Detection device, detection method, and program

Abstract

To provide a detection device, detection method and program capable of estimating the trunk portion of a tree stably with high accuracy without being affected by environmental factors, even in an environment where acquisition of position information of a moving body by GNSS or the like is unstable.SOLUTION: A detection device 1 comprises: a radar device 10 that transmits a transmission wave toward the surrounding environment of a moving body 2, receives a reflected wave from an object existing in the surrounding environment of the moving body 2, and periodically acquires point cloud information of the surroundings of the moving body 2 by processing the reflected wave; and a trunk position estimation unit 235 configured to estimate the position of a trunk portion of a tree based on the point cloud information acquired by the radar device 10.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a detection device, a detection method, and a program. [Background technology]

[0002] Conventionally, technologies related to mobile objects that autonomously navigate among rows of trees in orchards, etc., have been known. Particularly in environments where it is unstable to acquire position information of the mobile object using a Global Navigation Satellite System (GNSS) or the like, the traveling stability of the mobile object is improved by measuring the distance between the mobile object and trees and controlling the autonomous traveling of the mobile object so that the mobile object does not come into contact with the trees. For example, Patent Document 1 discloses a mobile object in an environment where GNSS cannot be used, which includes one or more Light Detection and Ranging (LiDAR) sensors that output sensor data indicating the distribution of objects in the environment around the mobile object, a storage device that stores environmental map data indicating the distribution of stems of multiple tree rows, a self-localization estimation device that detects stems of tree rows in the environment around the mobile object based on sensor data repeatedly output from the one or more LiDAR sensors while the mobile object is moving and matches the detected stems of the tree rows with the environmental map data to estimate the position of the mobile object, and a control device that controls the movement of the mobile object according to the estimated position of the mobile object. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] International Publication No. 2022 / 107586 Summary of the Invention [Problem to be solved by the invention]

[0004] LiDAR performs measurements using short-wavelength laser light, and therefore, generally, measuring an object is difficult when there is an obstacle between the object and the LiDAR. In this regard, Patent Document 1 describes that by placing the LiDAR at a position lower than the average height of the tree trunk, more laser beams can be irradiated onto the trunk without being blocked by leaves or branches. However, even when placed at a position lower than the average height, the beam may still be blocked by leaves or branches, or by weeds growing above ground. Furthermore, when using LiDAR, there is also the issue that the laser beam is scattered by environmental factors such as fog and dust, making it difficult to detect tree trunks with sufficient accuracy.

[0005] One object of the present invention is to provide a detection device, a detection method, and a program that can estimate the trunk portion of a tree stably and with high accuracy, without being affected by environmental factors, even in an environment where it is unstable to obtain position information of a moving object using GNSS or the like. [Means for solving the problem]

[0006] In order to achieve the above and other objects, the detection device of the present invention comprises a radar processing unit that transmits a transmission wave toward the surrounding environment of a moving body, receives reflected waves from targets present in the surrounding environment of the moving body, and processes the reflected waves to periodically acquire information on a point cloud around the moving body, and an estimation unit that estimates the position of a trunk portion of a tree based on the information on the point cloud acquired by the radar processing unit.

[0007] The estimation unit may estimate a position of a trunk portion of the tree based on height information of the point cloud or information on intensity of the point cloud acquired by the radar processing unit.

[0008] The detection device may further include a storage unit that stores information about the point cloud, and a filter unit that filters the information about the point cloud stored by the storage unit, and the filter unit removes information about the point cloud of targets other than the tree by filtering the information about the point cloud with respect to at least one of information about the intensity of the point cloud, information about the amplitude of the point cloud, information about the velocity of the point cloud, and information about the three-dimensional position of the point cloud acquired by the radar processing unit, and the estimation unit may estimate the position of the trunk of the tree based on the filtered information about the point cloud.

[0009] The detection device may further include a clustering unit that performs clustering processing on the point cloud information filtered by the filtering unit and classifies the information into one or more clusters.

[0010] The estimation unit may select one of the one or more clusters classified by the clustering unit, and estimate the position of the trunk portion of the tree using information on the number of point clouds and the intensity of the point clouds in the selected one cluster.

[0011] The estimation unit may weight the information of the point cloud using at least height information of the point cloud or information regarding the intensity of the point cloud, calculate a weighted center of gravity for the weighted point cloud information, and estimate the calculated weighted center of gravity to be the position of the trunk portion of the tree.

[0012] The information regarding the intensity of the point cloud is an RCS (Radar Cross Section) that indicates the reflective ability of the target to the incident transmission wave, and the estimation unit may assign a weight to the point cloud information using the height information of the point cloud and the RCS, and may increase the weight assigned to the point cloud information when the RCS is a relatively large value and when the detected height in the height information of the point cloud is a relatively low value.

[0013] The detection device may further include a travel control unit that controls travel of the moving object based on the result of estimation by the estimation unit.

[0014] Another aspect of the present invention is a detection method comprising the steps of: transmitting a transmission wave toward the surrounding environment of a moving body using a detection device having a radar processing unit for detecting targets and an estimation unit for estimating the position of a specific target from the detected targets; receiving reflected waves from targets present in the surrounding environment of the moving body; and periodically acquiring point cloud information around the moving body by processing the reflected waves; and estimating the position of a trunk portion of a tree based on the point cloud information acquired by the radar processing unit.

