may move to the vicinity of the predicted position FP, contrary to the tendency of the object's movement. The determination unit 104 of this embodiment determines that the object is an obstacle when the position of the object identified by the detection unit 102 is within the prediction range FA from the predicted position FP to a position away from the predicted position FP by a predetermined distance D. As described above, the prediction range FA is, for example, a range in which the object may move against the tendency of the object's movement. As a result, the determination unit 104 can determine whether the object is an obstacle that hinders the progress of the forklift 10 even if the object is moving. Therefore, the obstacle detection device 40 can prevent the travel of the forklift 10 from being impeded by determining that an object that does not impede the progress of the forklift 10 is an obstacle.

The above-described embodiments may be modified as follows: The above-described embodiments and the following examples may be combined with each other as long as they are not technically inconsistent.
In the above description, the determination unit 104 determines that an object exists when the position of the object detected by the detection unit 102 matches the predicted position FP or is within the predicted range FA, but this is not limited to the above. The determination unit 104 may perform a determination of whether or not an object exists multiple times, and may determine that the object is an obstacle when it is determined that the object exists a predetermined number of times or more among the multiple determinations. In this example, a case will be described in which the determination unit 104 determines that the object is an obstacle when it is determined that the object exists two or more times in a row.

In FIG. 8, the prediction unit 103 predicts the predicted position FP based on the detection result at the nth detection timing and the detection result at the n+1th detection timing. In this case, the prediction unit 103 predicts the predicted position FP1 based on the detection results of the objects P1 and P2 detected by the detection unit 102. The determination unit 104 determines that the object is an obstacle because the position of the object P3 detected by the detection unit 102 is within the prediction range FA1 from the predicted position FP1 to a position away from the predicted position FP1 by a predetermined distance D. Next, the prediction unit 103 predicts the predicted position FP based on the detection result at the n+1th detection timing and the detection result at the n+2nd detection timing. In this case, the prediction unit 103 predicts the predicted position FP2 based on the detection results of the objects P2 and P3 detected by the detection unit 102. The determination unit 104 determines that the object P4 detected by the detection unit 102 is an obstacle because the position of the object P4 is within the prediction range FA2, which is a predetermined distance D away from the predicted position FP2.

Here, the position of object P4 is a position within the monitoring range SA. In this example, when an object is present within the monitoring range SA and the object is determined to be an obstacle twice in succession, the determination unit 104 outputs a detection result including information indicating that an object is present within the monitoring range SA to the control device 22. As described above, the determination unit 104 determined that the object was an obstacle both times in succession, when it determined that object P3 was present and when it determined that object P4 was present. Therefore, the determination unit 104 outputs a detection result including information indicating that an object is present within the monitoring range SA to the control device 22.

Here, depending on the situation, an object that does not impede the progress of the forklift 10 may be detected by chance at a position that coincides with the predicted position FP or within the predicted range FA. On the other hand, it is unlikely that an object that does not impede the progress of the forklift 10 will be detected multiple times at a position that coincides with the predicted position FP or within the predicted range FA. With this configuration, the determination unit 104 can determine the presence or absence of an obstacle with higher accuracy than if it were to make a determination only once.

The predetermined number of times that an object is determined to be an obstacle is not limited to two times, and may be three or more times. Furthermore, the determination unit 104 may determine that an object is an obstacle when the total number of times that an object is determined to be an obstacle is equal to or greater than a predetermined number.

In the above, the prediction unit 103 predicts the predicted position FP for one object detected by the detection unit 102, but this is not limited to the above. When there are multiple objects detected by the detection unit 102, the prediction unit 103 may predict the predicted position FP for each of the multiple objects.

In FIG. 9, the detection unit 102 detects an object P1, and at the next detection timing, detects three objects, object P2, object Q1, and object R1. The prediction unit 103 predicts a predicted position FP1 based on the detection result of object P1 and the detection result of object P2. The prediction unit 103 also predicts a predicted position FQ1 based on the detection result of object P1 and the detection result of object Q1. The prediction unit 103 also predicts a predicted position FR1 based on the detection result of object P1 and the detection result of object R1.

