JP2018092483A - Object recognition device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an object recognition device capable of reducing an erroneous detection of a target of a roadside object as a target within a lane of a preceding vehicle or the like.SOLUTION: A control device (object recognition device) of an automatic driving system 1 detects an object existing in the vicinity of the own vehicle by a LIDAR (Laser Imaging Detection Ranging) and also by a millimeter wave radar. A lane marker adjacent to a road boundary is recognized, and a LIDAR's target detected by the LIDAR is narrowed down to an area inside the lane boundary line. Then, only when the radar's target detected by the millimeter wave radar overlaps an in-lane region of the LIDAR's target, it is determined that an object specified by the in-lane region of the LIDAR's target and the radar's target is an object located in the lane.SELECTED DRAWING: Figure 1

Description

  The present invention relates to an object recognition apparatus suitable for use in an automatic driving system.

  Patent Document 1 discloses a method in which an image target detected based on information from an image sensor and a radar target detected based on information from a radar are collated and output as a fusion target if they are identical. Is disclosed.

JP 2006-266927 A

  The objects to be detected by the object recognition device include roadside objects such as road walls and guardrails. By detecting the roadside object, the position and shape of the road boundary can be recognized. However, road boundaries that are actually captured by the image sensor, such as road walls and joints between guardrails and grass covering guardrails, are complex. For this reason, when a roadside object is converted into a target based on information from the image sensor, a target frame that represents the position and range of the target may protrude beyond the lane boundary line into the lane.

  Such a problem is not limited to the case where an image sensor is used to detect an object. This is also a problem that may occur when an object existing around the host vehicle is detected using a rider (LIDAR: Laser Imaging Detection and Ranging). According to the rider, it is possible to measure the position and shape of an object existing around the host vehicle by performing scanning with a laser beam in the horizontal direction and the vertical direction. However, due to the influence of the complicated shape of the road boundary and the influence of noise, the target frame is set so that the target frame crosses the lane boundary line even though it is a roadside object outside the lane boundary line. May end up. If a roadside target is erroneously detected as a target in a lane such as a preceding vehicle, the control of the vehicle by the automatic driving system may be adversely affected.

  The present invention has been made in view of the above-described problems, and an object thereof is to provide an object recognition device that can reduce erroneous detection of a roadside target as a target in a lane. To do.

In order to achieve the above object, an object recognition apparatus according to the present invention includes:
First target detecting means for detecting a first target by a first sensor that measures the position and shape of an object existing around the host vehicle;
Second target detection means for detecting a second target by a second sensor for measuring the position of an object existing around the host vehicle;
Lane boundary recognition means for recognizing a lane boundary adjacent to a road boundary;
Based on the positional relationship between the first target and the lane boundary line, a lane area extraction means for extracting a lane area that is inside the lane boundary line of the first target;
When the second target overlaps the lane area, a determination unit that determines that the object specified by the second target and the lane area is an object inside the lane boundary line;
It is characterized by providing.

  According to the object recognition device of the present invention, the first target detected by the first sensor that measures the position and shape of the object protrudes into the lane even though it is an object outside the lane boundary. Even if the target is a shape, the range of the first target is narrowed down to the area inside the lane boundary, and the second target detected by the second sensor is selected for the narrowed range. By collating, erroneous detection of the object as an object inside the lane boundary line can be suppressed.

