WO2024018337A1

WO2024018337A1 - Object detection device, object detection method, and computer program

Info

Publication number: WO2024018337A1
Application number: PCT/IB2023/057175
Authority: WO
Inventors: 明宇李
Original assignee: ロベルト·ボッシュ·ゲゼルシャフト·ミト•ベシュレンクテル·ハフツング
Priority date: 2022-07-22
Filing date: 2023-07-13
Publication date: 2024-01-25

Abstract

Provided are an object detection device, an object detection method, and a computer program that can improve the accuracy of moving body detection using a LiDAR. An object detection device (100) detects an object on the basis of measurement point data acquired by a LiDAR (11). The object detection device: extracts, from primary measurement point data that is measurement point data acquired by the LiDAR (11), a measurement point group for which the distance between neighboring measurement points is within a prescribed distance; excludes, from the primary measurement point data, a non-moving-body measurement point group that is, among the extracted measurement point groups, a measurement point group with the length in any direction being at least a prescribed length threshold value; extracts dynamic measurement points, which are measurement points obtained by excluding static measurement points that are measurement points overlapping with the primary measurement point data used in a previous processing cycle, from secondary measurement point data that is measurement point data obtained by excluding the non-moving-body measurement point group from the primary measurement point data; extracts, from the secondary measurement point data, a measurement point group for which the distance between neighboring measurement points is within a prescribed distance; identifies a first moving body measurement point group that comprises the dynamic measurement points and a second moving body measurement point group that comprises the dynamic measurement points and the static measurement points; and, through clustering processing, associates the identified first moving body measurement point group and second moving body measurement point group with moving body data for which detection was completed in or prior to the previous processing cycle.

Description

[Document name] Statement

[Name of the invention] Object detection device, object detection method, and computer program

【Technical field】

[. [001] The present invention relates to an object detection device, an object detection method, and a computer program that are applied to a moving object such as a vehicle.

[Background technology]

[. [002] In recent years, moving objects such as vehicles are equipped with object detection devices that use distance measuring sensors to detect the surrounding environment of the moving object. For example, in a vehicle, cruise control (AC C) is used to automatically drive the vehicle while maintaining the target distance to the vehicle in front based on detected information about the surrounding environment. Collision safety functions are implemented to avoid collisions with obstacles such as vehicles in front or to reduce the impact of a collision. Li DAR (L i DAR), which uses optical waves such as lasers and infrared rays, is known as one type of distance measurement sensor used in such object detection devices.

[Prior art documents]

[Patent document]

【〇 0 0 3】

[Patent Document 1] Japanese Patent Application Laid-Open No. 2021-163213

[Summary of the invention]

[Problem to be solved by the invention]

[〇 0 0 4] However, although L i DAR can measure the distance and direction to the object to be detected, it cannot measure the speed of the object to be detected like a radar sensor, which is also known as a ranging sensor. cannot be measured. For this reason, there is a risk that irregularities on the road surface or installed objects such as guardrails may be mistakenly detected as moving objects, or that tracking of moving objects may be interrupted.

[〇 0 0 5] An object of the present invention is to provide an object detection device, an object detection method, and a computer program that can improve the accuracy of detecting a moving object using LiDAR.

[Means to solve the problem]

[〇 0 0 6] In order to solve the above problems, according to one aspect of the present invention,

In an object detection device (1 〇〇) that detects an object based on the data of the measurement point acquired by L i DAR ( 1 1 ), the data of the measurement point acquired by L i DAR ( 1 1 ) is From certain primary measurement point data, extract a measurement point group in which the distance between adjacent measurement points is within a predetermined distance, and among the extracted measurement point group, the length in any direction is a predetermined length. a filtering process that removes the non-moving body measurement point group from the primary measurement point data by treating the measurement point group having a value equal to or higher as the non-moving body measurement point group; and filtering the non-moving body measurement point group from the primary measurement point data. Among the secondary measurement point data, which is the data of the measurement points excluding the static measurement points, the measurement points excluding the static measurement points, which are the measurement points that overlap with the primary measurement point data used in the previous processing cycle, are called dynamic measurement points. A dynamic measurement point extraction process is performed to extract a dynamic measurement point from the secondary measurement point data, and the distance between adjacent measurement points is within the predetermined distance. a first moving body measurement point group; a second moving body measurement point group constituted by the dynamic measurement points and the static measurement points; a clustering process for identifying the first moving body measurement point group; a map for associating the first moving object measurement point group and the second moving object measurement point group with moving object data detected by the previous processing cycle; An object detection device is provided that performs the etching process.