The determination unit 104 determines whether or not the object is an obstacle based on the position of the predicted position FP1, a prediction range FA-1 from the predicted position FP1 to a position a predetermined distance D away, and the detection result of the detection unit 102. The determination unit 104 also determines whether or not the object is an obstacle based on the position of the predicted position FQ1, a prediction range FA-2 from the predicted position FQ1 to a position a predetermined distance D away, and the detection result of the detection unit 102. The determination unit 104 also determines whether or not the object is an obstacle based on the position of the predicted position FR1, a prediction range FA-3 from the predicted position FR1 to a position a predetermined distance D away, and the detection result of the detection unit 102.

In FIG. 9, the detection unit 102 detects objects P3, Q2, and R2 at the next detection timing. Furthermore, all of these objects are present within the prediction ranges FA-1 to FA-3. Therefore, the determination unit 104 determines that each of objects P3, Q2, and R2 is an obstacle.

According to this configuration, even if a plurality of objects are detected by the detection unit 102, the determination unit 104 can determine whether each of the objects is an obstacle or not.
In the above description, the predicted range FA is defined as a range from the predicted position to a position that is a predetermined distance D away, but this is not limited to the above. The determination unit 104 may change the width of the predicted range FA according to, for example, the length of a predetermined time interval during which the detection unit 102 performs the detection process. In more detail, the longer the predetermined time interval is, the wider the predicted range FA is set by the determination unit 104.

Here, when an object is moving, the longer the predetermined time interval during which the detection unit 102 performs the detection process, the wider the range in which the object may move. With this configuration, the determination unit 104 can determine the width of the prediction range FA according to the length of the predetermined time interval during which the detection unit 102 performs the detection process, thereby making it possible to determine the presence of an obstacle according to the movement of the object.

In the above, the monitoring range SA is described as being a rectangular range, but this is not limited to this. The range may be of any shape as long as the presence of an obstacle will hinder the forklift 10. For example, the monitoring range SA may be the same shape as the detectable range TA.

〇 The prediction unit 103 has been described as using the coordinates of the predicted position FP obtained by subtracting the movement amount Δx and the movement amount Δy from the coordinates of the object detected in the most recent detection process of the detection unit 102, but this is not limited to the above. For example, the prediction unit 103 may correct the movement amount Δx and the movement amount Δy based on the movement amount of the forklift 10. In detail, the prediction unit 103 calculates the degree of increase in the movement amount of the forklift 10. The degree of increase is, for example, a value obtained by dividing the movement amount of the forklift 10 between the n+1th detection timing and the n+2th detection timing by the movement amount of the forklift 10 between the nth detection timing and the n+1th detection timing. With this configuration, the prediction unit 103 can predict the predicted position FP according to the movement of the object and/or the movement amount of the forklift 10 even when the object is moving and the forklift 10 is not moving at a constant speed.

○The sensor 31 may be a stereo camera, a ToF (Time of Flight) camera, or a millimeter wave radar. The stereo camera has two cameras arranged at a distance from each other, and derives the coordinates of an object using the pixel difference of the same object captured in the images captured by both cameras. When a stereo camera is used as the sensor 31, the obstacle detection device 40 acquires a parallax image from the images captured by the two cameras. The parallax image is a pixel-to-pixel correspondence. The parallax is obtained by comparing the images captured by the two cameras of the stereo camera and deriving the difference in the number of pixels between the images for the same feature point captured in each image. A feature point is a part that can be recognized as a boundary, such as an edge of an object. The feature point can be detected from brightness information, etc. The obstacle detection device 40 derives the coordinates of the feature point in the camera coordinate system from the baseline length of the stereo camera, the focal length of the stereo camera, and the parallax image. The obstacle detection device 40 converts the coordinates of the feature point in the camera coordinate system into coordinates in a coordinate system in real space. A feature point is a point that represents a part of an object. Therefore, it can be said that the coordinates of the feature point in the coordinate system in real space represent the distance and direction to the point that represents a part of the object. When a stereo camera is used as the sensor 31, it can be said that points are detected continuously in the horizontal direction when parallax is continuously associated with pixels arranged in a direction corresponding to the horizontal direction in the parallax image. For example, when imaging is performed so that the horizontal direction and the lateral direction of the image coincide, if parallax is associated with adjacent pixels in the lateral direction of the image, it can be said that points are detected continuously in the horizontal direction.