1 is a block diagram showing a configuration of an automatic driving system according to a first embodiment. 3 is a flowchart illustrating a function of a lane in / outside area determination unit according to the first embodiment. 6 is a diagram illustrating a method of detecting an intersection position by a lane in / outside area determination unit according to Embodiment 1. FIG. 6 is a diagram for explaining a method for extracting an in-lane region by an in-and-outside region determining unit according to Embodiment 1. FIG. 3 is a flowchart illustrating functions of a consistency determination unit according to the first embodiment. FIG. 10 is a diagram for explaining a correspondence method by a consistency determination unit according to the first embodiment. 6 is a diagram for describing a consistency determination method by a consistency determination unit according to Embodiment 1. FIG. FIG. 10 is a diagram for explaining the effect of the object recognition process of the first embodiment. FIG. 10 is a diagram for describing advantages over the comparative example of the object recognition process of the first embodiment. FIG. 10 is a diagram for describing advantages over the comparative example of the object recognition processing according to the first embodiment. 6 is a flowchart illustrating functions of a lane in / outside area determination unit according to a second embodiment. FIG. 10 is a diagram for explaining a method of lane area synthesis by a lane in / outside area determination unit according to the second embodiment. 10 is a flowchart illustrating functions of a consistency determination unit according to the third embodiment. 10 is a flowchart illustrating functions of a lane in / outside area determination unit according to a fourth embodiment. It is a figure explaining the candidate limitation method by the lane inside / outside area determination part of Embodiment 4. FIG.

  Embodiments of the present invention will be described below with reference to the drawings. However, in the embodiment shown below, when referring to the number of each element, quantity, quantity, range, etc., unless otherwise specified or clearly specified in principle, the reference However, the present invention is not limited to these numbers. Further, the structures described in the embodiments described below are not necessarily essential to the present invention unless otherwise specified or clearly specified in principle.

Embodiment 1
FIG. 1 is a diagram illustrating a configuration of an automatic driving system according to an embodiment. The automatic driving system performs automatic driving of a vehicle on which the automatic driving system is mounted. The automatic driving system 1 includes a control device 10 that functions as an object recognition device and an autonomous recognition sensor. The autonomous recognition sensor includes a rider 2, a millimeter wave radar 3, and a stereo camera 4. These autonomous recognition sensors are connected to the control device 10 directly or via a communication network such as a CAN (Controller Area Network) built in the vehicle. The control device 10 can grasp the situation of substantially the entire periphery of the vehicle from the object information obtained by the autonomous recognition sensor.

  The automatic driving system 1 further includes a GPS (Global Positioning System) receiver 5, a map database 6, a navigation system 7, an HMI (Human Machine Interface) 8, and an actuator 9. The GPS receiver 5 is means for acquiring position information indicating the position of the host vehicle based on a signal transmitted from a GPS satellite. The map database 6 is formed in a storage such as an HDD or an SSD mounted on the vehicle, for example. The map information included in the map database 6 includes road position information, road shape information, intersection and branch point position information, road lane information, and the like. The navigation system 7 calculates the route traveled by the host vehicle based on the position information of the host vehicle measured by the GPS receiver 5 and the map information in the map database 6, and the calculated route information is transmitted via the HMI 8. While transmitting to a driver | operator, it outputs to the control apparatus 10. The HMI 8 is an interface for outputting and inputting information between the occupant and the control device 10. The actuator 9 is a device that operates according to an operation signal from the control device 10 and changes the traveling state of the vehicle by the operation. The actuator 9 is provided in each of a drive system, a braking system, and a steering system, for example. In addition to these, the automatic driving system 1 is provided with an internal sensor for obtaining information on the traveling state of the host vehicle, such as a vehicle speed sensor and an acceleration sensor.

  The control device 10 is an ECU (Electronic Control Unit) having at least one CPU, at least one ROM, and at least one RAM. The ROM stores various types of data including various programs and maps for automatic operation. Various functions are realized in the control device 10 by loading a program stored in the ROM into the RAM and executing it by the CPU. In addition, the control apparatus 10 may be comprised from several ECU.

  In FIG. 1, among the functions for automatic driving that the control device 10 has, in particular, functions related to object recognition are represented by blocks. The illustration of other functions for automatic operation of the control device 10 is omitted.

