Object Tracking Device

Abstract

An object tracking device (20) according to one aspect of the present disclosure includes a detection unit (21), a prediction unit (22), a first range setting unit (23), an estimation unit (25), a registration unit (30), a suppression range setting unit (27), and a registration suppression unit (29). The registration suppression unit (29) suppresses a case where an observation value within a suppression range set by the suppression range setting unit (27) among at least one observation value detected by the detection unit (21) is registered by the registration unit (30) as the new target.

Description

[0001] Cross - reference to related applications

[0002] This international application claims priority based on Japanese Patent Application No. 2019 - 188667 filed with the Japan Patent Office on October 15, 2019, and by reference, the entire contents of Japanese Patent Application No. 2019 - 188667 are incorporated into this international application. Technical field

[0003] This disclosure relates to an object tracking device for tracking an object. Background art

[0004] The target motion estimation device described in Patent Document 1 below generates a target based on a radar signal from a radar device and performs tracking of the target. Specifically, the above - mentioned device calculates a predicted value of the position of the target in the current processing cycle based on the estimated value of the position of the target in the previous processing cycle, and sets a prediction gate centered on the predicted value. Moreover, the above - mentioned device associates the observed value of the position observed in the current processing cycle that exists within the set prediction gate and is closest to the predicted value with the predicted value, and calculates the estimated value in the current processing cycle based on the associated observed value and predicted value. In such a device for tracking an object, in the case where there is an observed value that is not associated with the predicted value, it is processed as a new target.

[0005] Patent Document 1: Japanese Patent No. 3629328 Gazette.

[0006] In the case of using a high - resolution radar, multiple observed values are obtained from the same object, and thus sometimes multiple targets are generated from the same object. The inventors' detailed research results have found the following problem: In a processing cycle after a processing cycle in which multiple targets are generated, when the number of observed values obtained is less than the number of targets being recognized, there is a conflict in the observed values between the target being tracked and the new target generated from the same object. As a result of the conflict in the observed values, if the observed value is associated with the new target, the tracking of the target being tracked cannot continue, and the new target is re - tracked.

[0007] On the other hand, as a vehicle driving assistance system, there is a driving assistance system in which the longer the duration of tracking a target, the higher the reliability of the tracking result, and if the reliability is above a threshold value, the tracking result is used for vehicle control. In such a driving assistance system, if the tracking of the target is interrupted and re - tracked, it may cause a delay in vehicle control. Summary of the invention

[0008] One aspect of this disclosure preferably enables stable tracking of a target.

[0009] An object tracking device according to one aspect of the present disclosure estimates the state quantity of at least one object in each processing cycle set in advance, and includes a detection unit, a prediction unit, a first range setting unit, a determination unit, an estimation unit, a registration unit, a suppression range setting unit, and a registration suppression unit. The detection unit is configured to detect at least one observation value based on an observation signal observed by a sensor. The at least one observation value is information about at least one object around a vehicle. The prediction unit is configured to calculate a predicted value of the current state quantity based on the estimated value of the state quantity in the past for each object included in the at least one object. The first range setting unit is configured to set a first range based on the predicted value for each predicted value calculated by the prediction unit. The first range is a range estimated to be the observation value acquired this time. The determination unit is configured to determine an observation value within the first range set by the first range setting unit based on at least one observation value detected by the detection unit for each predicted value calculated by the prediction unit, and the observation value is associated with the predicted value. The estimation unit is configured to calculate an estimated value of the current state quantity based on the observed value and the predicted value determined by the determination unit for each predicted value calculated by the prediction unit. The registration unit is configured to register an observation value that is not associated with any predicted value among at least one observation value detected by the detection unit as a new object mark. The suppression range setting unit is configured to set a suppression range for each predicted value calculated by the prediction unit. The suppression range is a range of observation values ​​that suppress registration of new object marks. The registration suppression unit is configured to suppress the situation in which an observation value within the suppression range set by the suppression range setting unit among at least one observation value detected by the detection unit is registered as a new object mark by the registration unit.

