Target recognition processing device

Abstract

The purpose of the present invention is to provide a target recognition processing device capable of accurately estimating the reliability of target information received from another vehicle and thereby handling the target information with appropriate priority. For this purpose, this target recognition processing device 100 is provided with: a reception unit 110 that receives at least one set of first target information that is transmitted from at least one of the other vehicles 302 traveling around a host vehicle 300, and that relates to a target 301 recognized by the other vehicle 302; a first target processing unit 122 that calculates a first trajectory 602 of the target 301 on the basis of the first target information, and calculates the reliability of the first trajectory 602; and a second target processing unit 160 that sets the priority of the first target information according to the reliability of the first trajectory 602, and processes the first target information according to the priority.

Description

Target recognition processing device

[0001] The present invention relates to a target recognition processing device that recognizes targets present around a vehicle.

[0002] The rapid development of autonomous driving and driver assistance technologies has made a significant contribution to reducing accidents, such as collisions between vehicles or between vehicles and pedestrians.

[0003] To achieve even more advanced autonomous driving and driving assistance, it is necessary to avoid collisions with targets that cannot be recognized by external sensors installed on the vehicle alone (other vehicles, pedestrians, roadside objects, and other objects whose presence must be recognized while driving), and to take driving actions that coordinate with the movement of such targets. For example, targets that are outside the recognition range of external sensors or that exist in blind spots such as other vehicles or buildings are targets that the vehicle cannot recognize.

[0004] In order to obtain information about targets that cannot be recognized by the vehicle itself, as described above, the application of information sharing technology using V2X (Vehicle to Anything) communication is being considered. Depending on the communication partner, V2X communication may be called V2V (Vehicle to Vehicle, communication with other vehicles), V2I (Vehicle to Infrastructure, communication with road equipment such as traffic lights), V2N (Vehicle to Network, communication with a cloud network), or V2P (Vehicle to Pedestrian, communication with communication equipment held by pedestrians).

[0005] Through V2X, it is possible to obtain information held by the communication partner, such as the communication partner's own location information and target information recognized by the communication partner using sensors, etc. This makes it possible to obtain information from the communication partner about targets in positions that cannot be recognized by the vehicle itself.

[0006] In this case, the reliability of the target information obtained from the communication partner is not guaranteed. Therefore, it is useful to estimate the reliability of the information obtained, select information according to the reliability, or combine information by weighting the information with higher reliability.

[0007] Patent Document 1 discloses a method of estimating position errors in target information acquired through communication and assigning a high reliability to information with small errors.

[0008] International Publication No. 2020 / 189069

[0009] However, the method disclosed in Patent Document 1 may result in erroneous decisions on driving assistance or autonomous driving actions when the reliability of the movement direction or speed of the obtained target information is not high. For example, when the trajectory of the target is estimated from the time progression of the target information obtained through communication, the trajectory may be unstable compared to the original trajectory even when the position error is small. In this case, the movement direction of the target may be estimated incorrectly, and the possibility of a collision with the target may also be estimated incorrectly.

[0010] The present invention has been made in consideration of the above-mentioned problems, and its purpose is to provide a target recognition processing device that can handle target information received from other vehicles with appropriate priority by estimating the reliability of the target information with high accuracy.

[0011] In order to achieve the above object, the present invention comprises a receiving unit that receives at least one first target information item related to a target recognized by at least one other vehicle traveling around the host vehicle, the first target information item being transmitted from the other vehicle; a first target processing unit that calculates a first trajectory of the target based on the first target information item and calculates a reliability of the first trajectory; and a second target processing unit that sets a priority of the first target information item according to the reliability of the first trajectory and processes the first target information according to the priority.

[0012] According to the present invention, the reliability of target information received from other vehicles can be estimated with high accuracy, and therefore the target information can be handled with appropriate priority.