[0015] Yet another aspect of the present invention is a computer program that causes a computer to execute the steps of periodically acquiring point cloud information around a moving body by transmitting a transmission wave toward the surrounding environment of the moving body, receiving reflected waves from targets in the surrounding environment of the moving body, and processing the reflected waves, and estimating the position of a trunk portion of a tree based on the acquired point cloud information. [Effects of the Invention]

[0016] According to the present invention, it is possible to provide a detection device, a detection method, and a program that can detect the trunk portion of a tree stably and with high accuracy, without being affected by environmental factors, even in an environment where it is unstable to obtain location information of a moving object. [Brief explanation of the drawings]

[0017] [Figure 1] 1 is a diagram showing a schematic configuration of a detection device according to an embodiment of the present invention. [Figure 2A] 1 is a schematic plan view of a mobile object on which a radar device is mounted, as viewed from above; [Figure 2B] FIG. 2 is a block diagram showing the hardware configuration of a computer in the radar device. [Figure 3] 10 is a flowchart illustrating an outline of processing by the detection device according to the present embodiment. [Figure 4]FIG. 10 is a diagram showing the relationship between the estimated trunk of a tree and a no-entry area. [Figure 5] 10 is a flowchart specifically illustrating the operation of radar detection processing. [Figure 6] 10 is a flowchart specifically illustrating the operation of a central part position estimation process. [Figure 7] FIG. 10 is a diagram showing an example of the determination in step S22 and step S23. [Figure 8] FIG. 10 is a diagram showing the relationship between the detected height of a tree and RCS. [Figure 9] 10 shows an example of detecting tree heights using a radar device. [Figure 10] FIG. 10 is a diagram showing the relationship between weight, RCS, and height from the ground. DETAILED DESCRIPTION OF THE INVENTION

[0018] A detection device according to one embodiment of the present invention will be described below. However, the present invention is not limited to the following embodiment. Furthermore, the drawings referred to in the following description merely show the shape, size, and positional relationship in a schematic manner to enable understanding of the contents of the present invention. In other words, the present invention is not limited to only the shape, size, and positional relationship exemplified in each drawing.

[0019] 1 is a diagram showing a schematic configuration of a detection device 1 according to this embodiment. The detection device 1 includes a radar device 10, a specific target determination unit 23, and a travel control unit 24.

[0020] 1 shows only one radar device 10, the detection device 1 may include multiple radar devices. The devices may be connected by wire or via a wireless network. The detection device 1 may include, for example, one or more processors as a control device, and one or more memories, hard disk drives, etc. as storage devices.

[0021] In this embodiment, as will be described later, the detection device 1 uses the radar device 10 to transmit a transmission wave toward the surrounding environment of the moving body 2, receives reflected waves from targets in the surrounding environment of the moving body 2, and processes the reflected waves to periodically acquire point cloud information around the moving body 2, and estimates the position of the trunk of a tree (see reference numeral 3 in FIG. 2A ) based on the point cloud information acquired by the radar device 10. Note that FIG. 2A shows an example in which the radar devices 10 are arranged on both the left and right sides in front of the moving body 2, but the present invention is not limited to this, and for example, only one radar device 10 may be arranged in front of the moving body 2.

[0022] 1, the radar device 10 includes a transmitting antenna 11, a receiving antenna 12, a signal generating unit 13, a transmission control unit 14, an oscillator 15, a switch 16, an ADC (Analog to Digital Converter) 17, a distance calculating unit 18, a speed calculating unit 19, a threshold setting unit 20, a detecting unit 21, and an angle calculating unit 22. The radar device 10 employs an FCM (Fast Chirp Modulation) method, but is not limited to the FCM method and may employ a pulse method.

[0023] The transmitting antenna 11 is composed of one or more antenna elements, and emits a transmission wave S1 oscillated by an oscillator 15 toward a target 40. The receiving antenna 12 is composed of one or more antenna elements, and receives an electromagnetic wave R1 (reflected signal) reflected by the target 40.

[0024] The signal generating unit 13 generates a signal pattern of a periodic transmission signal. The transmission control unit 14 is connected to the signal generating unit 13 and controls the operation of the signal generating unit 13. The oscillation unit 15 is connected to the signal generating unit 13 and oscillates the transmission signal based on the signal pattern from the signal generating unit 13.

[0025] The ADC 17 is connected to the receiving antenna 12 and converts the signal received by the receiving antenna 12 into a digital signal. The oscillator 15 and the ADC 17 can be directly connected by a switch 16. The distance calculation unit 18 calculates the relative distance of the target 40 with respect to the radar device 10 based on the converted digital signal.

[0026] A speed calculation unit 19 calculates the speed of the moving object 2 based on the converted digital signal. A threshold setting unit 20 sets a detection threshold to be used by a detection unit 21. The detection unit 21 detects peaks in the distance direction. An angle calculation unit 22 calculates the angle of arrival (horizontal angle and elevation / depression angle) at the distance detected by the detection unit 21.

[0027] As described above, the radar device 10 according to this embodiment employs the FCM (Fast Chirp Modulation) method as an example, but is not limited to the FCM method and may employ a pulse method. The radar device 10 may be disposed below the moving object 2, above the moving object 2, or both above and below the moving object 2. The moving object 2 equipped with the radar device 10 may be an agricultural machine, a construction machine, or a general automobile.