[0010] According to one aspect of the present disclosure, for each object mark, a predicted value of the current state quantity is calculated based on the estimated value of the past state quantity, and for each calculated predicted value, a first range is set based on the predicted value. Moreover, for each predicted value, an observed value associated with the predicted value is determined based on the observed value within the first range among the detected observed values, and the estimated value of the current state quantity is calculated based on the determined observed value and the predicted value. Moreover, for each predicted value, a suppression range is set, and the observed value that is not associated with any predicted value among the acquired observed values, that is, the observed value outside the suppression range is registered as a new object mark. On the other hand, even if it is an observed value that is not associated with any predicted value, the observed value within the suppression range is likely to be observed from the same object as the observed value associated with the predicted value. Therefore, the situation in which the observed value within the suppression range is registered as a new object mark is suppressed. Therefore, the situation in which multiple object marks are generated from the same object can be suppressed. Furthermore, the object mark can be tracked stably. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 This is a block diagram showing the structure of the object tracking device according to the first embodiment.

[0012] Figure 2 is a block diagram showing the functions of the object tracking device according to the first embodiment.

[0013] Figure 3 is a flowchart showing the object tracking process executed by the object tracking device according to the first embodiment.

[0014] Figure 4 is a subroutine showing the registration inhibition determination process executed by the object tracking device according to the first embodiment.

[0015] Figure 5 is a diagram showing the first range for associating the predicted value and the observed value according to the first embodiment, and the inhibition range for suppressing the registration of new targets.

[0016] Figure 6 is a flowchart showing the object tracking process executed by the object tracking device according to the second embodiment.

[0017] Figure 7 is a subroutine showing the inhibition range setting process according to the second embodiment.

[0018] Figure 8 is a diagram showing an example of the inhibition ranges, i.e., the first to fourth ranges, for suppressing the registration of new targets according to the second embodiment.

[0019] Figure 9 is a diagram showing another example of the inhibition ranges, i.e., the first to fourth ranges, for suppressing the registration of new targets according to the second embodiment.

[0020] Figure 10 is a diagram showing the first range for suppressing the registration of new targets according to the second embodiment, and the noise range for suppressing the registration of noise.

[0021] Figure 11 is a diagram showing the third range for suppressing the registration of new targets according to the second embodiment, and the noise range for suppressing the registration of noise.

[0022] Figure 12 is a diagram showing the state of estimating the shape of the target according to the second embodiment.

[0023] Figure 13 is a diagram showing the shapes of the first range, the inhibition range, and the noise range according to other embodiments.

[0024] Figure 14 is a diagram showing the shapes of the first range, the inhibition range, and the noise range according to other embodiments. Detailed Embodiments

[0025] Hereinafter, an exemplary embodiment for implementing the present disclosure will be described with reference to the accompanying drawings.

[0026] (First Embodiment)

[0027] <1-1. Configuration>

[0028] First, refer to Figure 1 to describe the configuration of the driving assistance system 100 of the present embodiment. The driving assistance system 100 includes a radar device 10, an object tracking device 20, and a driving assistance device 50.

[0029] The radar device 10 may be mounted at the center of the front of the vehicle 80 (for example, the center of the front bumper), and has the area at the center of the front of the vehicle 80 as the detection area. In addition, the radar device 10 may be respectively mounted on the left front side and the right front side of the vehicle 80 (for example, the left end and the right end of the front bumper), and has each of the areas at the left front and the right front of the vehicle 80 as the detection area. In addition, the radar device 10 may be respectively mounted on the left rear side and the right rear side of the vehicle 80 (for example, the left end and the right end of the rear bumper), and has each of the areas at the left rear and the right rear of the vehicle 80 as the detection area. It is not necessary to mount all of these five radar devices 10 on the vehicle 80. One of the five radar devices 10 may be mounted on the vehicle 80, or two or more of the five radar devices 10 may be mounted on the vehicle 80.

[0030] The radar device 10 is a high-resolution millimeter-wave radar. The radar device 10 has a transmitting array antenna including a plurality of antenna elements and a receiving array antenna including a plurality of antenna elements. The radar device 10 repeatedly transmits a transmitted wave at a predetermined cycle, and receives a reflected wave generated by the reflection of the transmitted wave by an object. Then, the radar device 10 mixes the transmitted wave and the reflected wave to generate a beat signal, and outputs the sampled beat signal (i.e., the observation signal) to the object tracking device 20. The radar device 10 may be any modulation method such as the FMCW method or the multi-frequency CW method.

