Target detection device

Abstract

In object fusion, which combines detection results from multiple different sensors to detect objects, it is difficult to correctly determine if an object is the same object when it moves left or right relative to the vehicle's direction of travel. Therefore, an object detection device is provided that can easily determine if an object is the same object even when the difference in left-right position detected by each sensor increases due to the object's left-right movement relative to the vehicle's direction of travel. The object detection device uses the object information detected by each sensor to determine whether the object has moved left or right relative to the vehicle's direction of travel. If the object has moved left or right relative to the vehicle's direction of travel, the threshold value for the difference in left-right position used to determine if the object is the same object is changed.

Description

Technical Field

[0001] This invention relates to an object detection device for detecting objects in the vicinity of vehicles, pedestrians, etc. Background Technology

[0002] In recent years, object detection devices have been installed in automobiles for purposes such as preventing accidents, mitigating damage, and assisting driving. These devices use sensors such as cameras that process images to detect objects and millimeter-wave radar that uses reflected millimeter waves to detect objects. They detect people and other vehicles in the vicinity of the vehicle and are used for vehicle control.

[0003] Furthermore, to improve the detection performance of object detection devices, technologies such as object fusion have been developed, which combine the detection results of multiple different sensors rather than just a single sensor to detect objects. In one previous object fusion method, the coordinate system of the object information detected by each sensor was first unified and the time was synchronized. Then, by comparing the front-back and left-right positions of the object detected by each sensor, it was determined whether they were the same object. If they were the same object, the detection results of each sensor were combined to calculate the object information.

[0004] When implementing object fusion, deviations in object position information arise due to differences in the characteristics of various sensors. For example, when an object is located at a distance far from the vehicle, the difference in the forward and backward detection accuracy of the camera and radar tends to increase the difference in forward and backward position between the object detected by the camera and the object detected by the radar. Therefore, in Patent Document 1, in determining the same object during object fusion, the threshold for the difference in forward and backward position used to determine if the object is the same object is changed based on whether the distance between the object and the vehicle is above a certain value, thereby making it easier to determine that they are the same object.

[0005] Existing technical documents

[0006] Patent documents

[0007] Patent Document 1: Japanese Patent Application Publication No. 2006-292475 Summary of the Invention

[0008] The problem the invention aims to solve

[0009] In the approach described in Patent Document 1, by switching the threshold used for determining the forward and backward directions, it becomes easier to determine that distant objects are the same object. However, the threshold used for determining the left and right directions remains fixed. Therefore, when a vehicle in front of the vehicle changes lanes, a vehicle crosses in front of the vehicle, or when an object moves left or right relative to the vehicle's direction of travel, the difference in the left and right positions of the object detected by multiple sensors increases, making it impossible to correctly determine that the object is the same.

[0010] The present invention was made in view of the above circumstances, and its object is to provide a target detection device that can easily determine that the target is the same target even when the difference in the left and right positions of the target detected by each sensor increases due to the target moving in the left and right direction relative to the direction of travel of the vehicle.

[0011] Problem-solving methods

[0012] To address the aforementioned issues, the object detection device of this invention acquires object information around the vehicle detected by a first sensor and object information around the vehicle detected by a second sensor. Regarding objects determined to be the same in processing frames prior to a specified processing frame, based on the object information detected by the first sensor and the object information detected by the second sensor in the specified processing frame, if an object detected by the second sensor exists within a specified threshold position from the left-right position of the object detected by the first sensor, it is determined that the object detected by the first sensor and the object detected by the second sensor are the same object. Based on the object information detected by the first and second sensors in the specified processing frame, it is determined that the object is moving in the left-right direction from the perspective of the vehicle, and the specified threshold for determining it to be the same object is changed.

[0013] The effects of the invention

[0014] According to the present invention, even if the difference in the left and right positions of the object detected by each sensor increases due to the object moving in the left and right direction relative to the direction of travel of the vehicle, it can be easily determined that they are the same object. Attached Figure Description

[0015] Figure 1 This is a diagram illustrating the configuration of a driving assistance system according to an embodiment of the present invention.

[0016] Figure 2 This is a flowchart illustrating the object fusion process.