[0028] The specific target determination unit 23 determines a specific target 40 based on the point cloud acquired by the radar device 10. The specific target determination unit 23 includes a coordinate conversion unit 231, a point cloud accumulation unit 232, a filter unit 233, a clustering unit 234, and a trunk position estimation unit 235.

[0029] The coordinate conversion unit 231 converts the point cloud information obtained by the radar device 10 from the coordinate system of the radar device 10 to the coordinate system of the center of the moving object 2. The point cloud storage unit 232 stores information on the point cloud transformed by the coordinate transformation unit 231.

[0030] The filter unit 233 filters the point cloud information accumulated by the point cloud accumulation unit 232. Specifically, the filter unit 233 filters the point cloud information on at least one of the information on the intensity of the point cloud, the amplitude information of the point cloud, the velocity information of the point cloud, and the three-dimensional position information of the point cloud acquired by the radar device 10, thereby removing the point cloud information of targets other than trees.

[0031] The clustering unit 234 performs clustering processing on the point cloud information filtered by the filter unit 233 .

[0032] The trunk position estimation unit 235 estimates the position of the trunk of a tree based on the information of the point cloud acquired by the radar device 10 and then subjected to filtering and clustering processes. Here, for example, the position of the trunk of a tree is defined by the distance and angle from the center of a coordinate system (a vehicle-body fixed coordinate system) that is set to the center of gravity of the moving body 2, but is not limited to this. The vehicle body coordinate system is a coordinate system that has its origin at the center of gravity of the vehicle body, the X axis representing the direction in which the vehicle body moves forward and backward, the Y axis representing the left-right direction of the vehicle body, and the Z axis representing the direction perpendicular to the X axis and Y axis.

[0033] More specifically, the stem position estimation unit 235 estimates the position of the stem of the tree based on the point cloud information, such as height information of the point cloud or information on the intensity of the point cloud (for example, RCS (Radar Cross Section) described later).

[0034] In addition, the stem position estimation unit 235 selects one of the one or more clusters classified by the clustering unit 234, and estimates the position of the stem of the tree using information on the number of point clouds and the intensity of the point clouds in the selected cluster.

[0035] Furthermore, the stem position estimation unit 235 assigns weights to the point cloud information using at least the height information of the point cloud or the information about the intensity of the point cloud, calculates a weighted center of gravity for the weighted point cloud information, and estimates the calculated weighted center of gravity to be the position of the stem part of the tree.

[0036] The information about the intensity of the point cloud is an RCS (Radar Cross Section) that indicates the reflectivity of the target to the incident transmission wave. The trunk position estimation unit 235 assigns a weight to the point cloud information using, for example, the height information and RCS of the point cloud, and increases the weight assigned to the point cloud information when the RCS is a relatively large value and when the detected height in the height information of the point cloud is a relatively low value.

[0037] The travel control unit 24 controls the travel of the moving object 2 based on the results estimated by the trunk position estimation unit 235. Specifically, the travel control unit 24 determines the travel direction and speed of the moving object 2 based on the estimation results of the trunk portion of the tree and the point cloud information acquired by the radar device 10, and causes the moving object 2 to travel in the determined travel direction and speed. The specific target determination unit 23 and the travel control unit 24 may be realized by an ECU (Electronic Control Unit).

[0038] The storage device 30 is configured with one or more memories, a hard disk drive, etc., and stores various types of information in the detection device 1. The storage device 30 stores, for example, point clouds acquired by the radar device 10, etc.

[0039] The specific target determination unit 23 and the driving control unit 24 implement their functions by programming the functions of each unit and executing the programs on a computer equipped with a processor and a storage device, but each unit may also be configured as individual hardware or software. The programs are stored in the storage device of the computer. This computer also includes a communication unit that performs data communication with a higher-level device, etc., and an input / output unit that constitutes a data input / output interface to the detection device 1.

[0040] 2B is a block diagram showing the hardware configuration of a computer in the radar device 10. As shown in FIG. 2B, the radar device 10 includes a radar device control unit 100, an input / output unit 106, a communication unit 107, and a storage unit 108. The radar device control unit 100 includes a processor 101, a read-only memory (ROM) 102, a random access memory (RAM) 103, a bus 104, and an input / output interface 105. The radar device control unit 100 of the radar device 10 may be a general-purpose personal computer capable of executing various functions by installing various programs therein, or may be a computer incorporated in dedicated hardware.

[0041] The processor 101 performs various calculations and processes. The processor 101 is, for example, a central processing unit (CPU), a micro processing unit (MPU), a system on a chip (SoC), a digital signal processor (DSP), a graphics processing unit (GPU), an application specific integrated circuit (ASIC), a programmable logic device (PLD), or a field-programmable gate array (FPGA). Alternatively, the processor 101 is a combination of two or more of these. Furthermore, the processor 101 may be a combination of these with a hardware accelerator or the like.

[0042] The processor 101, ROM 102, and RAM 103 are connected to one another via a bus 104. The processor 101 executes various processes according to a program recorded in the ROM 102 or a program loaded into the RAM 103. A part or all of the program may be incorporated into the circuitry of the processor 101.

[0043] The bus 104 is also connected to an input / output interface 105. To the input / output interface 105, an input / output unit 106, a communication unit 107, and a storage unit 108 are connected.