[0031] The object tracking device 20 includes a microcomputer having a CPU and semiconductor memories such as a ROM and a RAM. The object tracking device 20 realizes various functions by the CPU executing various programs stored in the ROM. Specifically, as Figure 2 shown by the solid line in, the object tracking device 20 realizes the functions of a detection unit 21, a prediction unit 22, a first range setting unit 23, a decision unit 24, an estimation unit 25, a suppression range setting unit 27, a registration suppression unit 29, and a registration unit 30 to perform object tracking processing. Moreover, the object tracking device 20 outputs the target information generated by the execution of the object tracking processing to the driving assistance device 50. In addition, the details of the object tracking processing will be described later.

[0032] The driving assistance device 50 uses the target information generated by the object tracking device 20 and the state information and operation status information of the vehicle 80 obtained from various sensors mounted on the vehicle 80 to control the vehicle 80 to achieve driving assistance.

[0033] <1-2. Processing>

[0034] Next, with reference to Figure 3 the flowchart, the object tracking process executed by the object tracking device 20 of the first embodiment will be described. The object tracking device 20 repeatedly executes this process at a predetermined cycle.

[0035] First, in S10, the detection unit 21 detects the observation values of each target existing around the vehicle 80 based on the observation signals obtained from the radar device 10. The observation values include the power value of the observation signal, the distance from the vehicle 80 to the target, the azimuth of the target relative to the vehicle 80, and the relative speed of the target relative to the vehicle 80. In addition, the observation value may include the ground speed of the target calculated based on the relative speed and the speed of the vehicle 80 instead of the relative speed of the target.

[0036] Next, in S20, the prediction unit 22 determines whether there is unprocessed target information. Specifically, it is determined whether there is a target among the registered targets for which the subsequent processing in S30 to S80 has not been performed. If it is determined that there is unprocessed target information, the process proceeds to S30.

[0037] In S30, for one of the unprocessed targets, the prediction unit 22 calculates the predicted value of the state quantity of the target in the current processing cycle based on the estimated value of the state quantity of the target calculated in the previous processing cycle. Similar to the observation value, the predicted value of the state quantity of the target may have the power value P of the observation signal, the distance R of the target, the azimuth θ, and the speed Vr as elements, or may have the power value P of the observation signal, the X-axis coordinate value Cx, the Y-axis coordinate value Cy, the X-direction speed Vx, and the Y-direction speed Vy as elements. The X-axis is the axis along the width direction of the vehicle 80, and the Y-axis is orthogonal to the X-axis and along the long side direction of the vehicle 80. In addition, the speed Vr may be the relative speed relative to the vehicle 80 or the ground speed. The X-direction speed Vx may be the X-direction component of the relative speed of the target relative to the vehicle 80 or the X-direction component of the ground speed of the target. The Y-direction speed Vy may be the Y-direction component of the relative speed of the target relative to the vehicle 80 or the Y-direction component of the ground speed of the target.

[0038] Next, in S40, the first range setting unit 23 sets a first range based on at least one element of the predicted value calculated in S30. The first range is a range presumed to obtain an observed value in the current processing cycle. The observed value detected from the same object as the predicted value should be a value close to the predicted value. Therefore, as Figure 5 shown, as the first range, with the predicted value calculated in S30 as the center, a range of observed values presumed to be detected from the same object as the predicted value is set. The first range setting unit 23, for example, assumes the object to be a preceding vehicle traveling in the same direction as the vehicle 80 and sets the first range. In the present embodiment, based on two elements in the state quantity, a square range centered on the predicted value is set as the first range. It is also possible to set a range based on three or more elements in the state quantity as the first range.

[0039] Next, in S50, the determination unit 24 performs an association establishment process. The determination unit 24 determines, from the observed values detected in S10, the observed values associated with the predicted value calculated in S30. Specifically, the observed values within the first range set in S40, that is, the observed values closest to the predicted value calculated in S30, are determined as the observed values associated with the predicted value. In Figure 5 the example shown, within the first range, the estimated value A1 and the observed value A2 in the current processing cycle are detected. Among the observed value corresponding to the estimated value A1 and the observed value A2, the observed value corresponding to the estimated value A1 is closer to the predicted value, so this observed value is determined as the object associated with the predicted value. The observed value A2 is not associated with the predicted value.

[0040] Next, in S60, the estimation unit 25 calculates the estimated value in the current processing cycle using a Kalman filter or the like based on the predicted value calculated in S30 and the observed value determined to be the association object in S50. The estimated value of the state quantity may have elements such as the power value P of the observation signal, the distance R of the target, the azimuth θ, and the speed Vr, or may have elements such as the power value P of the observation signal, the X-axis coordinate value Cx, the Y-axis coordinate value Cy, the X-direction speed Vx, and the Y-direction speed Vy. The estimated value of the state quantity may have the same elements as the observed value or the predicted value, or may have elements different from the observed value or the predicted value.