[0017] Figure 3It is a graph showing the difference in the front-back and left-right positions of the target as detected by the camera and millimeter-wave radar, respectively.

[0018] Figure 4 This is a schematic diagram illustrating an example of a scenario where an embodiment of the present invention is effective.

[0019] Figure 5 This is a diagram showing the change in the left-right position of an object as it moves to the left in front of the vehicle.

[0020] Figure 6 This is a diagram showing the change in the relative velocity of an object in the left and right directions when the object in front of the vehicle moves to the left.

[0021] Figure 7 This is a flowchart illustrating an example of the same object determination process in an embodiment of the present invention.

[0022] Figure 8A This is a schematic diagram of the left and right position threshold switching in the same object determination process of an embodiment of the present invention.

[0023] Figure 8B This is a schematic diagram showing the threshold switching of the left and right position in the same object determination process of an embodiment of the present invention.

[0024] Figure 9 This is a diagram showing the change in the relative acceleration of an object in the left and right directions when the object in front of the vehicle moves to the left.

[0025] Figure 10 This is a diagram showing the change in the direction of an object as seen from this vehicle when the object in front of this vehicle moves to the left. Detailed Implementation

[0026] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.

[0027] [Example 1]

[0028] The first embodiment of the present invention will be described. Figure 1 One example of the configuration of a driving assistance system for implementing the present invention is shown.

[0029] The driver assistance system 1 consists of a camera 11 and a millimeter-wave radar 12, which are external sensors that detect surrounding objects; an ECU 13 that performs object fusion and vehicle control information calculation; and a vehicle control unit 16 that performs vehicle control such as throttle, brake, and steering.

[0030] The camera 11, millimeter-wave radar 12, ECU 13, and vehicle control unit 16 transmit and receive data via CAN communication.

[0031] Camera 11 is installed, for example, inside the vehicle at the center of the windshield, to detect objects in front of the vehicle. Based on the images captured by the camera, camera 11 calculates object information such as the position and relative speed of the object and sends the object information to ECU 13.

[0032] Millimeter-wave radar 12 is installed, for example, at the four corners of the vehicle to detect objects around the vehicle. The millimeter-wave radar 12 uses millimeter waves to calculate the position, relative speed, and other object information of the objects and sends the object information to ECU 13.

[0033] The ECU 13 performs object fusion and vehicle control information calculation based on information obtained from various external sensors installed in the vehicle. As an internal function, the ECU 13 has an object fusion unit 14 for object fusion processing and a control information calculation unit 15 for calculating vehicle control information.

[0034] Furthermore, the object fusion unit 14 consists of an object data preprocessing unit 14a, a same object determination unit 14b, and an object fusion output value calculation unit 14c.

[0035] The control information calculation unit 15 includes an ACC control unit 15a and an AEB control unit 15b. The ACC control unit 15a is used to perform ACC (Adaptive Cruise Control) control, which enables the vehicle to travel at a constant speed when there is no preceding vehicle and to automatically maintain a distance from the preceding vehicle and follow it when there is a preceding vehicle. The AEB control unit 15b is used to perform collision damage mitigation braking control.

[0036] use Figure 2 This section will explain the object fusion process performed by the object fusion unit 14.

[0037] Figure 2 The processing flow is periodically executed by the Object Integration Department 14.

[0038] First, the object data preprocessing unit 14a receives object information detected by the camera 11 and the millimeter-wave radar 12 via CAN communication (S101).

[0039] Next, the object data preprocessing unit 14a unifies the coordinate systems of the object information received from the camera 11 and the millimeter-wave radar 12, and synchronizes them in time. At this time, as... Figure 3As shown, the object data preprocessing unit 14a sets up a coordinate system with the center of the front end of the vehicle 20 as the origin, the direction of travel of the vehicle 20 as the Z-axis, and the rightward direction as observed from the vehicle 20 as the X-axis. In addition, the object data preprocessing unit 14a converts the object information so that the coordinates representing the position of the object are the rear end of the object in the front-rear direction and the center of the object in the left-right direction, and calculates the camera object position 21 (hereinafter also referred to as the camera object) as the position of the object detected by the camera 11 and the radar object position 22 (hereinafter also referred to as the radar object) as the position of the object detected by the millimeter-wave radar 12 (S102).