[ 0 0 0 7 ] Furthermore, in order to solve the above problems, according to another aspect of the present invention,

In an object detection method that detects an object based on measurement point data acquired by L i D A R ( 1 1 ), primary measurement point data that is data of measurement points acquired by L i D A R ( 1 1 ) is , extracting a group of measurement points in which the distance between adjacent measurement points is within a predetermined distance, and among the extracted measurement point group, a measurement point whose length in any direction is equal to or greater than a predetermined length value. removing the non-moving object measurement point group from the primary measurement point data by setting the point group as a non-moving object measurement point group (step S11); Among the secondary measurement point data, which is the data of the measurement points excluding the static measurement points, the measurement points excluding the static measurement points, which are the measurement points that overlap with the primary measurement point data used in the previous processing cycle, are called dynamic measurement points. (Step S15), from the secondary measurement point data, extract a group of measurement points in which the distance between adjacent measurement points is within the predetermined distance, and a clustering process (step S17) for identifying: a first moving body measurement point group consisting of the dynamic measurement points and the static measurement points; a matching process (step S 1 9 to step S 2 9 ); and an object detection method is provided.

[ 0 0 0 8 ] Furthermore, in order to solve the above problems, according to another aspect of the present invention, the computer detects an object based on data of measurement points acquired by L i D A R ( 1 1 ). A computer program that executes a process of determining, from primary measurement point data that is measurement point data acquired by the L i D A R ( 1 1 ), a measurement in which the distance between adjacent measurement points is within a predetermined distance. While extracting a point group, among the extracted measurement point groups, a measurement point group whose length in any direction is equal to or greater than a predetermined length value is defined as a non-moving object measurement point group from the primary measurement point data. Excluding the non-moving object measurement point group (step S11); and out of the secondary measurement point data that is measurement point data obtained by excluding the non-moving object measurement point group from the primary measurement point data, extracting measurement points excluding static measurement points that overlap with the primary measurement point data used in the previous processing cycle as dynamic measurement points (step S15); From the data, a group of measurement points in which the distance between adjacent measurement points is within the predetermined distance is extracted, and a first mobile measurement point group consisting of the dynamic measurement points, the dynamic measurement points, and a clustering process (step S17) for identifying a second moving body measurement point group constituted by the static measurement points; 2. A computer that executes a matching process (steps S 1 9 to S 2 9) that associates the measurement point group with moving object data detected up to the previous processing cycle, and a process that includes . program will be provided.

【Effect of the invention】

[0009] As explained above, according to the present invention, it is possible to improve the accuracy of detecting a moving object using LiDAR.

[Brief explanation of drawings]

[ 0 0 1 0 ] FIG. 1 is a schematic diagram showing a configuration example of a vehicle as an example of a moving body according to an embodiment of the present invention.

FIG. 2 is a block diagram showing a configuration example of an object detection device according to the same embodiment.

[Figure 3] A flowchart showing processing operations by the object detection device according to the embodiment.

[Fig. 4] Fig. 4 is a flowchart showing a moving object detection processing operation by the object detection device according to the same embodiment.

FIG. 5 is an explanatory diagram showing filtering processing by the object detection device according to the embodiment.

FIG. 6 is an explanatory diagram showing filtering processing by the object detection device according to the same embodiment.

FIG. 7 is an explanatory diagram showing dynamic measurement point extraction processing and clustering processing by the object detection device according to the embodiment.

FIG. 8 is an explanatory diagram showing dynamic measurement point extraction processing and clustering processing by the object detection device according to the same embodiment.