[0013] Functional block diagram of the target recognition processing device. Diagram showing a first example of the processing result of the target recognition processing device. Diagram showing a second example of the processing result of the target recognition processing device. Diagram showing a third example of the processing result of the target recognition processing device.

[0014] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. In each drawing, the same reference numerals are used to designate the same components, and redundant description will be omitted.

[0015] 1 is a functional block diagram of a target recognition processing device 100 according to this embodiment. The target recognition processing device 100 is mounted on a host vehicle 300 (shown in FIG. 2, etc.), processes information (target information) such as the position and speed of targets present in the vicinity of the host vehicle 300, and passes the information to a downstream application 200 (a driving assistance function such as emergency braking, or an automatic driving function such as lane changing). The target recognition processing device 100 includes a calculation device such as a CPU, storage devices such as a ROM and RAM, an input / output interface for inputting and outputting signals to and from external devices, and realizes each functional unit by executing a program stored in the ROM, etc.

[0016] The target recognition processing device 100 includes a V2X receiving unit 110, an external sensor 111, a GPS 112, an inertial measurement unit (IMU) 113, a plurality of transmitting terminal processing units 120, an external sensor processing unit 130, a self-position estimation unit 140, a V2X target association unit 150, a plurality of fusion target processing units 160 (second target processing units), and a target information fusion unit 170.

[0017] The transmitting terminal processing unit 120 has a target track association unit 121 and a plurality of target processing units (first target processing units) 122. The target processing unit 122 has a trajectory estimation unit 123, a driving model compatibility determination unit 124, and a map information compatibility determination unit 125.

[0018] The external sensor processing unit 130 includes a target track association unit 131 and a plurality of target processing units (third target processing units) 132. The target processing unit 132 includes a trajectory estimation unit 133, a driving model compatibility determination unit 134, and a map information compatibility determination unit 135.

[0019] The fusion target processing unit 160 has an own vehicle / V2X coincidence determination unit 161 and an information rejection determination unit 162.

[0020] The V2X receiver 110 receives target information from other vehicles, pedestrians, infrastructure facilities, and cloud networks (hereinafter, each of these will be referred to indistinguishably as a terminal) via communication. The target information includes position information of the target, and may also include speed, acceleration, yaw rate, etc. The received target information includes at least one of information about targets recognized by other terminals and information about the terminal itself.

[0021] The V2X receiving unit 110 separates the information for each terminal that transmitted the data and passes the information to the transmitting terminal processing unit 120 corresponding to each terminal. The target recognition processing device 100 is equipped with L transmitting terminal processing units 120 and can support up to L terminals as V2X communication partners. The transmitting terminal processing unit 120 estimates the trajectory of the target from target information acquired in the past and stores the estimated trajectory as a target track.

[0022] The target track association unit 121 determines which of the stored target tracks the target information delivered from the V2X receiving unit 110 corresponds to, and associates the target information with the corresponding target track. For example, it estimates the target position at the current time from the target track information, and searches for the target track that is closest to the target information delivered from the V2X receiving unit 110. If no target track can be associated with the target information, a new target track is generated. Next, target processing is performed for each target track. The transmitting terminal processing unit 120 has M1 target processing units 122 and can handle a maximum of M1 targets.

[0023] The trajectory estimation unit 123 estimates the trajectory of the movement of the target based on past target position information of the target track, etc. The trajectory may be estimated by directly using target information previously acquired by V2X, or by applying filter processing such as a Kalman filter.