[0040] Next, the same object determination unit 14b determines whether the same object detected by the camera 11 and the millimeter-wave radar 12, which was determined to be the same in the previous preprocessing frame, is also the same in this processing frame (S103).

[0041] Here we use Figure 3 This describes the method by which the same object determination unit 14b determines that the objects detected by the camera 11 and the millimeter-wave radar 12 are the same object.

[0042] The same object determination unit 14b compares the front-rear and left-right positions of the camera object position 21 and the radar object position 22 as observed from the vehicle 20. If the difference ΔZ between the front-rear positions of the camera object position 21 and the radar object position 22 is below the front-rear threshold and the difference ΔX between the left-right positions is below the left-right threshold, the same object determination unit 14b determines that the object detected by the camera 11 and the object detected by the millimeter-wave radar 12 are the same.

[0043] Here, as Figure 4 As shown, a scenario is considered where a leading vehicle 31, which is an object traveling in front of the right lane of the main vehicle 20, changes lanes in front of the main vehicle 20. Figure 5 This is a diagram showing the changes in the left and right positions of the object detected by camera 11 and the object detected by millimeter-wave radar 12 in the above-mentioned lane change scenario.

[0044] According to this Figure 5 When a vehicle changes lanes from the adjacent lane on the right to this lane, there is a portion where the difference in the left-right position of the camera target and the radar target increases.

[0045] here, Figure 6This is a graph showing the changes in the relative left and right velocities of the object detected by camera 11 and the object detected by millimeter-wave radar 12 in a scenario where, as described above, a preceding vehicle 31 traveling in the lane adjacent to the right of vehicle 20 changes lanes in front of vehicle 20. Figure 6 When comparing the relative speeds of camera and radar targets in the left and right directions, there is a period of time during which the difference in relative speeds in the left and right directions increases as the preceding vehicle changes lanes from the adjacent lane on the right to its own lane.

[0046] Therefore, in this embodiment, in S103 ( Figure 2 In the same target determination unit 14b, the difference in relative velocity in the left and right directions is used, according to... Figure 7 The processing flow shown changes the left and right thresholds used for determining the same object.

[0047] First, the same object determination unit 14b determines whether the difference between the relative speeds in the left and right directions observed from the vehicle is lower than a predetermined value for objects that were determined to be the same in the previous preprocessing frame of the processing frame (S201).

[0048] If the difference between the left and right relative velocities of the object detected by camera 11 and the object detected by millimeter-wave radar 12 is lower than a specified value, then the same object determination unit 14b will... Figure 8A As shown, a predetermined threshold (TH_X) is set for the threshold of the difference between the left and right positions used for the determination of the same object (S202).

[0049] If the difference between the left and right relative velocities of the object detected by camera 11 and the object detected by millimeter-wave radar 12 is greater than or equal to a predetermined value, then the same object determination unit 14b will... Figure 8B As shown, the threshold setting for left and right used for determining the same object is larger than the predetermined threshold (TH_X) by a threshold (TH_Xw(>TH_X)) (S203).

[0050] Furthermore, if the difference ΔZ between the front-back position of the object detected by camera 11 and the object detected by millimeter-wave radar 12 is below the front-back threshold, and the difference ΔX between the left-right positions is below the left-right threshold set in S202 or S203, the same object determination unit 14b determines that the object detected by camera 11 and the object detected by millimeter-wave radar 12 are the same object (S204). In other words, if there is a front-back position of the object detected by millimeter-wave radar 12 within the front-back threshold from the front-back position of the object detected by camera 11, and a left-right position of the object detected by millimeter-wave radar 12 within the left-right threshold set in S202 or S203 from the left-right position of the object detected by camera 11, the same object determination unit 14b determines that the object detected by camera 11 and the object detected by millimeter-wave radar 12 are the same object (S204).

[0051] Finally, based on the determination result of the same object determination unit 14b, the object fusion output value calculation unit 14c calculates the fused object information such as the position information and relative speed in the front-back and left-right directions of the object (S104), and outputs the above-mentioned fused object information to the control information calculation unit 15 of ECU13 (S105).

[0052] The ACC control unit 15a and AEB control unit 15b of the control information calculation unit 15 calculate the vehicle control information of the vehicle based on the fused object information (fusion result) output from the object fusion unit 14.