[0041] Next, in S70, the suppression range setting unit 27 sets a suppression range. The suppression range is a range of observed values that suppresses the registration of new targets. In the present embodiment, the first range set in S40 is set as the suppression range.

[0042] Next, in S80, the registration suppression unit 29 performs a registration suppression determination process. Specifically, the registration suppression unit 29 performs Figure 4The subroutine shown below. First, in S200, it is determined whether there are unprocessed observations. Specifically, it is determined whether there are observations among the observations detected in S10 that are not associated with the predicted values and for which the subsequent processing of S210 to S220 has not been executed.

[0043] Here, in a high-resolution radar, multiple observations are sometimes detected from the same object. One of these multiple observations is associated with the predicted value in S340, and the remaining observations exist as unprocessed observations. In addition, the observations of an object detected for the first time in the current processing cycle also exist as unprocessed observations.

[0044] In S200, if it is determined that there are no unprocessed observations, this subroutine ends and the process returns to S20. On the other hand, in S200, if it is determined that there are unprocessed observations, the process proceeds to S210.

[0045] In S210, for one of the unprocessed observations, it is determined whether it is an observation within the suppression range set in S70. That is, the registration suppression unit 29 determines whether the unprocessed observation is an observation among the multiple observations detected from the same object that is not associated with the predicted value or an observation of an object detected for the first time. The suppression range is a range for determining whether an observation is detected from the same object as the object corresponding to the predicted value.

[0046] In Figure 5 In the example shown, the observation A2 is determined to be an observation within the suppression range. In S210, if it is determined that the observation is within the suppression range, the process proceeds to S220. On the other hand, in S210, if it is determined that the observation is outside the suppression range, the process returns to S200 and this subroutine is executed for the next unprocessed observation.

[0047] In S220, a suppression flag is set for the observation determined to be within the suppression range in S210. The suppression flag is a flag for suppressing the case of registering an observation as a new target.

[0048] For example, in addition to the landmark T1 in the tracking generated for the object O1, it is assumed that landmarks T2 and T3 are also generated based on the observed values B1 and B2 detected from the object O1. In the processing cycles after the next time, when two observed values C1 and C2 are detected from the object O1, the three landmarks T1, T2, and T3 compete for the two observed values C1 and C2. Moreover, if the predicted values of the landmarks T2 and T3 are associated with the observed values C1 and C2, the tracking of the landmark T1 is interrupted, and the tracking of the landmarks T2 and T3 is started. Therefore, the tracking of the landmark T1 cannot be continued. That is, if multiple landmarks are generated from one object, when fewer observed values than the number of landmarks are detected in subsequent processing cycles, the tracking of the landmarks may not be able to be continued.

[0049] Therefore, in order not to generate multiple landmarks from the same object, an inhibition flag is set for the observed values that are not associated with the predicted values among the multiple observed values that are presumably detected from the same object. In Figure 5 the example shown, an inhibition flag is set for the observed value A2.

[0050] However, the registration inhibition unit 29 does not set an inhibition flag for the observed values within the pre-set distance threshold from the landmark among the observed values determined to be within the inhibition range in S210. The registration inhibition unit 29 stops inhibiting the registration of the observed values detected from an object at a distance relatively close to 80 from the vehicle as a new landmark. That is, in the distance relatively close to 80 from the vehicle, the registration inhibition unit 29 prioritizes the landmarking performance over the situation of inhibiting the generation of multiple landmarks from one object. Then, the process returns to the process of S200.

[0051] If the Figure 4 subroutine shown ends, the process returns to Figure 3 the process of S20 in the object tracking process shown. Moreover, during the period when there is unprocessed landmark information, the processes of S20 to S80 are repeatedly executed. On the other hand, when there is no unprocessed landmark information and it is determined in S20 that there is no unprocessed landmark information, the process proceeds to the process of S90.

[0052] In S90, the registration unit 30 determines whether there are unused observed values among the observed values detected in S10. That is, it determines whether there are observed values that are not associated with any predicted values among the observed values detected in S10. In S90, if it is determined that there are no unused observed values, this process ends. On the other hand, in S90, if it is determined that there are unused observed values, the process proceeds to the process of S100.