[0044] The input / output unit 106 is electrically connected to the input / output interface 105 via a wired or wireless connection. The input / output unit 106 is configured with an input unit such as a keyboard and a mouse, and an output unit such as a display for displaying captured images and a speaker for amplifying audio. The input / output unit 106 may be configured such that the display function and the input function are integrated, such as a touch panel.

[0045] The communication unit 107 is a device that allows the processor 101 to communicate with other devices. The communication unit 107 is, for example, a short-range communication means such as Wi-Fi. The other devices include, for example, the detection device 1 and a higher-level device (not shown). These devices are controlled by the radar device control unit 100 via the communication unit 107. However, these devices may also be controlled by the radar device control unit 100 via a wired connection without via the communication unit 107. The storage unit 108 is, for example, a storage device such as a hard disk drive (HDD) or a solid-state drive (SSD).

[0046] The hardware configuration shown in Fig. 2B is merely an example, and is not limited to this configuration. In addition to configurations consisting of various processing devices, such as a single processor, a multiprocessor, and a multi-core processor, a combination of these processing devices with processing circuits, such as an ASIC (Application Specific Integrated Circuit) and an FPGA (Field-Programmable Gate Array), may be used to realize the functional configuration of a processor. Instead of the radar device 10 having the memory unit 108, a configuration in which the memory unit 108 is separately provided may be used.

[0047] The processing executed by the detection device 1 having the above configuration and functions will be described. Fig. 3 is a flowchart illustrating an example of the outline of the processing by the detection device 1 in this embodiment. Through the processing illustrated in Fig. 3, the mobile object 2 detects a row of trees using the radar device 10, separates them into individual trees, and then estimates the positions 3 of the trunks of the trees based on intensity information and height information of the detected point cloud. The processing by the detection device 1 is started, for example, by starting up the mobile object 2 and supplying power to the detection device 1.

[0048] In step S1, the specific target determination unit 23 acquires radar installation information, such as the position, height, and angle relative to the ground, at which the radar device 10 is installed on the moving body 2, from a storage device (not shown) of the moving body 2. The radar installation information is used in coordinate conversion, which will be described later.

[0049] In step S2, the radar device 10 periodically detects detection points on the target 40 around the moving object 2 and provides the detection information to the specific target determination unit 23 (radar detection process (described later with reference to FIG. 4)). The detection information includes the relative speed between the target 40 and the radar device 10, three-dimensional position information including the height of the target 40 for each detection point, and the reception intensity of the reflected wave. The detection information is stored in a storage device (not shown) of the detection device 1 in chronological order for each frame corresponding to each period.

[0050] In step S3, the coordinate conversion unit 231 acquires information about the point cloud detected by the radar device 10. The point cloud information includes three-dimensional position information (position information in the x, y, and z directions) with the position of the radar device 10 as the origin, angle, amplitude, and velocity. Next, the coordinate conversion unit 231 converts the center coordinates of the point cloud information from the coordinate system of the radar device 10 to the coordinate system of the moving object 2. Specifically, the coordinate conversion unit 231 converts the position information of the point cloud from the coordinate system of the radar device 10 to the coordinate system of the center of the moving object 2, using the radar installation information acquired in step S1.

[0051] Specifically, when a radar device 10 is mounted on the left and right sides of a moving body 2, in order to be able to identify that the targets 40 in the overlapping detection areas detected by each of the two radar devices 10 are the same, the position information of the point clouds detected by each device must be converted into a common coordinate system.

[0052] For example, the position information of the point clouds detected by each of the two radar devices 10 is expressed in a polar coordinate system consisting of distance and azimuth angle, and a common coordinate system is expressed in an orthogonal coordinate system (xy coordinate system) with the origin set at the midpoint between the two radar devices 10. Note that the coordinate systems used by each radar device 10 and the common coordinate system are not limited to this and can be selected arbitrarily.

[0053] The coordinate conversion unit 231 converts the position information of the point clouds detected by each of the two radar devices 10 into a coordinate system common to the two radar devices 10. In the above example, the coordinate conversion unit 231 converts the position information of the point clouds in the polar coordinate systems set by each of the two radar devices 10 into position information of the point clouds in a common Cartesian coordinate system. As a result, the position information of the point clouds detected by each of the two radar devices 10 is converted into position information (x, y) of the point clouds in the Cartesian coordinate system.

[0054] The coordinate conversion unit 231 corrects the position information (x, y) of the converted point group in the Cartesian coordinate system into a coordinate system (Cartesian coordinate system) of the center of the moving object 2 (center of gravity of the moving object 2).

[0055] Furthermore, the coordinate conversion unit 231 adjusts the position information (x, y) of the point cloud using the radar installation information acquired in step S1 described above. As a result, the position information (x, y) of the point cloud in the Cartesian coordinate system becomes a value that takes into consideration the position, height, and angle with respect to the ground at which the radar device 10 is installed on the moving body 2. The above describes the coordinate conversion when one radar device 10 is installed on each of the left and right sides of the moving body 2, but this is not limited to this and the coordinate conversion unit 231 can also be applied to a case where only one radar device 10 is installed on the front side of the moving body 2. The coordinate conversion unit 231 may also correct the position information of the point cloud by taking into consideration the attitude of the moving body 2 using position information of the moving body 2 and IMU (Inertial Measurement Unit) information.