[0053] As shown in this embodiment, the object is determined to be moving in the left and right directions based on the difference between the left and right relative speeds of the object detected by the camera 11 and the object detected by the millimeter-wave radar 12. By changing (increasing) the left and right threshold used for determining the same object, it can be easily and correctly determined to be the same object regardless of whether the object is moving in the left and right directions.

[0054] In this embodiment, camera 11 is a stereo camera or a monocular camera.

[0055] Furthermore, in this embodiment, the same object determination unit 14b determines whether objects detected by the camera 11 and the millimeter-wave radar 12 that were determined to be the same in the previous preprocessing frame are also the same in this processing frame. However, it is also possible to continuously determine whether combinations that were determined to be different in the previous processing frame are the same objects.

[0056] In this embodiment, as Figure 4As shown, an example is given of a scenario where a preceding vehicle 31, which is a landmark, traveling in front of the right lane of the vehicle 20 changes lanes in front of the vehicle 20. However, in addition to this, it can also be applied to scenarios where other vehicles cross in front of the vehicle, scenarios where the vehicle turns left or right, causing the landmark to move relatively in the left or right direction from the perspective of the vehicle, and even scenarios where the relative position of the landmark changes due to the influence of the road surface when driving on bad roads.

[0057] In this embodiment, the same object determination unit 14b determines the relative speed of the left and right directions of the object detected by the camera and the object detected by the millimeter-wave radar for each processing frame and switches the left and right thresholds. However, once the left and right thresholds are set to an expanded threshold (TH_Xw), the expanded threshold (TH_Xw) can be used without relying on the above determination for a certain period of subsequent processing frames.

[0058] In this embodiment, the same object determination unit 14b switches between two thresholds for determining the left and right sides of the same object: a predetermined threshold (TH_X) and an expanded threshold (TH_Xw). However, for example, the threshold can be set as follows: if the difference between the relative velocities in the left and right directions is less than A, it can be set as a threshold (TH_X); if it is greater than A and less than B (>A), it can be set as a threshold (TH_Xw1); if it is greater than B, it can be set as a threshold (TH_Xw2), and so on. In this way, three thresholds can be switched, and more than four thresholds can be switched.

[0059] Furthermore, the same object determination unit 14b may also make the threshold variable, such as calculating the threshold based on the object information detected by the camera 11 and the object information detected by the millimeter-wave radar 12. For example, by integrating the difference in the relative velocities of the object in the left and right directions over time, the difference in the left and right positions can be calculated to what extent. The same object determination unit 14b may also calculate and set the threshold based on the difference in the relative velocities of the object detected by the camera 11 and the object detected by the millimeter-wave radar 12 in the left and right directions.

[0060] [Example 2]

[0061] In Embodiment 1, the same object determination unit 14b uses the difference between the left and right relative velocities of the object detected by the camera 11 and the object detected by the millimeter-wave radar 12 to switch the left and right thresholds for determining the same object. However, it is also possible not to use the difference between the left and right relative velocities of the object, but instead to use the difference between the left and right positions of the object (see [reference]). Figure 5 This can be used to switch the left and right thresholds used for determining the same object.

[0062] [Example 3]

[0063] In Embodiment 1, the same object determination unit 14b uses the difference between the left and right relative velocities of the object detected by the camera 11 and the object detected by the millimeter-wave radar 12 to switch the left and right thresholds for determining the same object. However, it is also possible not to use the difference between the left and right relative velocities of the object, but to use the difference between the left and right relative accelerations of the object to switch the left and right thresholds for determining the same object.

[0064] Figure 9 This diagram illustrates the change in the relative acceleration in the left-right direction between camera target position 21 (camera target) and radar target position 22 (radar target) in a scenario where a preceding vehicle, acting as an object traveling in front of this vehicle in the right lane, changes lanes in front of this vehicle. According to this... Figure 9 When comparing the relative accelerations of the object detected by camera 11 and the object detected by millimeter-wave radar 12 in the left and right directions, it can be seen that there are situations where the difference in relative accelerations in the left and right directions increases as the object moves in the left and right directions. Therefore, the same object determination unit 14b can use the difference in relative accelerations in the left and right directions of the object to perform the switching determination of the left and right thresholds for the same object determination.