  The control device 10 has a function of recognizing an object existing in the vicinity of the host vehicle and discriminating between an object located in the lane and an object located outside the lane. This function is realized by the runway estimation unit 11, the first target detection unit 12, the second target detection unit 13, the lane inside / outside region determination unit 14, and the consistency determination unit 15 that the control device 10 includes. As will be described in detail later, the lane estimation unit 11 corresponds to the lane boundary recognition unit according to the present invention, the first target detection unit 12 corresponds to the first target detection unit according to the present invention, and the second target. The detection unit 13 corresponds to a second target detection unit according to the present invention, the lane inside / outside region determination unit 14 corresponds to a lane region extraction unit according to the present invention, and the consistency determination unit 15 includes a determination unit according to the present invention. It corresponds to. These units 11, 12, 13, 14, and 15 do not exist as hardware in the control device 10, but are realized by software when a program stored in the ROM is executed by the CPU.

  The travel path estimation unit 11 estimates the position and posture of the host vehicle from the GPS signal received by the GPS receiver 5. Based on the estimated position and orientation of the host vehicle, map data around the host vehicle is searched from the map database 6. The map data includes information on the running road, specifically, information on the position and shape of the lane boundary line that divides the lane. The runway estimation unit 11 coordinates the position and shape of the lane boundary line around the host vehicle from the GPS coordinate system to a reference coordinate system centered on the host vehicle. Thereby, the position of each lane boundary line with respect to the host vehicle and the shape of each lane boundary line are estimated.

  The first target detection unit 12 acquires distance measurement data of an object around the host vehicle by the rider 2 as a first sensor. The distance measurement data acquired by the rider 2 includes information regarding the distance from the rider 2 to the distance measurement point, information regarding the height of the distance measurement point, and information regarding the direction of the distance measurement point. The first target detection unit 12 clusters the acquired distance measurement data. Clustering is performed based on the position and height of each distance measuring point on the reference coordinate system. Further, it is performed so as to maintain continuity with the result of clustering in the previous frame. The specific method of clustering is not particularly limited, and a known clustering method can be used. The clustered distance measuring point group is surrounded by a rectangular target frame in the reference coordinate system to form one target. Hereinafter, this target is referred to as a rider target or a first target.

  The second target detection unit 13 acquires distance measurement data of an object around the host vehicle by the millimeter wave radar 3 as a second sensor. The distance measurement data acquired by the millimeter wave radar 3 includes information regarding the distance from the millimeter wave radar 3 to the distance measurement point and information regarding the direction of the distance measurement point. In the object detection by the millimeter wave radar 3, the object is detected when the reflected millimeter wave can be received. Each time a reflected millimeter wave is received, one target is obtained. Hereinafter, this target is referred to as a radar target or a second target. The aforementioned rider target is a target having information on the position and shape, whereas the radar target is a target having only information on the position.

  The position and shape of each lane boundary estimated by the lane estimation unit 11 and the position and shape of the rider target detected by the first target detection unit 12 are input to the lane inside / outside region determination unit 14. . Based on the position and shape of each lane boundary line and the position and shape of the rider target, the lane inside / outside area determination unit 14 determines which area of the rider target is an area inside the lane boundary and which area is It is determined whether the area is outside the lane boundary. Here, when the function which the lane inside / outside area | region determination part 14 has is represented with a flowchart, it will become like FIG. As shown in FIG. 2, the lane in / outside area determination unit 14 performs intersection position detection as the first process, and performs lane area extraction as the next process. Details of each process will be described below with reference to the drawings.

  FIG. 3 is a diagram for explaining a method of detecting an intersection position by the lane in / outside area determination unit 14. In FIG. 3, a pair of lane boundary lines that form a lane, two rider targets, and two radar targets are drawn in a reference coordinate system centered on the subject vehicle. Here, it is assumed that the road on which the vehicle travels is a one-lane road, and the road boundary is in contact with both sides of the pair of lane boundary lines. In the intersection position detection process, the intersection position between the lane boundary line adjacent to the road boundary (both lane boundary lines in FIG. 3) and the frame line of the rider target is detected. When the rider target crosses the lane boundary line, there are two intersection positions for each rider target. In the intersection position detection process, a margin is drawn on the inner side of the lane boundary line, and a boundary line for determination (curved line in FIG. 3) is drawn to detect the intersection position of the boundary line for determination and the rider target. You may make it do.