A ToF camera includes a camera and a light source that emits light, and derives the depth distance for each pixel of an image captured by the camera from the time it takes for the reflected light emitted from the light source to be received. By using a ToF camera, a distance image in which pixels are associated with depth distances can be obtained. As in the case of using a stereo camera, the obstacle detection device 40 can derive the coordinates of a point in a coordinate system in real space. When a ToF camera is used, a point that represents a part of an object is a location where light hits. When a ToF camera is used as the sensor 31, it can be said that points are detected continuously in the horizontal direction when distances are continuously associated with pixels that are lined up in the corresponding horizontal direction in the distance image.

Millimeter wave radar is a sensor that can recognize the surrounding environment by irradiating the surroundings with radio waves in a specific frequency band. When using millimeter wave radar, a point that represents part of an object is the point where the radio waves hit. Millimeter wave radar differs from laser range finders in that it uses radio waves instead of lasers. Millimeter wave radar has the same functions as laser range finders.

The sensor 31 may be a combination of multiple sensors, such as a laser range finder and a stereo camera.
The sensor 31 may be one capable of performing three-dimensional measurement. The three-dimensional measurement is to measure the X-, Y-, and Z-coordinates of a point. The Z-coordinate indicates the position of the point in the vertical direction. Examples of the sensor 31 capable of performing three-dimensional measurement include a laser range finder, a stereo camera, a ToF camera, and a millimeter wave radar that can change the irradiation angle in the vertical direction in addition to the horizontal direction. When a sensor capable of performing three-dimensional measurement is used as the sensor 31, the obstacle detection device 40 may detect a point group in which the number of points detected continuously in the horizontal direction is equal to or greater than a threshold value, and the number of points detected continuously in the vertical direction is equal to or greater than a threshold value, as an object. That is, when the vertical position of a point can be detected, the object may be detected including the vertical position. The threshold value set for the number of points detected continuously in the vertical direction may be the same value as the threshold value set for the number of points detected continuously in the horizontal direction, or may be a different value.

There may be only one sensor 31. In this case, the predetermined range A may be set within the detectable range TA of the sensor 31.
The process performed by the control device 22 after the obstacle detection device 40 detects an obstacle is arbitrary. For example, the control device 22 may notify the driver of the forklift 10 of the presence of an obstacle by emitting a sound or turning on a lamp. If the forklift 10 is an autonomous forklift 10, the control device 22 may generate a route to avoid the obstacle.

The obstacle detection device 40 can be mounted on any moving object. For example, the obstacle detection device 40 may be mounted on an industrial vehicle other than the forklift 10, such as a towing tractor, a passenger vehicle, or an aircraft such as a drone.

10...forklift, 11...vehicle body, 12...drive wheels, 13...steering wheels, 14...loading device, 21...drive mechanism, 22...control device, 31, 31A, 31B, 31C...sensors, 40...obstacle detection device, 100...control unit, 101...acquisition unit, 102...detection unit, 103...prediction unit, 104...determination unit, 200...storage unit, FA, FA1, FA-1, FA2, FA-2, FA-3...prediction range, FP, FP1, FP2, FQ1, FR1...predicted position, SA...monitoring range, TA...detectable range.

Claims (4)

An obstacle detection device mounted on a moving object,
a sensor for detecting distance and direction to a point representing a portion of an object;
An acquisition unit that acquires a detection result from the sensor;
a detection unit that detects a position of the object based on the detection result;
A prediction unit that predicts a future predicted position of the object based on a detection result of the detection unit;
a determination unit that determines whether the object is an obstacle based on the predicted position and a detection result of the detection unit;
Equipped with
when predicting a destination of the moving object, the prediction unit predicts the predicted position based on detection results of a plurality of the detection units;
An obstacle detection device , wherein the determination unit determines that the object is an obstacle when the detection result of the detection unit indicates that the object is present within a predicted range up to a predetermined distance from the predicted position . When predicting a destination of the object that is stopped, the prediction unit predicts the predicted position based on a movement of the moving body;
the determination unit determines that the object is an obstacle when the detection result of the detection unit indicates that the object is present at the predicted position.
The obstacle detection device according to claim 1 . the detection unit performs a process of detecting the object at predetermined time intervals,
The prediction unit widens the width of the prediction range as the predetermined time interval is longer.
3. An obstacle detection device according to claim 1 or 2 . the determination unit determines whether the object is an obstacle a plurality of times, and when the object is determined to be an obstacle a predetermined number of times or more among the plurality of determinations, the determination unit determines the object to be an obstacle.
An obstacle detection device according to any one of claims 1 to 3 .

Images