FIG. 9 is an explanatory diagram showing matching processing by the object detection device according to the embodiment.

[Form for carrying out the invention]

[ 0 0 1 1 ] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In addition, in this specification and the drawings, the same reference numerals are given to the constituent elements having substantially the same functional configuration to omit redundant explanation.

[ 0 0 1 2 ]

<1. Configuration example of moving body> First, a configuration example of a moving body to which the object detection device according to the present embodiment can be applied will be described.

[ 0 0 1 3 ] The object detection device according to this embodiment can be applied to various moving bodies such as ships, aircraft, and robots, as well as vehicles such as four-wheeled vehicles and motorcycles. In this embodiment, an example in which an object detection device is applied to a four-wheeled vehicle as a moving object will be described.

[ 0 0 1 4 ] FIG. 1 is a schematic diagram showing a configuration example of a vehicle 1 as an example of a moving object. The vehicle 1 shown in Fig. 1 is configured as a two-wheel drive vehicle 1 that transmits drive torque output from a drive power source 2〇, such as an internal combustion engine or a drive motor, to the left and right front wheels. The vehicle 1 may be a four-wheel drive vehicle that transmits driving force to front, rear, left and right wheels. In addition, when vehicle 1 is an electric vehicle or a hybrid electric vehicle, vehicle 1 includes a secondary battery that stores power supplied to the drive motor, and a battery that stores power supplied to the drive motor and charges the battery. It is equipped with a generator that generates the electric power generated by the vehicle, an inverter device that controls the driving of the drive motor, etc.

[ 0 0 1 5 ] The vehicle 1 includes a driving force source 20, an electric steering device 25, and a brake fluid pressure control unit 30 as devices for controlling the running of the vehicle 1. The driving power source 20 outputs driving torque that is transmitted to the wheels via a transmission or differential mechanism (not shown). The drive power source 2〇 and the transmission are controlled by the vehicle control device 4〇.

[ 0 0 1 6 ] The electric steering device 2 5 includes an electric motor and a gear mechanism (not shown), and is controlled by the vehicle control device 4 〇 to adjust the steering angle of the left and right front wheels. During manual driving, the vehicle control device 4 〇 controls the electric steering device 2 5 based on the steering angle of the steering wheel by the driver. Furthermore, during automatic driving, the vehicle control device 40 controls the electric steering device 25 based on the target steering angle.

[ 0 0 1 7 ] The brake fluid pressure control unit 30 adjusts the hydraulic pressure supplied to the brake caliper provided on each wheel to generate braking force. The brake fluid pressure control unit 30 is driven by the car Controlled by both control devices 4 〇. When the vehicle 1 is an electric vehicle or a hybrid electric vehicle, the brake fluid pressure control unit 30 is used in combination with regenerative braking by the drive motor.

[ 0 0 1 8 ] The vehicle control device 40 includes one or more electronic control devices that control the driving of the driving power source 20, the electric steering device 25, and the brake fluid pressure control unit 30. The vehicle control device 40 is configured to be able to acquire a signal transmitted from the object detection device 50, and is configured to be able to execute automatic driving control of the vehicle 1. Furthermore, automatic driving control shall include emergency brake control and A C C (Ada p t i v e C r u i se Cont r 〇!). Furthermore, during manual operation of the vehicle 1, the vehicle control device 40 acquires information on the driver's driving operation amount, and controls the driving force source 20, the electric steering device 25, and the brake fluid pressure control unit 30. control the drive

〇

The object detection processing unit 63 acquires the data (primary measurement point data) mpts of the measurement points detected by L i DAR 1 1, and performs filtering processing to exclude non-moving body measurement points from the primary measurement point data. Execute (Step S 1 1). Specifically, the object detection processing unit 63 extracts one or more measurement point groups in which the distance between adjacent measurement points is within a predetermined distance from the primary measurement point data mpts. For example, the object detection processing unit 63 calculates the distance between each measurement point converted to the coordinates of the fixed coordinate system, and selects a plurality of distances between adjacent measurement points within a predetermined distance. Extract a measurement point group that combines measurement points. Typically, Euclidean distance may be used as the distance, but other distances may also be used.