[0065] [Example 4]

[0066] In Embodiment 1, the same object determination unit 14b uses the difference between the left and right relative velocities of the object detected by the camera 11 and the object detected by the millimeter-wave radar 12 to switch the left and right thresholds for determining the same object. However, it is also possible not to use the difference between the left and right relative velocities of the object, but to use the difference between the directions of the object when viewed in a polar coordinate system centered on the vehicle to switch the left and right thresholds for determining the same object.

[0067] Figure 10 This diagram illustrates the change in the direction of the object detected by camera 11 and the object detected by millimeter-wave radar 12 from the perspective of this vehicle in a scenario where a preceding vehicle, acting as an object traveling in front of this vehicle in the right lane, changes lanes in front of this vehicle. According to this... Figure 10 When comparing the direction of the object detected by camera 11 and the direction of the object detected by millimeter-wave radar 12 from the perspective of the vehicle, it can be seen that there are situations where the difference between the direction of the object and the direction of the object from the perspective of the vehicle increases as the object moves in the left and right directions. Therefore, the same object determination unit 14b can use the difference between the direction of the object and the direction of the object from the perspective of the vehicle to make a switching determination of the left and right thresholds for the same object determination.

[0068] Furthermore, even if the same object determination unit 14b uses the angular velocity and angular acceleration of the object detected by the camera 11 and the object detected by the millimeter-wave radar 12 as observed from the vehicle, it can still perform the same left and right threshold switching determination for the same object determination.

[0069] [Example 5]

[0070] In Example 1, the combination of external sensors is a camera and a millimeter-wave radar, but it can also be a combination of a camera and LiDAR (Light Detection and Ranging), etc.

[0071] [Summarize]

[0072] As described above, the object detection device (ECU13) of this embodiment acquires object information around the vehicle detected by the first sensor and object information around the vehicle detected by the second sensor (object data preprocessing unit 14a). Regarding objects that are determined to be the same in the processing frame before the specified processing frame, based on the object information detected by the first sensor and the object information detected by the second sensor in the specified processing frame, if there is a left-right position of the object detected by the second sensor within a specified threshold (TH_X) from the left-right position of the object detected by the first sensor, it is determined that the object detected by the first sensor and the object detected by the second sensor are the same object (same object determination unit 14b). Based on the object information detected by the first sensor and the second sensor in the specified processing frame, the object detection device (ECU13) determines that the object is moving in the left-right direction from the perspective of the vehicle and changes (expands the change to TH_Xw(>TH_X)) the specified threshold (TH_X) at which it is determined to be the same object.

[0073] Furthermore, regarding objects determined to be the same in processing frames prior to the specified processing frame, the object detection device (ECU13) compares the left-right positions of the object as observed from the vehicle by the first sensor and the second sensor in the specified processing frame. If the difference between the left-right positions of the object is greater than or equal to a specified value, it determines that the object is moving in the left-right direction as observed from the vehicle, and changes (expands to TH_Xw(>TH_X)) the specified threshold (TH_X) for determining that it is the same object.

[0074] Furthermore, regarding objects determined to be the same object in the processing frame preceding the specified processing frame, the object detection device (ECU13) compares the relative speed of the object in the left and right directions as observed from the vehicle by the first sensor and the second sensor in the specified processing frame. If the difference between the relative speeds in the left and right directions of the object is greater than or equal to a specified value, it determines that the object is moving in the left and right directions as observed from the vehicle, and changes (expands to TH_Xw(>TH_X)) the specified threshold (TH_X) for determining that it is the same object.

[0075] Furthermore, regarding objects determined to be the same in processing frames prior to the specified processing frame, the object detection device (ECU13) compares the relative acceleration in the left and right directions of the object as observed from the vehicle by the first sensor and the second sensor in the specified processing frame. If the difference between the relative accelerations in the left and right directions of the object is greater than or equal to a specified value, it determines that the object is moving in the left and right directions as observed from the vehicle, and changes (expands to TH_Xw(>TH_X)) the specified threshold (TH_X) for determining that it is the same object.