  FIG. 4 is a diagram for explaining a method of extracting the in-lane area by the in-and-outside area determining unit 14. In the lane region extraction process, the rider target is divided into two at the intersection position detected in the intersection position detection process, and the region on the side that is in the lane is extracted. Specifically, as indicated by an arrow in FIG. 4, the vehicle proceeds in a predetermined direction (for example, clockwise direction) from one intersection position along the border line of the rider target, and the other intersection position is searched. When the other intersection position is searched, a closed region is formed by a line segment connecting the two intersection positions. A region where all the vertices are in the lane, such as a closed region indicated by a dotted line in FIG. 4, is extracted as the lane region. If the margin is to be taken as described above, the margin is increased as the expected accuracy of the lane boundary is lower in order to prevent the area outside the lane boundary from being erroneously determined as the inner area. May be.

  When a plurality of rider targets are detected by the first target detection unit 12, the lane in / outside region determination unit 14 performs the above-described intersection position detection process and lane region extraction for each rider target. Thereby, the lane area is extracted for each rider target. In addition, when all the regions of the rider target are inside the lane boundary line, the entire rider target is extracted as the lane region.

  The consistency determination unit 15 includes the position and shape of the in-lane region of the rider target extracted by the lane inside / outside region determination unit 14 and the position and shape of the radar target detected by the second target detection unit 13. Is entered. The consistency determination unit 15 determines the consistency between two types of targets, that is, the radar target and the rider target, based on the position and shape of the lane area of the rider target and the position and shape of the radar target. Judge about. Here, when the function which the lane inside / outside area | region determination part 14 has is represented with a flowchart, it will become like FIG. As shown in FIG. 5, the consistency determination unit 15 performs association as the first process, and performs consistency determination as the next process. Details of each process will be described below with reference to the drawings.

  FIG. 6 is a diagram for explaining a matching method by the consistency determination unit 15. In the association process, the in-lane area of the rider target is associated with the radar target. When a radar target in the lane area of a certain rider target enters, it is said that “both correspond as the same object”, and a radar target in the lane area of a certain rider target does not enter , “They do not correspond as the same thing”. In the example shown in FIG. 6, since the left radar target RA is not included in the lane area LA of the left rider target, the lane area LA and the left radar target RA are associated with each other as the same object. No. On the other hand, since the right radar target RB is contained in the lane region LB of the right rider target, the lane region LB and the left radar target RB are associated with each other as the same object. In the association processing, a margin may be provided inside the lane area of the rider target, and the area that is limited by the margin and the area limited to the lane area may be associated with the rider target.

  FIG. 7 is a diagram illustrating a method for determining consistency by the consistency determining unit 15. In the consistency determination process, only the area corresponding to the radar target is extracted from the lane area of the rider target. The lane area of the rider target extracted by the consistency determination process is a fusion target duplicated by the radar target, and is output as a final target for automatic driving. The position of the extracted rider target area in the lane becomes the final target position, and the shape of the extracted rider target area in the lane (in the example shown in FIG. 7, the shape of the range surrounded by the dotted line). Is the final target shape.

  In the consistency determination process, the area in the lane of the rider target that has not been associated with the radar target is not output as the final target. That is, even if a target is detected in the lane by the rider 2, it is not used as a final target for automatic driving if it is not supported by the detection result of the radar 3.

  Since the respective units 11, 12, 13, 14, and 15 having the above functions are provided, the control device 10 functions as an object recognition device. The control device 10 as the object recognition device narrows down the rider target detected by the rider 2 to the area inside the lane boundary, and the radar target detected by the millimeter wave radar 3 overlaps the lane area of the rider target. Only the object specified by the lane area of the rider target and the radar target is determined to be an object in the lane.