[ 0 0 4 0 ] Furthermore, the object detection processing unit 63 selects a measurement point group whose length in any direction is equal to or greater than a predetermined length value from among the extracted one or more measurement point groups. Exclude from the primary measurement point data as a non-moving object measurement point group. For example, the object detection processing unit 63 calculates the maximum length along an appropriate direction for each of the extracted one or more measurement point groups. The appropriate direction may be along any direction on the fixed coordinate system. Then, the object detection processing unit 63 selects a group of measurement points whose calculated maximum length is equal to or greater than a predetermined length value, which is set to a value exceeding the maximum length assumed for a moving object, as non-moving object measurement points. Exclude as a group from the primary measurement point cloud data. The predetermined length value may be a value that exceeds the maximum length expected for a detected vehicle, such as a bus or a truck, but may be arbitrarily set depending on the size of the moving object to be detected. The predetermined length value may be set to a value within a range of 10 to 12 m, for example.

[ 0 0 4 1 ] FIGS. 5 to 6 are explanatory diagrams showing the filtering process in step S11. FIG. 5 shows a two-dimensional image of primary measurement point data detected when a plurality of LiDARs are provided on the front, rear, left, and right sides of the vehicle 1. The object detection processing unit 63 extracts a measurement point group that combines a plurality of measurement points in which the distance between adjacent measurement points is within a predetermined distance from the primary measurement point data, and extracts a measurement point group in which the distance between adjacent measurement points is within a predetermined distance. Measurement point groups with a value greater than or equal to the value are excluded from the primary measurement point data as non-moving object measurement point groups. Figure 6 shows measurement point data (secondary measurement point data) after excluding the non-moving object measurement point group from the primary measurement point data.

[ 0 0 4 2 ] The non-moving object measurement points excluded in step s i 1 are considered to be road surface irregularities such as curbs or roadside objects such as guardrails, and are not used as measurement points for detecting moving objects. . Therefore, the object detection processing unit 63 labels and records the non-moving object measurement points excluded in step S11 as road surface irregularities or road-installed objects (step S13).

[ 0 0 4 3 ]

(Dynamic measurement point extraction process) The object detection processing unit 63 executes a dynamic measurement point extraction process on the secondary measurement point data excluding the non-first moving object measurement point group in step S11. (Step S 1 5). For example, the object detection processing unit 63 detects data acquired from L i D A R 1 1 in the previous processing cycle among the secondary measurement point data, which is measurement point data obtained by excluding non-moving object measurement points from the primary measurement point data. Static measurement points that overlap with the primary measurement point data are excluded, and the remaining measurement points are extracted as dynamic measurement points. More specifically, the object detection processing unit 63 uses the secondary measurement point data and the primary measurement point data acquired in the previous processing cycle to calculate the position from the previous processing cycle to the current processing cycle. Identify and exclude static measurement points that do not change. The object detection processing unit 63 extracts, from the secondary measurement point data, measurement points whose positions have changed from the previous processing cycle to the current processing cycle as dynamic measurement points.

[ 0 0 4 4 ]

(Clustering Process) Next, the object detection processing unit 63 extracts a group of measurement points in which the distance between adjacent measurement points is within a predetermined distance from the secondary measurement point data, and extracts a group of measurement points in which the distance between adjacent measurement points is within a predetermined distance. Clustering processing is performed to identify a first moving object measurement point group consisting of dynamic measurement points and a static measurement point, and a second moving object measurement point group consisting of dynamic measurement points and static measurement points. Specifically, the object detection processing unit 63 performs secondary measurement. From the fixed point data, one or more measurement point groups in which the distance between adjacent measurement points is within a predetermined distance are extracted. The predetermined distance may be the same as the distance used in the filtering in step s11. However, the predetermined distance may be different from the distance used in the filtering in step S11. The object detection processing unit 63 sets the measurement point group consisting of the dynamic measurement points extracted in step S15 as the first moving object measurement point group among the extracted measurement point groups, and divides the dynamic measurement points and static measurement points into a first moving object measurement point group. A measurement point group composed of the target measurement point group is specified as a second moving object measurement point group.