[0076] Furthermore, regarding objects determined to be the same object in the processing frame preceding the specified processing frame, the object detection device (ECU13) compares the direction of the object as observed from the vehicle by the first sensor and the second sensor in the specified processing frame. If the difference between the direction of the object as observed from the vehicle is greater than or equal to a specified value, it determines that the object is moving in the left-right direction as observed from the vehicle, and changes (expands to TH_Xw(>TH_X)) the specified threshold (TH_X) for determining that it is the same object.

[0077] Furthermore, regarding objects determined to be the same in processing frames prior to the specified processing frame, the object detection device (ECU13) compares the left-right positions of the object as observed from the vehicle by the first sensor and the second sensor in the specified processing frame. Based on the difference in the left-right positions of the object, it determines that the object is moving in the left-right direction as observed from the vehicle, and switches (switching between TH_X and TH_Xw) the specified threshold for determining that it is the same object.

[0078] Furthermore, regarding objects determined to be the same in processing frames prior to the specified processing frame, the object detection device (ECU13) compares the relative speeds of the object in the left and right directions as observed from the vehicle by the first sensor and the second sensor in the specified processing frame. Based on the difference in the relative speeds of the object in the left and right directions, it determines that the object is moving in the left and right directions as observed from the vehicle, and switches (switching between TH_X and TH_Xw) the specified threshold for determining that it is the same object.

[0079] Furthermore, regarding objects determined to be the same in processing frames prior to the specified processing frame, the object detection device (ECU13) compares the relative acceleration in the left and right directions of the object as observed from the vehicle by the first sensor and the second sensor in the specified processing frame. Based on the difference in the relative acceleration in the left and right directions of the object, it determines that the object is moving in the left and right directions as observed from the vehicle, and switches (switching between TH_X and TH_Xw) the specified threshold for determining that it is the same object.

[0080] Furthermore, regarding objects determined to be the same in processing frames prior to the specified processing frame, the object detection device (ECU13) compares the direction of the object as observed from the vehicle by the first sensor and the second sensor in the specified processing frame, and determines, based on the difference in the direction of the object as observed from the vehicle, that the object is moving in the left-right direction as observed from the vehicle, and switches (switching between TH_X and TH_Xw) the specified threshold for determining that it is the same object.

[0081] In other words, when the object detection device (ECU13) of this embodiment is moving in the left and right direction, the difference in the left and right position of the object detected by different sensors may make it impossible to correctly determine whether the object detected by each sensor is the same object. Therefore, it determines whether the object is moving in the left and right direction as observed from the vehicle, and expands the threshold of the difference in the left and right position of the object detected by each sensor as the same object when it is moving.

[0082] According to this embodiment, even if the difference in the left and right positions of the object detected by each sensor increases due to the object moving in the left and right direction relative to the direction of travel of the vehicle, it can still be easily determined that they are the same object.

[0083] Furthermore, the present invention is not limited to the embodiments described above, but includes various modifications. For example, the embodiments described above are given in detail for ease of understanding of the present invention and are not necessarily limited to having all the described configurations.

[0084] Furthermore, some or all of the aforementioned components, functions, processing units, and processing methods can be implemented in hardware, for example, using integrated circuit design. Additionally, the aforementioned components and functions can be implemented in software by a processor interpreting and executing programs that implement each function. The programs, tables, files, and other information implementing each function can be stored in recording devices such as memory, hard disks, SSDs (Solid State Drives), or recording media such as IC cards, SD cards, and DVDs.

[0085] Furthermore, control lines and information lines are shown only those deemed necessary in the description, but not all control lines and information lines are necessarily shown on the product. In fact, it can be assumed that almost all components are interconnected.

[0086] Symbol Explanation

[0087] 1: Driving assistance system; 11: Camera (first sensor); 12: Millimeter-wave radar (second sensor); 13: ECU (object detection device); 14: Object fusion unit; 15: Control information calculation unit; 16: Vehicle control unit; 20: Main vehicle; 21: Object detected by camera (camera object position); 22: Object detected by millimeter-wave radar (radar object position); 31: Precursor vehicle that becomes the target.