[0053] In S100, the registration unit 30 determines whether one of the unused observed values is outside the object of registration suppression. Specifically, when a suppression flag is set for an observed value, it is determined that it is an object of registration suppression, and when a suppression flag is not set for the observed value, it is determined that it is outside the object of registration suppression.

[0054] In S100, when it is determined that it is an object of registration suppression, the process returns to the process of S90, and when it is determined that it is outside the object of registration suppression, the process proceeds to the process of S110.

[0055] In S110, the observed value outside the object of registration suppression is registered as a new target. Then, the process returns to the process of S90, and while there are unused observed values for which the processes of S90 to S110 have not been performed, the processes of S90 to S110 are repeatedly executed. This concludes this process.

[0056] <1-3. Effects>

[0057] According to the first embodiment described above, the following effects can be obtained.

[0058] (1) For each target, the predicted value of the current state quantity is calculated based on the estimated value of the past state quantity, and for each calculated predicted value, a first range is set based on the predicted value. Moreover, for each predicted value, based on the observed values within the first range among the acquired observed values, the observed values associated with the predicted value are determined, and based on the determined observed values and the predicted value, the estimated value of the current state quantity is calculated. And for each predicted value, a suppression range is set, and the observed values outside the suppression range, that is, the observed values not associated with any of the predicted values among the acquired observed values, are registered as new targets. On the other hand, even for the observed values not associated with any of the predicted values, the observed values within the suppression range are likely to be observed from the same object as the observed values associated with the predicted value. Therefore, the situation where the observed values within the suppression range are registered as new targets is suppressed. Thus, the situation where multiple targets are generated from the same object can be suppressed. Furthermore, the target can be stably tracked.

[0059] (2) Within the range of the observed values associated with the predicted value, that is, the first range, the observed values not associated with the predicted value are likely to be the observed values observed from the same object as the observed values associated with the predicted value. Therefore, by setting the first range as the suppression range, the situation where multiple targets are generated from the same object can be appropriately suppressed.

[0060] (3) The range of the observed values associated with the predicted value and the range for suppressing the observed values registered as new targets can be set based on the physical quantity observable by the radar device 10.

[0061] (4) Even if the observed values within the short distance within the vehicle 80 distance threshold enter the suppression range, the registration as a new target is not suppressed. Thus, in the short distance, compared with the case of suppressing the generation of multiple targets from the same object, the targetization performance can be prioritized.

[0062] (Second Embodiment)

[0063] <2-1. Differences from the First Embodiment>

[0064] The basic configuration of the second embodiment is the same as that of the first embodiment. Therefore, the description of the same configuration is omitted, and the description will be centered on the differences. In addition, the same reference numerals as those in the first embodiment represent the same configuration, and refer to the previous description.

[0065] As Figure 2 shown by the dashed line in, the difference between the object tracking device 20 of the second embodiment and the object tracking device 20 of the first embodiment is that in addition to the functions of the object tracking device 20 of the first embodiment, the functions of the determination unit 26, the noise range setting unit 28, and the shape estimation unit 31 are also implemented. The details of the determination unit 26, the noise range setting unit 28, and the shape estimation unit 31 will be described later.

[0066] <2-2. Processing>

[0067] Next, with reference to Figure 6 the flowchart of, the object tracking process performed by the object tracking device 20 of the second embodiment will be described. The object tracking device 20 repeatedly executes this process at a prescribed cycle.

[0068] First, in S300 to S350, the same processing as S10 to S60 is executed.

[0069] Next, in S360, the determination unit 26 determines the type of the object corresponding to the target. Specifically, the speed Vr, distance R, and azimuth θ of any one of the observed value, estimated value, and predicted value of the target are used to determine the type of the object. The X-axis coordinate value Cx, Y-axis coordinate value Cy, X-direction speed Vx, and Y-direction speed Vy can also be used instead of the speed Vr, distance R, and azimuth θ. The types of objects include pedestrians, bicycles crossing in front of the vehicle 80, and cars crossing in front of the vehicle 80.

[0070] Next, in S370, the suppression range setting unit 27 sets the suppression range according to the type of the object determined in S360. Specifically, the suppression range setting unit 27 executes Figure 7 the subroutine shown.

[0071] First, in S500, it is determined whether the type of the object is a pedestrian. In S500, if it is determined that the type of the object is a pedestrian, the process proceeds to S510.