[0056] The point cloud accumulation unit 232 acquires and accumulates three-dimensional position information, intensity information, and velocity information for the point cloud information converted by the coordinate conversion unit 231. The acquired three-dimensional position information, intensity information, and velocity information may be for one frame of point cloud information, or may accumulate for multiple frames. Alternatively, the point cloud accumulation unit 232 may first determine the number of frames of point cloud information, and delete older frames each time a new frame is accumulated so that the number of frames always remains constant. Alternatively, the point cloud accumulation unit 232 may acquire point cloud information while the mobile object 2 is moving, and accumulate the three-dimensional position information, intensity information, and velocity information for the acquired point cloud information.

[0057] In step S4, the filter unit 233 performs filtering on the point cloud information using position information, amplitude information, height information, velocity information, RCS, etc. This filtering process removes point cloud information of objects other than the specific target 40 that is the main detection target of the detection device 1. In this way, the filter unit 233 removes point clouds of other objects using height information and amplitude information, thereby removing noise from the point cloud and enabling highly accurate detection of the target 40. A predetermined value may be set for the amplitude filter, or a relative threshold may be created using the amplitude values ​​of the point clouds of surrounding points.

[0058] Furthermore, since tree trunks are easily detected at positions relatively low from the ground, the filter unit 233 may perform filtering using height information about the point cloud information. Furthermore, the filter unit 233 may filter point cloud information corresponding to relatively low positions using height information about the point cloud information in order to remove the influence of detected weeds, the ground, etc. Furthermore, the filter unit 233 may perform filtering using RCS on the point cloud information in order to remove the influence of pesticide spraying, fog, etc., and the influence of dust.

[0059] In step S5, the clustering unit 234 performs clustering processing on the point cloud information filtered by the filter unit 233 using a known clustering algorithm such as DBSCAN (Density-based spatial clustering of applications with noise) to classify the information into one or more clusters.

[0060] The clustering unit 234 performs clustering processing on the point cloud information using either three-dimensional position information or two-dimensional position information. Furthermore, the clustering unit 234 may perform clustering processing on the point cloud information by adding intensity information and velocity information. Furthermore, the clustering unit 234 may determine parameters for the clustering method based on the range of branch spread or, if the trees are lined up in a row, the spacing between the rows of trees. By performing the above-described clustering processing, the clustering unit 234 can separate the point cloud information for each tree and classify it into one or more clusters.

[0061] In step S6, the stem position estimation unit 235 estimates the position of the stem of the tree based on the information of the clustered point cloud. The process of the stem position estimation unit in step S6 will be described later with reference to FIG.

[0062] In step S7, the stem position estimation unit 235 determines whether one or more tree trunks have been detected in the stem position estimation unit processing. If it is determined that one or more tree trunks have been detected (step S7: YES), the processing proceeds to step S8, and if it is determined that no tree trunks have been detected (step S7: NO), the processing returns to step S2.

[0063] In step S8, the travel control unit 24 determines (updates) the travel direction and speed of the moving object 2 based on the estimation result of the tree trunk portion and the point cloud information acquired by the radar device 10, and then ends the process.

[0064] FIG. 4 is a diagram showing the relationship between the position 5 of the trunk of a tree estimated by the trunk position estimation unit 235 and the no-entry area 7. As shown in FIG. 4, the traveling control unit 24 may, for example, determine the radius of a circle from the two-dimensional coordinate positions of the information on the point cloud estimated to be the trunk, and set the circle with that radius as the no-entry area 7 for other vehicles. Furthermore, the traveling control unit 24 may set the no-entry area 7 for other vehicles using not only the information on the point cloud estimated to be the trunk, but also the information on the point cloud acquired by the radar device 10. In this case, the traveling control unit 24 may perform control such as setting a route that prevents the moving object 2 from entering within a predetermined distance from the position 5 of the trunk, or stopping the moving object 2 if it enters within the predetermined distance.

[0065] Furthermore, when performing work such as spraying pesticides, it may be desirable for the travel control unit 24 to avoid tree trunks and get as close as possible to tree branches and leaves. In this way, the travel control unit 24 may set no-entry areas and tree parts to be avoided depending on the purpose of the work. For example, when the position 5 of the trunk is no longer detected ahead of the mobile object 2, it may be determined that the mobile object 2 has reached the operating range of the mobile object 2, for example, the edge of an orchard, and the mobile object 2 may be controlled to stop.

[0066] Next, the radar detection process will be described. Fig. 5 is a flowchart specifically showing the operation of the radar detection process (step S2) in the processing flow of the detection device 1 illustrated in Fig. 3. In step S11, the signal generation unit 13, the transmission control unit 14, and the oscillator 15 generate a transmission signal of a predetermined frequency and perform a transmission process of emitting a transmission wave from the transmission antenna 11 toward the front of the moving body 2.

[0067] In step S12, the receiving antenna 12 receives, as a received signal, a wave reflected by one or more targets 40 from the transmitted wave, and the ADC 17 performs a receiving process of converting the received signal into a digital signal.

[0068] In step S13, the distance calculation unit 18 performs distance calculation to obtain the relative distance between the radar device 10 and the target 40 based on the digital signal.

[0069] In step S14, the speed calculation unit 19 performs speed calculation based on the digital signal. Specifically, the speed calculation unit 19 calculates relative speed information between one or more targets 40 and the radar device 10 based on the received signal. In this way, the speed calculation unit 19 can detect the speed relative to the targets 40 by speed calculation.