[ 0 0 4 5 ] In other words, the object detection processing unit 6 3 uses the measurement point group extracted from the secondary measurement point data in the dynamic measurement point extraction process ( S 1 5 ) and the clustering process ( S 1 7 ). is classified into a first moving object measurement point group whose position has changed significantly since the previous processing cycle, and a second moving object measurement point group whose position has changed less.

[0046] FIGS. 7 to 8 are explanatory diagrams showing the dynamic measurement point extraction process (S15) and the clustering process (S17). Each black circle indicates the measurement point group extracted in the current processing cycle, and the white circle indicates the primary measurement point data acquired in the previous processing cycle. In the example shown in Figure 7, the measurement point group extracted from the secondary measurement point data does not include static measurement points that overlap with the primary measurement point data of the previous processing cycle. Therefore, the measurement point group is classified into the first moving object measurement point group composed of dynamic measurement points. On the other hand, in the example shown in Figure 8, the measurement point group extracted from the secondary measurement point data is a static measurement point (measurement point within the dashed line H 1) that overlaps with the primary measurement point data of the previous processing cycle. , The measurement points excluding the static measurement points are composed of the dynamic measurement points (the measurement points indicated by the broken line H 2). Therefore, the measurement point group is classified as a second mobile measurement point group, which is composed of dynamic measurement points and static measurement points.

[ 0 0 4 7 ]

(Matching Process) Next, the object detection processing unit 63 matches the first moving object measurement point group and the second moving object measurement point group identified by the clustering process with the moving objects detected by the previous processing cycle. Matching processing for association is executed (steps S19 to S29). In this embodiment, the matching process includes a first matching process (steps S19 to S23) that is a simple process, and a first mobile measurement point that could not be associated by the first matching process. A second matching process (steps S25 to S29) for associating groups is included.

[ 0 0 4 8 ]

"Associating the first moving object measurement point group and the second moving object measurement point group with the moving objects detected by the previous processing cycle" means that the first moving object measurement point group and the second moving object measurement point group This indicates that the measurement point group where the same moving object as the data of the detected moving object is detected is to be identified and recorded from among the group. For example, if data of detected moving objects labeled with identification information to identify individual moving objects was recorded along with information estimating the predetermined moving speed and direction of movement up to the previous processing cycle, the current Among the first moving object measurement point group and the second moving object measurement point group identified in the processing cycle, the measurement point group that detected the same moving object as the detected moving object data is identified and labeled with the same identification information. The process of recording the information along with the moving speed and moving direction information is executed.

[ 0 0 4 9 ] The first matching process is a matching process that can be executed more easily (in a shorter time) than the second matching process described later, and can be easily and reliably performed without performing the second matching process. High Executed for the purpose of completing the association for the first moving object measurement point group and the second moving object measurement point group that can be associated.

[ 0 0 5 0 ] The first matching process may be performed using existing matching technology. For example, the object detection processing unit 63 collects the first moving object measurement point group, the second moving object measurement point group, and the moving object data that has been labeled as a moving object by the previous processing cycle (detected moving object data). ) to search for the first moving object measurement point group and the second moving object measurement point group that can be associated with the detected moving object data.

The distance D obtained for the moving data, the velocity component of each measurement point group V x _ c 1 ( n ), V y — c 1 ( n ), and the velocity component of each detected moving object data V x — m〇 ( m ), V ym 〇 ( m ) are input to the machine learning model 70 to obtain a group of measurement points and detected moving object data that can be correlated with each other. The machine learning model 70 learns the movement patterns of moving objects using data from measurement point clouds for learning that have been collected or generated in advance for various types of moving objects moving at various speeds, turning speeds, and directions. This is a model that has Typically, a support vector machine may be used as the machine learning model, but other machine learning models such as random forest or dual network may also be used.