[0072] In S510, as Figure 8 and Figure 9 shown, a second range wider than the first range is set as the suppression range. Since a pedestrian has swinging of hands and feet, the expansion of the speed Vr, or the X-direction speed Vx and the Y-direction speed Vy is larger than that of the preceding vehicle. Therefore, the second range is a range in which the range of the speed Vr, or the ranges of the X-direction speed Vx and the Y-direction speed Vy are expanded compared to the first range. After the process of S510, this subroutine ends and the process proceeds to S380.

[0073] In addition, in S500, if it is determined that the type of the object is not a pedestrian, the process proceeds to S520. In S520, it is determined whether the type of the object is a crossing bicycle. In S520, if it is determined that the type of the object is a crossing bicycle, the process proceeds to S530.

[0074] In S530, as Figure 8 and Figure 9 shown, a third range wider than the first range is set as the suppression range. Compared with the preceding vehicle, the length of the vehicle 80 in the width direction is longer for a crossing bicycle, so the expansion of the azimuth θ, or the expansion of the X-axis coordinate value Cx, is larger compared to the preceding vehicle. Therefore, the third range is a range in which the range of the azimuth θ or the range of the X-axis coordinate value Cx is expanded compared to the first range. After the process of S530, this subroutine ends and the process proceeds to S380.

[0075] In addition, in S520, if it is determined that the type of the object is not a crossing bicycle, the process proceeds to S540. In S540, it is determined whether the type of the object is a crossing car. In S540, if it is determined that the type of the object is a crossing car, the process proceeds to S550.

[0076] In S550, as Figure 8 and Figure 9 shown, a fourth range wider than the third range is set as the suppression range. Compared with a crossing bicycle, the length of the vehicle 80 in the width direction is longer for a crossing car, so the expansion of the azimuth θ, or the expansion of the X-axis coordinate value Cx, is larger compared to the crossing bicycle. Therefore, the fourth range is a range in which the range of the azimuth θ or the range of the X-axis coordinate value Cx is expanded after comparing with the third range. After the process of S550, this subroutine ends and the process proceeds to S380. On the other hand, in S540, if it is determined that the type of the object is not a crossing car, the first range is set as the suppression range, this subroutine ends, and the process proceeds to S380.

[0077] Next, in S380, the noise range setting unit 28 sets a noise range for the predicted value calculated in S320. The noise range is a range for suppressing the case where a noise peak is registered as a new target. As Figure 10 and Figure 11 shown, the noise range is a range wider than the suppression range set in S370. Moreover, for the observed values outside the suppression range and within the noise range, where the power difference between the power value P included in the observed value and the power value P included in the predicted value is equal to or greater than a preset power threshold, a suppression flag is set. Thereby, the case where a noise peak detected near an object is registered as a new target is suppressed.

[0078] Next, in S390, the same processing as in S80 is performed.

[0079] Next, in S400, the shape estimation unit 31 estimates the shape of the object indicated by the predicted value calculated in S320. Specifically, as Figure 12 shown, before the current processing cycle, the observed values in the case where registration as a new target is suppressed are used to estimate the shape of the object. That is, the information of the observed values that are not associated with the predicted value among the observed values within the suppression range set for the predicted value is stored in the memory in subsequent processing cycles. Moreover, in subsequent processing cycles, the shape of the object is estimated based on the positions of all the observed values associated with the predicted value.

[0080] In Figure 12 the example shown, within the suppression range set in S370, there are the estimated value A10 detected in the current processing cycle and the observed values A11 to A14 that were suppressed from being registered as new targets before the previous processing cycle. In this case, the shape of the object is estimated based on the positions of the stored observed values A11 to A14. Among the observed values A11 to A14, the observed values that were suppressed from being registered as new targets in the current processing cycle may also be included. The information of the observed values A11 to A14 is also stored in subsequent processing cycles after the next cycle.

[0081] Next, in S410 to S430, the same processing as in S90 to S110 is performed. Thus, this processing ends.

[0082] <2-3. Effects>

[0083] According to the second embodiment described above, in addition to the effects (1) to (4) of the above-described first embodiment, the following effects can also be obtained.

[0084] (5) Determine the type of the object corresponding to the target, and set the suppression range according to the type of the object. Therefore, an appropriate suppression range can be set according to the type of the object.

[0085] (6) A pedestrian has swinging hands and feet, so compared with the vehicle in front, the expansion of the speed Vr or the speed Vx in the X direction and the speed Vy in the Y direction is larger. Therefore, when it is determined that the type of the object is a pedestrian, the range of the speed Vr or the speed Vx in the X direction and the speed Vy in the Y direction of the suppression range is expanded compared with the first range. Thus, when the type of the object is a pedestrian, the situation of generating multiple targets from the same object can be appropriately suppressed.