[0070] In step S15, the threshold setting unit 20 sets a detection threshold for detecting peaks to be used by the detection unit 21. Specifically, the threshold setting unit 20 adjusts the detection threshold based on the installation position, height, angle, etc. of the receiving antenna 12.

[0071] In step S16, the detection unit 21 detects a peak in the distance direction using a predetermined method such as CFAR (Constant False Alarm Rate).

[0072] In step S17, the angle calculation unit 22 performs angle calculation processing to determine the arrival angle (horizontal angle and elevation / depression angle) of the reflected wave at the distance detected by the detection unit 21, and calculates the direction in which the target 40 exists in a virtual space with the position of the radar device 10 as the origin and the radiation direction as the reference axis.

[0073] In step S18, the radar device 10 corrects the deviation of the detection position that occurs when the radar device 10 tilts due to the influence of unevenness of the road, etc., by using IMU (Inertial Measurement Unit) information. The radar device 10 supplies the corrected detection information (i.e., point cloud information) to the specific target determination unit 23, and ends the process.

[0074] Next, a description will be given of the process of estimating the position of the main body in the processing of the detection device 1. Fig. 6 is a flowchart specifically showing the operation of the process of estimating the position of the main body (step S6) in the processing flow of the detection device 1 exemplified in Fig. 3 .

[0075] In step S21, the stem position estimation unit 235 selects a point cloud of one cluster (i.e., a cluster corresponding to each tree) from one or more clusters classified by the clustering unit 234, and proceeds to step S22. The selection order is not particularly limited. The stem position estimation unit 235 selects point clouds for all clusters and executes the processes of steps S22 to S27.

[0076] In step S22, the stem position estimation unit 235 determines whether the number of point clouds in one cluster selected in step S21 is equal to or greater than the point cloud number threshold. Here, the point cloud number threshold is a predetermined value, for example, 50. If the number of point clouds is equal to or greater than the point cloud number threshold (step S22: YES), the process proceeds to step S24, and if it is determined that the number of point clouds is less than the point cloud number threshold (step S22: NO), the process proceeds to step S23.

[0077] In step S23, the stem position estimation unit 235 determines whether the point cloud in the one cluster selected in step S21 has points equal to or greater than the RCS threshold. If the point cloud has points equal to or greater than the RCS threshold (step S23: YES), the process proceeds to step S24. If the point cloud has points less than the RCS threshold (step S23: NO), the process proceeds to step S27.

[0078] FIG. 7 is a diagram illustrating an example of the determinations made in steps S22 and S23. As shown in FIG. 7, the process of step S22 determines that if the number of point clouds in a cluster is equal to or greater than the point cloud number threshold, the cluster includes a tree trunk; if the number of point clouds is less than the point cloud number threshold, the cluster includes only tree branches, leaves, weeds, and the like, but does not include a trunk. When only leaves of a tree near the radar device 10 are detected, the stem position estimation unit 235 can exclude detection of the stem of the tree. Also, as shown in FIG. 7, the process of step S23 may cause the stem position estimation unit 235 to detect a tree trunk or an obstacle pillar, etc., for a target relatively far from the radar device 10, with a small number of point clouds and a large RCS. The process of step S23 is a process for detecting the tree trunk in such a case, using an RCS threshold corresponding to the case of detecting a tree trunk or pillar, etc., as a determination criterion.

[0079] 6, in step S24, the central node position estimation unit 235 assigns weights to all point cloud information in the one cluster selected in step S21. Here, the central node position estimation unit 235 assigns weights to the point cloud information using at least the height information or RCS of the point cloud as a weight parameter.

[0080] The trunk position estimation unit 235 may use only one of the weights depending on the surrounding environment and the point cloud information acquired by the radar device 10. Moreover, it is possible to arbitrarily set whether to prioritize either the height information of the point cloud or the RCS, and the trunk position estimation unit 235 may adjust the ratio when adding up the weights corresponding to the height information of the point cloud and the RCS depending on the surrounding environment.

[0081] Fig. 8 is a diagram showing the relationship between the detected height of a tree based on height information of a point cloud and RCS. Fig. 9 is a diagram showing an example of detection of the height of a tree by a radar device 10. As shown in Fig. 9, a transmission wave transmitted from the radar device 10 is transmitted to targets such as a trunk 3 of a tree, leaves and branches 4 of the tree, the ground 8, and weeds 9, and reflected waves from these targets are received.

[0082] Therefore, the detection height and RCS for each part of a tree have characteristics as shown in Figure 8. Specifically, the trunk of the tree has a large RCS and a low detection height, while the branch of the tree has a medium RCS and a medium to high detection height. The leaf of the tree has a small RCS and a medium to high detection height, while the weed part of the tree has a small RCS and a low detection height. The stem position estimation unit 235 may take into account the relationship between the detection height and RCS of the tree as shown in Figure 8 and set a large weight when the RCS is large and the detection height is low, in order to adopt a large proportion of the detection of the trunk.

[0083] Furthermore, the stem position estimation unit 235 may assign a larger weight as the RCS value increases. Regarding the detection height, the stem position estimation unit 235 may set the largest weight at a certain height from the ground because weeds and the like may be detected if the detection height is too close to the ground, and the specific height may be adjusted as a parameter according to the surrounding environment. Furthermore, the stem position estimation unit 235 may adjust the locations where the weight is set large and the locations where the weight is set small, and the gradient of the final weight as a parameter. Furthermore, the height information used may be an absolute height from the ground, or a relative height within one cluster.