[0024] A travel model is set in advance in the travel model compatibility determination unit 124. Examples of the travel model include a constant acceleration motion model and a constant velocity circular motion model. The travel model compatibility determination unit 124 fits the target trajectory estimated by the trajectory estimation unit 123 to the travel model to estimate parameter values ​​such as speed, and selects the trajectory that is closest to the target trajectory from the trajectory that conforms to the travel model as the estimated trajectory. The travel model compatibility determination unit 124 then sets the travel model compatibility reliability according to the degree of agreement between the estimated trajectory and the target trajectory. That is, the higher the degree of agreement between the estimated trajectory and the target trajectory, the higher the travel model compatibility reliability is set. One example of a method for calculating the degree of agreement between the estimated trajectory and the target trajectory is to determine the positions of both trajectories at the same time, calculate the distance or the square of the distance between them, calculate the average value over the trajectories, and use the reciprocal of the average value as the degree of agreement. Note that if the target information includes sensor accuracy information, the reliability may be adjusted according to the sensor accuracy. For example, if the sensor accuracy is low, the reliability may be reduced by multiplying the obtained reliability by a coefficient smaller than 1. Furthermore, if the target information includes sensor detection range information and the target position is near the boundary of the detection range, the reliability may be reduced by multiplying the obtained reliability by a coefficient smaller than 1. This is because detection accuracy generally decreases when the target position is near the boundary of the sensor detection range.

[0025] The map information compatibility determination unit 125 acquires map information about the area surrounding the traveling position. In this embodiment, it is assumed that the map information is stored in advance or acquired by means of communication or the like. If the target is a vehicle, the target should follow a trajectory that is close to the shape of the lane stored on the map. Therefore, the map compatibility reliability can be calculated from the degree of agreement between the trajectory on the map and the estimated trajectory of the target.

[0026] The target processing unit 122 stores the target track information for the next processing, and adds the obtained estimated trajectory, driving model compatibility reliability, and map compatibility reliability to the target information and passes it on to the V2X target association unit 150.

[0027] The external sensor processing unit 130 recognizes external targets using external sensors 111, such as cameras, mounted on the host vehicle 300. The external sensor processing unit 130 estimates the trajectory of a target from previously acquired target information and stores the estimated trajectory as a target track. The target track association unit 131 determines to which of the stored target tracks the target information delivered from the external sensor 111 corresponds, and associates the target information with the corresponding target track. If multiple external sensors 111 are provided, information fusion between the sensors, such as sensor fusion, may be performed. Next, target processing is performed for each target track. The external sensor processing unit 130 has M0 target processing units 132 and can handle a maximum of M0 targets. The processing of the trajectory estimation unit 133, the driving model compatibility determination unit 134, and the map information compatibility determination unit 135 is similar to the processing of the trajectory estimation unit 123, the driving model compatibility determination unit 124, and the map information compatibility determination unit 125 of the transmitting terminal processing unit 120, so explanations will be omitted.

[0028] The self-position estimation unit 140 estimates the self-position of the vehicle 300 using the GPS 112 and the inertial measurement unit (IMU) 113, and estimates the trajectory of the vehicle 300 from this information.

[0029] The V2X target association unit 150 receives all target information from the transmitting terminal processing unit 120 and the external sensor processing unit 130, determines which targets are the same target, and associates the target information of the same target. Information about the host vehicle 300, target information recognized by the host vehicle 300 using the external sensor 111, information about the other terminal transmitting information via V2X, and target information recognized by the other terminal may be associated as the same target, and a target associated in this manner is referred to as a fused target. The target recognition processing device 100 has N fused target processing units 160 and can handle a maximum of N fused targets.

[0030] A fused target may have multiple pieces of information associated with it. Since each piece of associated information holds an estimated trajectory, the consistency of the information between the multiple information sources can be determined from the degree of agreement between the estimated trajectories. This is called fusion reliability. As in the case of driving model adaptation, one example of a method for calculating the degree of agreement between the estimated trajectories is to, for example, determine the positions of both trajectories at the same time, calculate the distance or the square of the distance between them, calculate the average value over the trajectories, and use the reciprocal of the average value as the degree of agreement.