[ 0 0 5 7 ] As a result, detected moving object data that is associated with the first moving object measurement point group and the second moving object measurement point group that were not associated by the first matching process is obtained. The object detection processing unit 63 converts the first moving object measurement point group and the second moving object measurement point group, which have associated detected moving object data, into detected moving object data along with information on speed, moving direction, and orientation. Correlate and record (step S 2 7).

[ 0 0 5 8 ] On the other hand, the object detection processing unit 63 treats the first moving object measurement point group and the second moving object measurement point group for which there is no associated detected moving object data as new moving objects. Detected moving object data is generated and recorded (step S29). When newly generating detected moving object data, the object detection processing unit 63 generates the first moving object measurement point group and the second moving object measurement point group that could not be associated with the detected moving object data in each processing cycle. New detected moving object data may be generated using . Specifically, the object detection processing unit 63 calculates the distance between the first moving object measurement point group and the second moving object measurement point group between each processing cycle (between frames), the moving speed in each processing cycle, and the like. Identify the first moving object measurement point group and the second moving object measurement point group of multiple frames that are estimated to be the measurement point group of the same object based on the moving speed, and identify the first moving object measurement point group and the second moving object measurement point group. 2 Calculates the position, speed, and size (length and width) of the object in each processing cycle of the moving object measurement point group. The newly generated detected moving object data will be used in the matching process from the next processing cycle onward.

[ 0 0 5 9 ] The object detection processing unit 6 3 repeats the processing from step S 1 1 to step S 2 9 in each processing cycle, and extracts the data extracted from the measurement point data detected by L i D A R 1 1. The measurement point group is associated with the detected moving object data.

[ 0 0 6 0 ] In this way, in this embodiment, the object detection processing unit 63 detects the size of the moving object that clearly exceeds the size of the moving object from the data of the measurement points of LiDAR11 (primary measurement point data). In addition to excluding non-moving measurement points of size, dynamic measurement points are extracted by excluding static measurement points whose position has not changed since the previous processing cycle. In addition, the object detection processing unit 63 clusters the secondary measurement point data obtained by excluding the non-moving object measurement point group from the primary measurement point data, and clusters the secondary measurement point data, which is obtained by excluding the non-moving object measurement point group, from the primary measurement point data, and clusters the first moving object measurement point group consisting of the dynamic measurement point group and the moving object measurement point group. A second mobile object measurement point group consisting of a static measurement point group and a static measurement point group are classified, and a matching process is performed with the detected moving object data that has already been recorded.

[ 0 0 6 1 ] Therefore, the first moving object measurement point group is estimated to have a relatively fast moving speed, and the second moving object measurement point group is estimated to have a relatively slow moving speed. , it becomes possible to perform matching processing with already recorded detected mobile object data. Therefore, depending on the speed of the detected moving object data, it is possible to narrow down the matching targets to the first moving object measurement point group or the second moving object measurement point group, and it is possible to improve the accuracy of moving object detection. . In this way, the object detection device according to the present embodiment can detect data from the primary measurement point data detected in a series of processing cycles, even though the measurement point data of L i D A R 1 1 does not have velocity information. The same moving objects can be associated and detected.

[ 0 0 6 2 ] Furthermore, in the present embodiment, the object detection processing unit 63 detects objects with a predetermined speed among the detected moving object data. Determine whether or not a moving object with a speed value equal to or higher than a predetermined speed value can be associated with any of the first moving object measurement points, and select moving object data whose speed is less than a predetermined speed value among the detected moving object data. , executes a first matching process to determine whether or not it can be associated with either the first moving object measurement point group or the second moving object measurement point group. This allows matching targets to be narrowed down according to the speed of detected moving object data, thereby reducing the load on matching processing.

[ 0 0 6 3 ] Furthermore, in this embodiment, the object detection processing unit 63 collects the data of the first moving object measurement point group and the second moving object measurement point group, as well as the data detected by the previous processing cycle. Is it possible to input the moving object data into a machine learning model and associate the data of the first moving object measurement point group and the second moving object measurement point group with the moving object data detected by the previous processing cycle? Execute the second matching process to determine . As a result, based on the learning results of the movement patterns of various moving objects, the data of the first moving object measurement point group and the second moving object measurement point group are converted into moving object data detected by the previous processing cycle. It is possible to match with high accuracy.