[0084] Fig. 10 is a diagram showing an example of the relationship between the weight, RCS, and height from the ground determined by the above-mentioned processing. In the example shown in Fig. 10, if the detection height is too close to the ground, there will be weeds as shown in Fig. 9, so the stem position estimation unit 235 sets the weight so that the weight becomes large when the detection height is about 0.5 m from the ground.

[0085] 6, in step S25, the trunk position estimation unit 235 calculates a weighted center of gravity for the information of the point cloud to which weights have been assigned in step S24. The trunk position estimation unit 235 obtains the weighted center of gravity for the information of the point cloud to which weights have been assigned using either two-dimensional coordinates or three-dimensional coordinates. For example, when using three-dimensional coordinates, the weighted centroid can be found using the following formula: x-coordinate of weighted centroid = (x1*w1+x2*w2+x3*w3+...+xn*wn) / (w1+w2+w3+...+wn) y coordinate of weighted centroid = (y1*w1+y2*w2+y3*w3+...+yn*wn) / (w1+w2+w3+...+wn) Weighted center of gravity z coordinate = (z1*w1+z2*w2+z3*w3+...+zn*wn) / (w1+w2+w3+...+wn) Here, (x1, y1, z1), (x2, y2, z2), ..., (xn, yn, zn) represent the coordinates of each point, and w1, w2, ..., wn represent the weights of each point.

[0086] In step S26, the stem position estimation unit 235 detects the weighted center of gravity calculated in step S25 as the position of the stem of one tree corresponding to one cluster selected in step S21, and then ends the process.

[0087] In step S27, the stem position estimation unit 235 does not detect the one cluster selected in step S21 as the position of the stem of the tree because it was determined in step S23 that the point cloud has points less than the RCS threshold, and then terminates the processing. As described above, the trunk position estimation unit 235 selects point clouds for all clusters in step S21, executes the processes of steps S22 to S27, and ends all processes after processing all clusters.

[0088] According to the embodiment described above, the following effects are achieved. The detection device 1 of this embodiment includes a radar device 10 that transmits a transmission wave toward the surrounding environment of the moving body 2, receives reflected waves from targets in the surrounding environment of the moving body 2, and processes the reflected waves to periodically acquire information on a point cloud around the moving body 2, and a trunk position estimation unit 235 that estimates the position of the trunk portion of a tree based on the information on the point cloud acquired by the radar device 10.

[0089] With this configuration, the detection device 1 can detect the trunk of a tree without being affected by parts such as branches and leaves of the tree, enabling stable travel of the mobile object 2. Furthermore, the detection device 1 can estimate the position of the trunk of a tree that is located behind branches and leaves by using the point cloud information acquired by the radar device 10. Therefore, the detection device 1 can estimate the trunk of a tree stably and with high accuracy without being affected by environmental factors, even in an environment where acquisition of position information of the mobile object 2 by GNSS or the like is unstable.

[0090] Furthermore, the trunk position estimation unit 235 estimates the position of the trunk of the tree based on the height information of the point cloud or the information on the intensity of the point cloud (e.g., RCS) acquired by the radar device 10. This allows the detection device 1 to estimate the trunk of the tree using the height information of the point cloud or the information on the intensity of the point cloud without being influenced by parts such as branches and leaves of the tree.

[0091] The detection device 1 further includes a point cloud storage unit 232 that stores point cloud information, and a filter unit 233 that filters the point cloud information stored by the point cloud storage unit 232. The filter unit 233 filters the point cloud information with respect to at least one of the information on the intensity of the point cloud, the amplitude information of the point cloud, the velocity information of the point cloud, and the three-dimensional position information of the point cloud acquired by the radar device 10, thereby removing point cloud information of targets other than trees. The stem position estimation unit 235 estimates the position of the stem of the tree based on the filtered point cloud information. With this configuration, the detection device 1 can estimate the stem of the tree with high accuracy using the point cloud information from which the point cloud information of targets other than trees has been removed by filtering.

[0092] The detection device 1 further includes a clustering unit 234 that performs clustering processing on the point cloud information filtered by the filter unit 233 and classifies the information into one or more clusters. With this configuration, the detection device 1 can separate the point cloud information for each tree through clustering processing and classify it into one or more clusters, thereby enabling the tree trunk to be estimated with high accuracy.

[0093] Furthermore, the stem position estimation unit 235 selects one of the one or more clusters classified by the clustering unit 234, and estimates the position of the trunk of the tree using information (e.g., RCS) relating to the number of point clouds and the intensity of the point cloud in the selected cluster. In this way, by using the number of point clouds, the detection device 1 can eliminate the detection of the trunk of a tree when only the leaves of the tree near the radar device 10 are detected, and by using the information relating to the intensity of the point cloud, it can estimate the trunk of the tree with high accuracy.

[0094] Furthermore, the stem position estimation unit 235 weights the point cloud information using at least the height information of the point cloud or the information on the intensity of the point cloud (for example, RCS), calculates a weighted center of gravity for the weighted point cloud information, and estimates the calculated weighted center of gravity as the position of the tree trunk. In this way, the detection device 1 can estimate the tree trunk with high accuracy by weighting the point cloud information that is likely to be the position of the tree trunk.