[0031] The information rejection determination unit 162 can reject target information from a specific information source using the driving model compatibility reliability, map compatibility reliability, and fusion reliability as judgment criteria. For example, thresholds may be set in advance for the driving model compatibility reliability and the map compatibility reliability, and target information from the information source may be rejected if the reliability is lower than the threshold. A threshold may also be set for the fusion reliability, and both pieces of target information to be fused may be rejected if the fusion reliability falls below the threshold. Alternatively, if one piece of information matches the estimated result of the vehicle's own position 300 or is recognized by the external sensor 111 of the vehicle 300, the other piece of target information may be rejected. Alternatively, the driving model compatibility reliability and the map compatibility reliability of both pieces of target information may be compared, and the target information with the lower reliability may be rejected.

[0032] The target information fusion unit 170 can perform a rejection decision for information in larger units based on all reliability information. For example, for each transmitting terminal, the average reliability of all targets can be obtained and used as the reliability of the transmitting terminal itself. In this case, if the reliability of a transmitting terminal is lower than a predetermined threshold, it can be determined that there is a problem with the recognition of that transmitting terminal, and all target information acquired from that transmitting terminal can be rejected.

[0033] The target information fusion unit 170 passes the target information that has not been rejected in the above process to the subsequent application 200 (a driving assistance function such as emergency braking or an automatic driving function such as lane change). When target information from multiple information sources is associated in the fused target, various reliability levels may be further examined and only target information with high reliability may be adopted, or the target information may be synthesized by a method such as simple averaging or weighted averaging in which weights are set according to reliability.

[0034] 2 is a diagram showing a first example of a processing result of the target object recognition processing device 100. A host vehicle 300 equipped with the target object recognition processing device 100 is traveling on a road 400. Other vehicles 301 and 302 are traveling on a road 401 that intersects with the road 400 toward an intersection. The other vehicles 301 and 302 are in the blind spot of a building 500 near the intersection of the roads 400 and 401. The other vehicle 302 traveling behind the other vehicle 301 has acquired target information of the other vehicle 301. The host vehicle 300 is capable of V2X communication with the other vehicle 302, but is unable to perform V2X communication with the other vehicle 301. The target object recognition processing device 100 acquires target information of the other vehicle 301 from the other vehicle 302 via V2X.

[0035] The target recognition processing device 100 calculates a trajectory 602 of the other vehicle 301 based on the target information received from the other vehicle 302, and calculates an estimated trajectory 700 of the other vehicle 301 using a travel model. Then, the travel model compatibility reliability of the trajectory 602 is calculated based on the deviation of the trajectory 602 from the estimated trajectory 700. In the example shown in FIG. 2 , the trajectory 602 meanders, and there is a large deviation from the linear estimated trajectory 700 calculated using the travel model, resulting in a low travel model compatibility reliability. As a result, the target information of the other vehicle 301 received from the other vehicle 302 is treated as a low priority. In this case, the accuracy of the position and speed of the other vehicle 301 included in the target information is low, and there is a risk of the host vehicle 300 colliding with the other vehicle 302. Therefore, the host vehicle 300 can take collision avoidance action, such as reducing its speed. In addition, if the target recognition processing device 100 also acquires other target information from other vehicles 302, the priority of the other target information may be set according to the reliability of the trajectory calculated based on the other target information, or may be set low in conjunction with the target information of other vehicles 301.

[0036] 3 is a diagram showing a second example of the processing result of the target object recognition processing device 100. The host vehicle 300 equipped with the target object recognition processing device 100 is traveling on a road 400. Other vehicles 301 and 302 are traveling on a road 401 that intersects with the road 400 toward an intersection. Other vehicle 303 is traveling in the oncoming lane of the host vehicle 300, and other vehicle 304 is traveling in the oncoming lane of the other vehicles 301 and 302. The other vehicles 301 and 302 are in the blind spot of a building 500 near the intersection of the roads 400 and 401. The other vehicles 302 to 304 have acquired target information of the other vehicle 301. The host vehicle 300 is capable of V2X communication with the other vehicles 302 to 304, but is unable to perform V2X communication with the other vehicle 301. The target recognition processing device 100 acquires target information of another vehicle 301 from other vehicles 302 to 304 via V2X.