[ 0 0 6 4 ] Furthermore, in the present embodiment, the object detection processing unit 63 detects the first moving object measurement point group and the second moving object measurement point group that can be associated by simple matching processing (first matching processing). After removing the group and the detected moving object data, a second matching process using a machine learning model is performed. Therefore, the processing speed and processing load required for the moving object detection processing can be further reduced, and the matching processing with the detected moving object data can be executed with high accuracy.

[ 0 0 6 5 ] Furthermore, in the present embodiment, the predetermined length value in the filtering process is set to the value of the assumed maximum length of the moving object, and the object detection processing unit 63 A group of measurement points whose length in either direction is equal to or greater than a predetermined length value are labeled and recorded as road surface irregularities or road-installed objects. As a result, a group of measurement points larger than the expected size of a moving object can be easily identified as road surface irregularities or roadside objects.

[ 0 0 6 6 ] The above effects achieved by the object detection device can also be achieved by a computer program that executes the object detection method and object detection processing.

[ 0 0 6 7 ] Note that in the matching process, the first matching process may be omitted and only the second matching process may be executed, and an object detection device that executes this processing method is also within the technical scope of the present disclosure. It is understood that it belongs.

[0068] Although preferred embodiments of the present invention have been described above in detail with reference to the accompanying drawings, the present invention is not limited to such examples. A person with ordinary knowledge in the technical field to which the present invention pertains will be able to come up with various changes and modifications within the scope of the technical idea stated in the claims. This is obvious, and it is understood that these naturally fall within the technical scope of the present invention.

[ 0 0 6 9 ] For example, in the above embodiment, the L i D A R 1 1 calculates the position and velocity of the measurement point, and transmits information on the calculated position and velocity to the object detection device 5 〇. The D A R 1 1 may calculate the position of the measurement point and transmit information on the calculated position to the object detection device 5 〇, and the object detection device 5 〇 may calculate information on the velocity of the measurement point.

[Explanation of symbols]

[ 0 0 7 0 ]

1: Vehicle, 1 1 • 1 1 LF: Radar sensor, 1 7: Position detection sensor, 4 〇: Vehicle control device, 5 0: Object detection device, 5 1: Communication section, 5 3: Processing section, 5 5: Storage unit, 6 1: Acquisition unit, 6 3: Object detection processing unit, C _R : Sensor coordinate system, C v: Vehicle coordinate system, C _w : Fixed coordinate system, P f: Erroneous measurement point, P a _ i • P b _ i: half-U measurement point, ar • av: azimuth angle

Claims

[Document name] Scope of claims

[Claim 1]

In an object detection device (1 〇〇) that detects an object based on the data of the measurement point acquired by L i D A R ( 1 1 ), the data of the measurement point acquired by L i D A R ( 1 1 ) is used. From certain primary measurement point data, extract a measurement point group in which the distance between adjacent measurement points is within a predetermined distance, and among the extracted measurement point group, the length in any direction is a predetermined length. a filtering process that removes the non-moving body measurement point group from the primary measurement point data by treating the measurement point group having a value equal to or higher as the non-moving body measurement point group; and filtering the non-moving body measurement point group from the primary measurement point data. Among the secondary measurement point data, which is the data of the measurement points excluding the static measurement points, the measurement points excluding the static measurement points, which are the measurement points that overlap with the primary measurement point data used in the previous processing cycle, are called dynamic measurement points. A dynamic measurement point extraction process is performed to extract a dynamic measurement point from the secondary measurement point data, and the distance between adjacent measurement points is within the predetermined distance. a first moving body measurement point group; a second moving body measurement point group constituted by the dynamic measurement points and the static measurement points; a clustering process for identifying the first moving body measurement point group; A matching process for associating a first moving object measurement point group and a second moving object measurement point group with moving object data detected up to the previous processing cycle; and an object detection device that performs the following steps.