[0086] (7) The length of the width direction of the vehicle 80 of the oncoming bicycle is longer than that of the vehicle in front. Therefore, when it is determined that the type of the object is an oncoming bicycle, the range of the width direction of the vehicle 80 or the range of the azimuth θ of the suppression range is expanded. Thus, when the type of the object is an oncoming bicycle, the situation of generating multiple targets from the same object can be appropriately suppressed.

[0087] (8) The length of the width direction of the vehicle of the oncoming automobile is further longer than that of the oncoming bicycle. Therefore, when it is determined that the type of the object is an oncoming automobile, the range of the width direction of the vehicle or the range of the azimuth θ of the suppression range is further expanded. Thus, when the type of the object is an oncoming automobile, the situation of generating multiple targets from the same object can be appropriately suppressed.

[0088] (9) The situation where an observation value within the noise range outside the suppression range and having a difference between the power value P included in the observation value and the power value P included in the predicted value being equal to or greater than the power threshold is registered as a new target is suppressed. Thus, even for an observation value outside the suppression range, the situation where a noise peak near the object is registered as a new target can be suppressed.

[0089] (10) For the same predicted value, the observation values for which the registration as a new target is suppressed are observed from different positions of the same object as the observation value associated with the predicted value. Therefore, by holding multiple observation values for which the registration as a new target is suppressed, multiple positions within the same object can be detected by using the held multiple observation values, and the shape of the object can be estimated. That is, the observation values that are suppressed from being registered as new targets can be effectively utilized.

[0090] (Other embodiments)

[0091] As described above, the embodiments for implementing the present disclosure have been described, but the present disclosure is not limited to the above embodiments and can be implemented with various modifications.

[0092] (a) In the above-described embodiment, the first range and the suppression range are set as square ranges, but the shape of the range is not limited to a square. For example, the first range and the suppression range may be set as a range like the range R1 shown in Figure 13 or may be set as a circular range like the range R2 shown in Figure 14 .

[0093] (b) The object tracking device 20 and its method described in the present disclosure can also be implemented by a dedicated computer, which is provided by a processor and a memory configured to execute one or more functions embodied by a computer program. Alternatively, the object tracking device 20 and its method described in the present disclosure can be implemented by a dedicated computer provided by a processor constituted by one or more dedicated hardware logic circuits. Alternatively, the object tracking device 20 and its method described in the present disclosure can be implemented by one or more dedicated computers, which are constituted by a combination of a processor and a memory programmed to execute one or more functions and a processor constituted by one or more hardware logic circuits. In addition, the computer program can also be stored as instructions executed by a computer in a computer-readable non-transitory tangible recording medium. In the method of implementing the functions of each part included in the object tracking device 20, software is not necessarily required, and all of its functions can also be implemented using one or more hardware.

[0094] (c) A plurality of functions of one component in the above-described embodiment can be implemented by a plurality of components, or one function of one component can be implemented by a plurality of components. In addition, a plurality of functions of a plurality of components can be implemented by one component, or one function implemented by a plurality of components can be implemented by one component. In addition, a part of the configuration of the above-described embodiment can be omitted. In addition, at least a part of the configuration of the above-described embodiment can be added to or replaced with the configuration of other above-described embodiments.

[0095] (c) In addition to the above-described object tracking device 20, the present disclosure can also be implemented in various forms such as a system including the object tracking device 20 as a component, a program for causing a computer to function as the object tracking device 20, a non-transitory physical recording medium such as a semiconductor memory recording the program, and an object tracking method.