[0037] The target recognition processing device 100 calculates trajectories 602 to 604 of the other vehicle 301 based on the target information received from the other vehicles 302 to 304. Then, it calculates the fusion reliability of the trajectories 602 to 604 according to the mutual deviation of the trajectories 602 to 604. In the example shown in FIG. 3 , the trajectories 602 and 603 are linear, while the trajectory 604 meanders. Therefore, the deviation between the trajectory 602 and the trajectory 603 is small, and the deviation between the trajectory 604 and the trajectory 602 or the trajectory 603 is large. Therefore, the fusion reliability of the trajectories 602 and 603 is high, and the fusion reliability of the trajectory 604 is low. As a result, the target information of the other vehicle 301 received from the other vehicle 304 is treated as having a lower priority than the target information of the other vehicle 301 received from the other vehicles 302 and 303. If the target recognition processing device 100 also acquires other target information from the other vehicle 304, the priority of the other target information may be calculated according to the reliability of the trajectory calculated based on the other target information, or may be set lower in conjunction with the target information of the other vehicle 304. Furthermore, if multiple target information pieces are acquired from the other vehicle 304, and the proportion of target information pieces with reliability lower than a predetermined value in the multiple target information pieces is greater than a predetermined proportion, the priorities of all of the multiple target information pieces may be lowered.

[0038] 4 is a diagram showing a third example of the processing result of the target object recognition processing device 100. The host vehicle 300 equipped with the target object recognition processing device 100 is traveling on a road 403. Other vehicles 301 and 302 are traveling on a road 404 that merges with the road 403 toward the junction. The other vehicles 301 and 302 are in the blind spot of a building 500 near the junction. The other vehicle 302 traveling behind the other vehicle 301 has acquired target information of the other vehicle 301. The host vehicle 300 is capable of V2X communication with the other vehicle 302, but is unable to perform V2X communication with the other vehicle 301. The target object recognition processing device 100 acquires target information of the other vehicle 301 from the other vehicle 302 via V2X.

[0039] The target object recognition processing device 100 calculates a trajectory 602 of the other vehicle 301 based on target object information received from the other vehicle 301, calculates an estimated trajectory 700 of the other vehicle 301 using a travel model, and calculates an estimated trajectory 800 of the other vehicle 301 using map information. Then, the device sets a travel model compatibility reliability of the trajectory 602 according to the deviation of the trajectory 602 from the estimated trajectory 700, and sets a map information compatibility reliability of the trajectory 602 according to the deviation of the trajectory 602 from the estimated trajectory 800. In the example shown in FIG. 4 , the trajectory 602 meanders and has a large deviation from the estimated trajectory 700 calculated using the travel model, resulting in a low travel model compatibility reliability. On the other hand, the estimated trajectory 800 calculated using the map information meanders along the road 404, and has a small deviation from the estimated trajectory 700, resulting in a high map information compatibility reliability. Therefore, if the travel model compatibility reliability is set to be more important than the map information compatibility reliability, the target object information of the other vehicle 301 received from the other vehicle 302 is treated with a higher priority. As a result, even if the travel model compatibility reliability is low, the host vehicle 300 can appropriately make a collision avoidance decision by adopting the target object information received from the other vehicle 302. Note that it is possible to appropriately change which of the travel model compatibility reliability, the map information compatibility reliability, and the fusion reliability is to be given more importance.

[0040] (Summary) In this embodiment, the target recognition processing device 100 includes a receiving unit 110 that receives at least one first target information item related to a target 301 recognized by at least one other vehicle 302 traveling around the host vehicle 300, the first target information item being transmitted from the at least one other vehicle 302; a first target processing unit 122 that calculates a first trajectory 602 of the target 301 based on the first target information item and calculates the reliability of the first trajectory 602; and a second target processing unit 160 that sets a priority of the first target information item in accordance with the reliability of the first trajectory 602 and processes the first target information in accordance with the priority.