  When the roadside object that forms the road boundary is converted into a target based on the distance measurement data of the rider 2, as shown in FIG. 8A, the rider target that should be a roadside object crosses the lane boundary line. It may stick out inside. However, according to the object recognition process by the control device 10, as shown in FIG. 8B, the range of the rider target is narrowed down to a region inside the lane boundary line. Then, the radar target detected by the millimeter wave radar 3 is collated with the narrowed range. The lane area of the rider target that could not be duplicated by the radar target is not established as a fusion target. For this reason, erroneous detection of the roadside object as an object inside the lane boundary line can be suppressed.

  As a data processing method for suppressing false detection, a method of removing in advance the distance measurement data obtained outside the lane boundary line from the distance measurement data of the rider 2 (hereinafter referred to as a comparative example method) is considered. It is done. In the example shown in FIG. 9, five rider targets A to E are obtained by clustering the distance measuring points detected by the rider 2 (points indicated by black circles in FIG. 9). Among these, the target corresponding to the object in the lane is only the rider target B. By removing the ranging points detected outside the lane boundary line and narrowing the clustering target to the ranging points inside the lane boundary line, only the target rider target B can be obtained. However, in the method of the comparative example, since it is necessary to determine whether all the ranging points are inside or outside the lane boundary line, a large amount of calculation processing is required and a large load is imposed on the control device. On the other hand, according to the object recognition processing of the first embodiment, the number of data that must be determined inside or outside the lane boundary line is greatly increased by targeting before narrowing the distance measuring points. Less. Therefore, the object recognition processing method according to the present embodiment is superior to the method of the comparative example in that the amount of calculation processing can be reduced to reduce the calculation load of the control device.

  Further, the method of the comparative example has a problem that it is easily influenced by an error in the position of the lane boundary line. For example, when a boundary is drawn with a margin inside the lane boundary as shown by a dotted line in FIG. 10, the distance measuring points detected inside the lane boundary are also removed. For this reason, in the example shown in FIG. 10, not only the rider targets A, C, D, and E but also the rider target B that is the target cannot be obtained. On the other hand, when the boundary line is drawn with a margin outside the lane boundary line as shown by the alternate long and short dash line in FIG. 10, ranging points detected outside the lane boundary line cannot be removed. For this reason, in the example shown in FIG. 10, not only the target rider target B but also the rider targets A and D are acquired as targets for automatic driving.

  As can be seen from the comparison with the method of the comparative example, according to the object recognition processing method of the first embodiment, by targeting before narrowing down the data, the target target is deleted. Can be prevented. Further, by evaluating the lane boundary line and the radar target detected by the millimeter wave radar 3 as a set, only the target to be removed can be removed. In addition, since all objects located outside and inside the lane boundary line are tracked as targets, there is an advantage that the influence of errors can be reduced by integrating with past determination results.

Embodiment 2. FIG.
Similar to the automatic driving system of the first embodiment, the automatic driving system of the present embodiment has the configuration shown in FIG. The difference between the automatic driving system of the second embodiment and the automatic driving system of the first embodiment is in the function of the lane inside / outside area determination unit 14.

  FIG. 11 is a flowchart illustrating the function of the lane in / outside area determination unit 14 according to the second embodiment. As shown in FIG. 11, the lane inside / outside area determination unit 14 according to the second embodiment detects the intersection position as the first process, extracts the lane area as the next process, and further combines the lane area as the next process. I do. The contents of the intersection position detection process and the lane region extraction process are the same as those of the first embodiment. Therefore, the details of the in-lane region composition process, which is a characteristic process of the second embodiment, will be described below with reference to the drawings.