Claims

1. An object tracking device, in each predetermined processing cycle, estimates the state of at least one object, wherein: The object tracking device (20) comprises: A detection unit (21) configured to detect at least one observation value based on an observation signal observed by the sensor (10), wherein the at least one observation value is information about the at least one object mark around the vehicle; a prediction unit (22) configured to calculate a current predicted value of the state quantity based on a past estimated value of the state quantity for each object included in the at least one object quantity; A first range setting unit (23) configured to set a first range for each of the predicted values ​​calculated by the prediction unit based on the predicted value, the first range being a range estimated to be a range for acquiring the at least one observation value this time; a determination unit (24) configured to determine, for each of the predicted values ​​calculated by the prediction unit, an observed value within the first range set by the first range setting unit based on the at least one observed value detected by the detection unit, wherein the observed value is associated with the predicted value; an estimating unit (25) for calculating the estimated value of the current state quantity based on the observed value and the predicted value determined by the determining unit for each of the predicted values ​​calculated by the predicting unit; a registering unit (30) configured to register, as a new object marker, the observed value that is not associated with any of the predicted values ​​among the at least one observed value detected by the detecting unit; a suppression range setting unit (27) configured to set a suppression range for each of the predicted values ​​calculated by the prediction unit, the suppression range being a range of the observed value for suppressing registration of the new target object; and A registration suppressing unit (29) is configured to suppress the registration unit from registering, as the new object marker, the observation value within the suppression range set by the suppression range setting unit, among the at least one observation value detected by the detection unit.

2. The object tracking device according to claim 1, wherein: The suppression range setting unit is configured to set the first range as the suppression range.

3. The object tracking device according to claim 1, wherein: The predicted value includes a coordinate value of a first axis of the object, a coordinate value of a second axis of the object orthogonal to the first axis, a first speed of the object in the direction of the first axis, and a second speed of the object in the direction of the second axis, or a distance to the object, an orientation of the object, and a speed of the object. The first range setting unit is configured to set the first range based on at least one element included in the predicted value. The suppression range setting unit is configured to set the suppression range based on at least one element included in the predicted value.

4. The object tracking device according to claim 2, wherein: The predicted values include the coordinate values of the first axis of the target, the coordinate values of the second axis of the target orthogonal to the first axis, the first velocity in the direction of the first axis of the target, and the second velocity in the direction of the second axis of the target, or the distance to the target, the azimuth of the target, and the velocity of the target. The first range setting unit is configured to set the first range based on at least one element included in the predicted values. The suppression range setting unit is configured to set the suppression range based on at least one element included in the predicted values.

5. The object tracking device according to claim 3, wherein it further includes a determination unit (26), and the determination unit is configured to determine the type of the object indicated by the target. The suppression range setting unit is configured to set the suppression range according to the type of the object determined by the determination unit.

6. The object tracking device according to claim 5, wherein when the determination unit determines that the type of the object is a pedestrian, the suppression range setting unit sets, for the first velocity and the second velocity, or the velocity of the target, a second range that is wider than the first range as the suppression range.

7. The object tracking device according to claim 5, wherein the sensor is mounted on the vehicle, the first axis is the axis in the width direction of the vehicle, when the determination unit determines that the type of the object is a cross - riding bicycle, the suppression range setting unit sets, for the coordinate value of the first axis or the azimuth of the target, a third range that is wider than the first range as the suppression range.

8. The object tracking device according to claim 6, wherein the sensor is mounted on the vehicle, the first axis is the axis in the width direction of the vehicle, when the determination unit determines that the type of the object is a cross - riding bicycle, the suppression range setting unit sets, for the coordinate value of the first axis or the azimuth of the target, a third range that is wider than the first range as the suppression range.

9. The object tracking device according to claim 7, wherein when the determination unit determines that the type of the object is a cross - riding automobile, the suppression range setting unit sets, for the coordinate value of the first axis or the azimuth of the target, a fourth range that is wider than the third range as the suppression range.

10. The object tracking device according to any one of claims 1 to 9, wherein the state quantity includes the power value based on the observation signal of the sensor, it further includes a noise range setting unit (28), and the noise range setting unit is configured to set, for each of the predicted values calculated by the prediction unit, a noise range that is wider than the suppression range. The registration suppression unit is configured to suppress a situation in which an observation value outside the suppression range and within the noise range among the at least one observation value detected by the detection unit is registered as the new target by the registration unit, and the observation value is an observation value in which the power difference between the power value included in the observation value and the power value included in the predicted value is equal to or greater than a power threshold value.

11. The object tracking device according to any one of claims 1 to 9, wherein it further includes a shape estimation unit (31), and the shape estimation unit is configured to hold information of the observation value that is suppressed from being registered as the new target by the registration suppression unit for each predicted value calculated by the prediction unit in a subsequent processing cycle, and is configured to estimate the shape of the object indicated by the predicted value using the held observation value.

12. The object tracking device according to any one of claims 1 to 9, wherein the registration suppression unit is configured to stop suppressing a situation in which the observation value within the suppression range is registered as the new target when the distance to the target is equal to or less than a distance threshold value.

Images