[0041] According to the present embodiment configured as described above, the reliability of the target information received from the other vehicle 302 can be estimated with high accuracy, and the target information can be handled with appropriate priority.

[0042] In this embodiment, the first target processing unit 122 calculates the reliability of each of the trajectories 602 to 604 of the target 301 based on the plurality of target information including the first target information transmitted from the plurality of other vehicles 302 to 304 including the other vehicle 302, and the second target processing unit 160 sets the priority of each of the plurality of target information in descending order of the reliability. This makes it possible to handle the plurality of target information received from the plurality of other vehicles 302 to 304 with appropriate priorities.

[0043] In this embodiment, the first target processing unit 122 calculates the reliability of each of the trajectories of the plurality of targets, including the target 301, based on the plurality of target information including the first target information transmitted from the other vehicle 302, and the second target processing unit 160 lowers the priority of all of the plurality of target information when the proportion of target information with the reliability lower than a predetermined value in the plurality of target information is greater than a predetermined proportion. This makes it possible to handle all of the plurality of target information with a low priority when the overall reliability of the plurality of target information acquired from the other vehicle 302 is low.

[0044] Furthermore, in this embodiment, the first target processing unit 122 calculates the estimated trajectory 700 of the target 301 using a traveling model, which is a motion model of the vehicle, and the second target processing unit 160 sets the reliability of the first target information to be higher the smaller the deviation of the first trajectory 602 from the estimated trajectory 700. This makes it possible to estimate the reliability of the target information received from the other vehicle 302 with high accuracy based on the traveling model, which is a motion model of the vehicle.

[0045] In this embodiment, the first target processing unit 122 calculates the estimated trajectory 800 of the target 301 based on map information about the vicinity of the host vehicle 300, and the second target processing unit 160 sets the reliability of the first target information to be higher the smaller the deviation of the first trajectory 602 from the estimated trajectory 800. This makes it possible to estimate the priority of the target information received from the other vehicle 302 with high accuracy based on the map information about the vicinity of the host vehicle 300.

[0046] In this embodiment, the first target processing unit 122 calculates target trajectories 602 to 604 based on a plurality of pieces of target information including the first target information transmitted from a plurality of other vehicles 302 to 304 including the other vehicle 302, and the second target processing unit 160 extracts, from the plurality of pieces of target information, two or more pieces of target information (received from the other vehicles 302 and 303 in FIG. 3 ) whose deviations from the trajectories 602 to 604 are equal to or less than a predetermined value, and sets the priority (fusion reliability) of each of the two or more pieces of target information higher than the priority of other target information (received from the other vehicle 304 in FIG. 3 ) excluding the two or more pieces of target information from the plurality of pieces of target information. This makes it possible to set the reliability of each of the plurality of pieces of target information received from the plurality of other vehicles 302 to 304 by majority vote.

[0047] Furthermore, in this embodiment, the target recognition processing device 100 includes an external sensor 111 mounted on the host vehicle 300, and a third target processing unit 132 that calculates a second trajectory of the target 301 based on second target information acquired by the external sensor 111 and calculates the reliability of the second trajectory, and when the second target processing unit 160 determines that the reliability of the first trajectory 602 is lower than the reliability of the second trajectory, it sets the priority of the first target information lower than the priority of the second target information. This makes it possible to set the priority of target information received from other vehicles 302 to 304 based on the reliability of the target information acquired by the external sensor 111.

[0048] In this embodiment, the first target processing unit 122 determines that the reliability of the first trajectory 602 is lower than the reliability of the second trajectory when the deviation of the first trajectory 602 from the second trajectory is greater than a predetermined value, thereby making it possible to appropriately compare the reliability of the first trajectory 602 with the reliability of the second trajectory.

[0049] Although the embodiments of the present invention have been described in detail above, the present invention is not limited to the above-described embodiments and can be modified and implemented in various ways. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and the present invention is not necessarily limited to those having all of the described configurations.