  FIG. 12 is a diagram for explaining a method for synthesizing the in-lane region by the in-lane / outer region determining unit 14 according to the second embodiment. In the in-lane region composition process, the in-lane region of the rider target extracted by the in-lane region extraction process is acquired for a plurality of times. The obtained extraction results at a plurality of times are overlaid, and the degree of overlap is calculated. Specifically, a score of 1 is given to an area extracted as an in-lane area within a rider target frame at a certain time t. A score of 1 is also given to the area extracted as the in-lane area at the next time t + 1. Thus, giving a score of 1 to the region extracted as the lane region at each time is continued for a predetermined time. The regions extracted as the lane region at each time are overlapped, and the score given to the region is added. As a result, the total score increases as the number of times extracted as the in-lane region in the frame of the rider target increases.

  In the lane region synthesis process, a region having a total score of a certain value or more is extracted as a final lane region. For example, in the example illustrated in FIG. 12, a region having a total score of 5 or more is extracted as a lane region. If there is an error in the position of the lane boundary line, the shape of the region extracted as the lane region by the lane region extraction process changes depending on the error. However, by performing the processing as described above, it is possible to reduce the influence of the error in the position of the lane boundary line on the determination of the in-lane region.

Embodiment 3 FIG.
Similar to the automatic driving system of the first embodiment, the automatic driving system of the present embodiment has the configuration shown in FIG. The difference between the automatic driving system of the third embodiment and the automatic driving system of the first embodiment is in the function of the consistency determination unit 15.

  FIG. 13 is a flowchart illustrating functions of the consistency determination unit 15 according to the third embodiment. As shown in FIG. 13, the consistency determination unit 15 of the third embodiment performs association as the first process, performs consistency determination as the next process, and further performs consistency determination integration as the next process. The contents of the associating process and the consistency determining process are the same as those of the first embodiment. Therefore, details of the consistency determination integration process, which is a characteristic process of the second embodiment, will be described below.

  In the consistency judgment integration process, the judgment results of consistency / non-consistency at each time are integrated. Specifically, the reliability of the consistency determination is calculated based on the determination result at each time, and when the reliability is equal to or higher than a threshold, it is output as a final target for automatic operation. As an example of the calculation method of reliability, count the number of times corresponding to the radar target for each area in the lane of the rider target, and the ratio of the number of times corresponding to the radar target to the total number of judgments as the reliability You may calculate. The reliability in this case is higher as the ratio value is larger. Alternatively, the distance between the center of gravity of the lane area of the rider target and the radar target may be calculated, and the average value of the distance may be calculated as the reliability. The reliability in this case is higher as the average value of the distance is smaller. Alternatively, a normal distribution having the distance between the center of gravity of the lane area of the rider target and the radar target as an input may be obtained, and parameters of the normal distribution (for example, standard deviation) may be calculated as the reliability. For example, when the standard deviation is a parameter, the reliability is higher as the standard deviation is smaller.

  An error in the position of the lane boundary affects the determination result of whether or not there is consistency by the consistency determination process. However, by performing the consistency determination integration process described above, the influence of the error in the position of the lane boundary line is reduced, so that the accuracy of the determination of whether the object is in the lane or the object outside the lane can be improved.

Embodiment 4 FIG.
Similar to the automatic driving system of the first embodiment, the automatic driving system of the fourth embodiment has the configuration shown in FIG. The difference between the automatic driving system of the fourth embodiment and the automatic driving system of the first embodiment is in the function of the lane inside / outside area determination unit 14.

  FIG. 14 is a flowchart illustrating the function of the lane in / outside area determination unit 14 according to the fourth embodiment. As shown in FIG. 14, the lane in / outside area determination unit 14 according to the fourth embodiment performs candidate limitation as the first process, detects the intersection position as the next process, and further extracts the lane area as the next process. . The contents of the intersection position detection process and the lane region extraction process are the same as those of the first embodiment. Therefore, details of the candidate limiting process, which is a characteristic process of the fourth embodiment, will be described below with reference to the drawings.