[0095] Furthermore, the information relating to the intensity of the point cloud is an RCS (Radar Cross Section) that indicates the reflectivity of the target 40 to the incident transmission wave, and the stem position estimation unit 235 assigns a weight to the point cloud information using the height information and RCS of the point cloud, and increases the weight assigned to the point cloud information when the RCS is a relatively large value and when the detected height in the height information of the point cloud is a relatively low value. In this way, the detection device 1 can reliably estimate the trunk of a tree with high accuracy by increasing the weight assigned to the information of a point cloud that is likely to be the position of the trunk of a tree.

[0096] The detection device 1 further includes a travel control unit 24 that controls the travel of the moving object 2 based on the results of estimation by the stem position estimation unit 235. With this configuration, the detection device 1 can, for example, determine the travel direction and speed of the moving object 2 from the results of estimation of the tree stem and point cloud information, and set a no-entry area for the moving object 2 from the two-dimensional coordinate position estimated to be the tree stem, thereby realizing stable travel of the moving object 2 even in an environment such as an orchard where acquisition of position information of the moving object 2 by GNSS or the like is unstable.

[0097] In the above-described embodiment, a case has been described in which the moving body 2 travels between rows of trees where the spacing between the trees is predetermined, but the present invention is not limited to this embodiment and can also be applied to cases in which the spacing between the trees is not predetermined.

[0098] Although some of the embodiments of the present application have been described in detail above with reference to the drawings, these are merely examples, and the present invention can be implemented in other forms that include the forms described in this specification and that have been modified and improved in various ways based on the knowledge of those skilled in the art. [Explanation of symbols]

[0099] 1. Detection device 2. Mobile 10 Radar equipment 11 Transmitting Antenna 12 receiving antenna 13 Signal generation unit 14 Transmission control section 15 Oscillator 16 Switch 17 ADC 18 Distance calculation section 19 Speed ​​calculation section 20 Threshold setting unit 21 Detection unit 22 Angle calculation unit 23 Specific target determination unit 24 Travel control unit 30 Storage device 231 Coordinate conversion unit 232 Point Cloud Storage Unit 233 Filter section 234 Clustering Department 235 Executive position estimation department

Claims

1. a radar processing unit that transmits a transmission wave toward the surrounding environment of the moving object, receives reflected waves from targets present in the surrounding environment of the moving object, and processes the reflected waves to periodically acquire information about a point cloud around the moving object; an estimation unit that estimates the position of a trunk of a tree based on the point cloud information acquired by the radar processing unit; A detection device comprising:

2. The detection device according to claim 1 , wherein the estimation unit estimates the position of the trunk of the tree based on height information of the point cloud or information on intensity of the point cloud acquired by the radar processing unit.

3. a storage unit that stores information about the point cloud; a filter unit that filters the point cloud information stored by the storage unit; Further provided with the filter unit filters the point cloud information with respect to at least one of information on the intensity of the point cloud, information on the amplitude of the point cloud, information on the velocity of the point cloud, and information on the three-dimensional position of the point cloud, all of which are acquired by the radar processing unit, to remove information on the point cloud of targets other than the trees; the estimation unit estimates a position of a trunk portion of the tree based on information of the filtered point cloud.

3. The detection device according to claim 1 or 2.

4. The detection device according to claim 3 , further comprising a clustering unit that performs clustering processing on the point cloud information filtered by the filtering unit and classifies the information into one or more clusters.

5. The estimation unit selecting one cluster from the one or more clusters classified by the clustering unit; The detection device according to claim 4 , wherein the position of the trunk of the tree is estimated using information on the number of points and the intensity of the points in the selected cluster.

6. The estimation unit weighting the information of the point cloud using at least height information of the point cloud or information about intensity of the point cloud; Calculating a weighted center of gravity for the weighted point cloud information; The detection device according to claim 5 , wherein the calculated weighted center of gravity is estimated to be the position of a trunk portion of the tree.

7. The information about the intensity of the point cloud is a radar cross section (RCS) that indicates the reflectivity of the target with respect to the incident transmission wave, 7. The detection device according to claim 6, wherein the estimation unit assigns a weight to the point cloud information using the height information of the point cloud and the RCS, and increases the weight assigned to the point cloud information when the RCS has a relatively large value and when a detected height in the height information of the point cloud has a relatively low value.

8. The detection device according to claim 1 or 2, further comprising a travel control unit that controls travel of the moving object based on the result of estimation by the estimation unit.

9. A detection device including a radar processing unit for detecting a target and an estimation unit for estimating the position of a specific target from the detected target, a step of periodically acquiring point cloud information around the moving body by transmitting a transmission wave toward the surrounding environment of the moving body, receiving a reflected wave from a target present in the surrounding environment of the moving body, and processing the reflected wave; estimating the position of a trunk of a tree based on the point cloud information acquired by the radar processing unit; A detection method comprising:

10. On the computer, a step of periodically acquiring point cloud information around the moving body by transmitting a transmission wave toward the surrounding environment of the moving body, receiving a reflected wave from a target present in the surrounding environment of the moving body, and processing the reflected wave; a step of estimating the position of a trunk portion of a tree based on the acquired point cloud information; A computer program that executes the following:

Images