[0050] 100... target recognition processing device, 110... V2X receiving unit, 111... external sensor, 113... IMU, 120... transmitting terminal processing unit, 121... target track association unit, 122... target processing unit (first target processing unit), 123... trajectory estimation unit, 124... driving model compatibility determination unit, 125... map information compatibility determination unit, 130... external sensor processing unit, 131... target track association unit, 132... target processing unit (third target processing unit), 133... trajectory estimation unit, 134... driving model compatibility determination unit, 135...map information compatibility determination unit, 140...self-position estimation unit, 150...target association unit, 160...fusion target processing unit (second target processing unit), 161...own vehicle / V2X match determination unit, 162...information rejection determination unit, 170...target information fusion unit, 200...application, 300...own vehicle, 301...other vehicle (target), 302-304...other vehicles, 400-402...road, 500...building, 600...estimated trajectory, 602-604...trajectory (first trajectory), 700, 800...estimated trajectory.

Claims

1. A target recognition processing device, comprising: a receiving unit that receives at least one piece of first target information regarding a target recognized by at least one other vehicle traveling around the host vehicle, the first target information being transmitted from the at least one other vehicle; a first target processing unit that calculates a first trajectory of the target based on the first target information and calculates a reliability of the first trajectory; and a second target processing unit that sets a priority of the first target information according to the reliability of the first trajectory and processes the first target information according to the priority.

2. The target recognition processing device according to claim 1, wherein the first target processing unit calculates reliabilities of trajectories of the target based on a plurality of pieces of target information including the first target information, the plurality of pieces of target information being transmitted from a plurality of other vehicles including the other vehicle, respectively, and the second target processing unit sets the priorities of the plurality of pieces of target information to be higher in descending order of the reliabilities.

3. The target recognition processing device according to claim 1, wherein the first target processing unit calculates reliabilities of trajectories of a plurality of targets including the target based on a plurality of pieces of target information including the first target information, the plurality of pieces of target information being transmitted from the other vehicle, and the second target processing unit lowers all the priorities of the plurality of pieces of target information when a ratio of target information having a reliability lower than a predetermined value in the plurality of pieces of target information is larger than a predetermined ratio.

4. The target recognition processing device according to claim 1, wherein the first target processing unit calculates an estimated trajectory of the target using a driving model which is a motion model of the vehicle, and the second target processing unit sets the reliability of the first target information to be higher as a deviation between the first trajectory and the estimated trajectory is smaller.

5. The target recognition processing device according to claim 1, wherein the first target processing unit calculates an estimated trajectory of the target based on map information around the host vehicle, and the second target processing unit sets the reliability of the first target information to be higher as a deviation between the first trajectory and the estimated trajectory is smaller.

6. In the object recognition processing apparatus according to claim 1, the first object processing unit calculates the trajectory of the object based on a plurality of object information including the first object information respectively transmitted from a plurality of other vehicles including the other vehicle, and the second object processing unit extracts two or more pieces of object information from the plurality of object information such that the deviation of the trajectory is equal to or less than a predetermined value, and sets the priority of each of the two or more pieces of object information to be higher than the priority of other object information excluding the two or more pieces of object information from the plurality of object information. An object recognition processing apparatus characterized by the above.

7. In the object recognition processing apparatus according to claim 1, it includes an external sensor mounted on the host vehicle, and a third object processing unit that calculates a second trajectory of the object based on the second object information acquired by the external sensor and calculates the reliability of the second trajectory. When the second object processing unit determines that the reliability of the first trajectory is lower than the reliability of the second trajectory, the second object processing unit sets the priority of the first object information to be lower than the priority of the second object information. An object recognition processing apparatus characterized by the above.

8. In the object recognition processing apparatus according to claim 7, the first object processing unit determines that the reliability of the first trajectory is lower than the reliability of the second trajectory when the deviation of the first trajectory with respect to the second trajectory is greater than a predetermined value. An object recognition processing apparatus characterized by the above.

Images