  FIG. 15 is a diagram for explaining a candidate limitation method by the lane in / outside area determination unit 14 according to the fourth embodiment. In the candidate limitation process, the direction of the rider target and the direction of the lane are compared, and a rider target whose direction does not coincide with the lane is extracted as a candidate for subsequent processing. Specifically, as shown in FIG. 15A, the lane having the shortest distance from the center of gravity of the rider target is extracted. The distance from the center of gravity of the rider target to the lane means the shortest distance from the center of gravity of the rider target to the lane center line. The position and shape of the lane center line are calculated from the position and shape of a pair of lane boundary lines constituting the lane. Next, as shown in FIG. 15B, the direction of the lane is extracted with respect to the lane having the shortest distance from the center of gravity of the rider target. Specifically, the point on the lane centerline that provides the shortest distance to the center of gravity of the rider target is identified, and the direction of the lane centerline at that point, that is, the direction of the tangent line that touches the lane centerline at that point is acquired. . Then, as shown in FIG. 15C, the direction of the rider target is acquired, and the angle formed by the direction of the rider target and the direction of the lane is calculated. The direction of the rider target means the long direction of the target frame represented by a rectangle.

  If the rider target is not oriented in the direction along the lane, the rider target is likely to straddle the lane boundary. Therefore, in the candidate limitation process, if the angle formed by the direction of the rider target and the direction of the lane is greater than or equal to a predetermined value, the rider target is extracted as a candidate for subsequent processes. In other words, a rider target whose angle with the direction of the lane is less than a predetermined value is excluded from candidates for subsequent processing. By limiting the processing candidates in this way, it is possible to reduce the amount of calculation processing and suppress the calculation load of the control device.

Other embodiments.
While the embodiments of the present invention have been described above, as described above, the present invention is not limited to these embodiments. For example, the first camera that measures the position and shape of an object present around the host vehicle may be the stereo camera 4. When a millimeter wave radar having a short wavelength is mounted on the vehicle, it may be used as the first sensor.

  Means for estimating the position and orientation of the host vehicle are not limited to GPS signals. For example, the position and orientation of the host vehicle may be estimated by associating feature points obtained from the output of the rider 2 or the stereo camera 4 with feature points such as signs and white lines registered on the map.

  Acquisition of lane boundary information is not limited to acquisition from the map database 6. For example, the front of the host vehicle may be photographed by the stereo camera 4 and the lane boundary information may be acquired by white line recognition processing for the obtained photographed image.

  Note that the function of the consistency determination unit 15 of the third embodiment can also be applied to the automatic driving system of the second embodiment. The function of the lane in / outside area determination unit 14 of the fourth embodiment can be applied to the automatic driving systems of the second and third embodiments.

1 Automatic driving system 2 Rider (first sensor)
3 Millimeter wave radar (second sensor)
4 Stereo camera 5 GPS receiver 6 Map database 7 Navigation system 8 HMI
9 Actuator 10 Control device (object recognition device)
11 Runway estimation unit (lane boundary recognition means)
12 1st target detection part (1st target detection means)
13 Second target detection unit (second target detection means)
14 Lane inside / outside area determination unit (lane area extracting means)
15 Consistency determination unit (determination means)

Claims (1)

First target detecting means for detecting a first target by a first sensor that measures the position and shape of an object existing around the host vehicle;
Second target detection means for detecting a second target by a second sensor for measuring the position of an object existing around the host vehicle;
Lane boundary recognition means for recognizing a lane boundary adjacent to a road boundary;
Based on the positional relationship between the first target and the lane boundary line, a lane area extraction means for extracting a lane area that is inside the lane boundary line of the first target;
When the second target overlaps the lane area, a determination unit that determines that the object specified by the second target and the lane area is an object inside the lane boundary line;
An object recognition apparatus comprising:

Images