Claims

1. A target detection device, characterized in that, The object detection device acquires object information around the vehicle detected by a first sensor and object information around the vehicle detected by a second sensor. Regarding objects determined to be the same in processing frames prior to a specified processing frame, based on the object information detected by the first sensor and the object information detected by the second sensor in the specified processing frame, if the left-right position of the object detected by the first sensor exists within a specified threshold from the left-right position of the object detected by the first sensor, then the object detected by the first sensor and the object detected by the second sensor are determined to be the same object. Based on the object information detected by the first sensor and the second sensor in the specified processing frame, it is determined that the object is moving in the left-right direction from the perspective of the vehicle, and the specified threshold for determining that it is the same object is changed.

2. The object detection device according to claim 1, characterized in that, Regarding objects determined to be the same in processing frames prior to the specified processing frame, the object detection device compares the left-right positions of the object as observed from the vehicle by the first sensor and the second sensor in the specified processing frame. If the difference between the left-right positions of the object is greater than or equal to a specified value, the device determines that the object is moving in the left-right direction as observed from the vehicle, and changes the specified threshold for determining that it is the same object.

3. The object detection device according to claim 1, characterized in that, Regarding objects determined to be the same in processing frames prior to the specified processing frame, the object detection device compares the relative speeds of the objects in the left and right directions as observed from the vehicle by the first sensor and the second sensor in the specified processing frame. If the difference between the relative speeds of the objects in the left and right directions is greater than or equal to a specified value, the device determines that the object is moving in the left and right directions as observed from the vehicle, and changes the specified threshold for determining that the object is the same.

4. The object detection device according to claim 1, characterized in that, Regarding objects determined to be the same in processing frames prior to the specified processing frame, the object detection device compares the relative acceleration of the object in the left and right directions as observed from the vehicle by the first sensor and the second sensor in the specified processing frame. If the difference between the relative accelerations in the left and right directions of the object is greater than or equal to a specified value, it determines that the object is moving in the left and right directions as observed from the vehicle, and changes the specified threshold for determining that it is the same object.

5. The object detection device according to claim 1, characterized in that, Regarding objects determined to be the same in processing frames prior to the specified processing frame, the object detection device compares the direction of the object as observed from the vehicle by the first sensor and the second sensor in the specified processing frame. If the difference between the direction of the object as observed from the vehicle is greater than or equal to a specified value, it determines that the object is moving in the left-right direction as observed from the vehicle, and changes the specified threshold for determining that it is the same object.

6. The object detection device according to claim 1, characterized in that, Regarding objects determined to be the same in processing frames prior to the specified processing frame, the object detection device compares the left-right positions of the object as observed from the vehicle by the first sensor and the second sensor in the specified processing frame. Based on the difference in the left-right positions of the object, it determines that the object is moving in the left-right direction as observed from the vehicle, and switches to the specified threshold for determining that it is the same object.

7. The object detection device according to claim 1, characterized in that, Regarding objects determined to be the same in processing frames prior to the specified processing frame, the object detection device compares the relative speeds of the objects in the left and right directions as observed from the vehicle by the first sensor and the second sensor in the specified processing frame. Based on the difference in the relative speeds of the objects in the left and right directions, it determines that the objects are moving in the left and right directions as observed from the vehicle, and switches to the specified threshold for determining that they are the same objects.

8. The object detection device according to claim 1, characterized in that, Regarding objects determined to be the same in processing frames prior to the specified processing frame, the object detection device compares the relative acceleration of the object in the left and right directions as observed from the vehicle by the first sensor and the second sensor in the specified processing frame. Based on the difference in the relative acceleration of the object in the left and right directions, it determines that the object is moving in the left and right directions as observed from the vehicle, and switches to the specified threshold for determining that it is the same object.

9. The object detection device according to claim 1, characterized in that, Regarding objects determined to be the same in processing frames prior to the specified processing frame, the object detection device compares the direction of the object as observed from the vehicle by the first sensor and the second sensor in the specified processing frame. Based on the difference in the direction of the object as observed from the vehicle, it determines that the object is moving in the left-right direction as observed from the vehicle, and switches to the specified threshold for determining that it is the same object.

10. The object detection device according to claim 1, characterized in that, The first sensor is a camera, and the second sensor is a millimeter-wave radar.

11. The object detection device according to claim 1, characterized in that, The first sensor is a camera, and the second sensor is a LiDAR.

Images