[Claim 2] In the matching process, it is determined whether or not a moving object whose speed is equal to or higher than a predetermined speed value among the detected moving object data can be associated with any of the first moving object measurement point group. and determine whether it is possible to associate moving object data whose speed is less than a predetermined speed threshold value among the detected moving object data with either the first moving object measurement point group or the second moving object measurement point group. The object detection device according to claim 1, wherein the object detection device executes a first matching process to determine whether or not the object is detected.

3. The object detection device includes a machine learning model that has learned motion patterns of various moving objects, and in the matching process, the first moving object measurement point that could not be associated by the first matching process input the data of the group and the second moving object measurement point group, as well as the moving object data detected by the previous processing cycle into the machine learning model, and input the data of the first moving object measurement point group and the second moving object measurement point group 3. The object detection device according to claim 2, wherein the second matching process is executed to determine whether data of the body measurement point group can be associated with moving body data detected by the previous processing cycle.

4. The object detection device includes a machine learning model that has learned motion patterns of various moving objects, and in the matching process, the first moving object measurement point group and the second moving object measurement point group are data, and the moving object data detected up to the previous processing cycle are input into the machine learning model, and the data of the first moving object measurement point group and the second moving object measurement point group are input to the previous processing cycle. 2. The object detection device according to claim 1, wherein the object detection device determines whether or not it can be associated with moving object data that has been detected up to the physical cycle.

[Claim 5] The predetermined length value in the filtering process is set to the maximum length of an assumed moving object, and in the filtering process, the length in either direction is the predetermined length. Min value and above The object detection device according to claim 1, wherein the above measurement point group is labeled and recorded as road surface irregularities or road-installed objects.

[Claim 6]

In an object detection method that detects an object based on measurement point data acquired by L i D A R ( 1 1 ), primary measurement point data that is data of measurement points acquired by L i D A R ( 1 1 ) is , extracting a group of measurement points in which the distance between adjacent measurement points is within a predetermined distance, and among the extracted measurement point group, a measurement point whose length in any direction is equal to or greater than a predetermined length value. removing the non-moving object measurement point group from the primary measurement point data by setting the point group as a non-moving object measurement point group (step S11); Among the secondary measurement point data, which is the data of the measurement points excluding the static measurement points, the measurement points excluding the static measurement points, which are the measurement points that overlap with the primary measurement point data used in the previous processing cycle, are called dynamic measurement points. (Step S15), from the secondary measurement point data, extract a group of measurement points in which the distance between adjacent measurement points is within the predetermined distance, and a clustering process (step S17) for identifying: a first moving body measurement point group consisting of the dynamic measurement points and the static measurement points; a matching process (step S 1 9~Step S 2 9 ) An object detection method, including .

[Claim?] A computer program that causes a computer to execute a process of detecting an object based on data of a measurement point acquired by L i D A R ( 1 1 ), the computer program comprising: From the primary measurement point data, which is the data of the measured measurement points, a group of measurement points in which the distance between adjacent measurement points is within a predetermined distance is extracted, and among the extracted measurement point group, a group of measurement points whose length is equal to or greater than a predetermined length value as a group of measurement points of a non-moving object, and excluding the group of measurement points of a non-moving object from the primary measurement point data (step S11); Of the secondary measurement point data, which is data of measurement points obtained by excluding the non-moving object measurement point group from the primary measurement point data, static data is the measurement point that overlaps with the primary measurement point data used in the previous processing cycle. The measurement points excluding the measurement points are extracted as dynamic measurement points (step S15), and from the secondary measurement point data, a group of measurement points where the distance between adjacent measurement points is within the predetermined distance is extracted. and specifying a first moving body measurement point group made up of the dynamic measurement points, and a second moving body measurement point group made up of the dynamic measurement points and the static measurement points. Clustering processing (step S17), the first moving object measurement point group and the second moving object measurement point group identified by the clustering processing, and the moving object data detected up to the previous processing cycle. A computer program that executes a matching process (step S19 to step S29) that associates , and a process that includes .