JP3692910B2 - Lane tracking control device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention belongs to a technical field of a lane tracking travel control device that performs control for causing a vehicle to follow a forward lane based on control road information.
[0002]
[Prior art]
Conventionally, as a lane tracking travel control device, for example, a device described in Japanese Patent Laid-Open No. 5-297941 is known.
[0003]
The purpose of this publication is to make it possible to accurately detect a road shape from an expected shape based on map data or the like by a navigation system even when there are a plurality of actually measured shapes of a road obtained from a white line image. Technology that derives the predicted shape of the road by the processing means based on the navigation data such as the current position of the car, the driving direction, and map data, and selects the one close to the predicted shape from a plurality of actually measured shapes as the control road information Is described.
[0004]
[Problems to be solved by the invention]
However, in the above-described conventional lane tracking travel control device, when a plurality of road actual measured shapes are detected, the control runway information is based on the comparison with the predicted shape of the road derived based on the navigation data. Although it is difficult to capture the road shape ahead by the photographing means, the road shape is calculated not only when multiple road shapes are detected. In such a situation, it may be difficult to extract a white line from an image. In such a situation, the conventional road information obtained from the white line image is used as control road information. There is a problem that the technology cannot be dealt with and control must be interrupted.
[0005]
On the other hand, Japanese Patent Laid-Open No. 10-273062 discloses a configuration in which steering control is performed by estimating the current curvature radius using the curvature radius of the past road when the white line of the forward running road cannot be obtained. In particular, there is a suggestion that the curvature radius may be obtained from the map information of the navigation system instead of the estimation method using the curvature radius of the past road. However, in order to use the map information of the navigation system, it is essential that the current vehicle position is accurately recognized on the map data (matching process).
[0006]
Here, Japanese Patent Laid-Open No. 6-74778 discloses a configuration for recognizing an intersection when the white line on the front runway is interrupted and matching the current position to the nearest intersection in the map data. However, it is difficult to apply to roads where there are no intersections where white lines are interrupted, such as highways.
[0007]
The present invention has been made paying attention to the above-described problems, and the object of the present invention is to perform matching with the navigation means while the runway is recognized based on the image information, so that the image can be obtained. Provided is a lane tracking control device that enables lane tracking control to continue as it is by using highly accurate track information from navigation means as control road information even if the road cannot be recognized due to information. There is.
[0008]
[Means for Solving the Problems]
  In order to achieve the above object, according to the first aspect of the present invention, in the lane follow-up travel control device that performs control for causing the vehicle to follow the front lane based on the control road information,
  Road information detection means for detecting road information ahead of the vehicle based on the image information;
  Navigation means for determining the own vehicle traveling position on the map based on the map information and the traveling position information from the satellite, and outputting the traveling road information before and after the own vehicle traveling position;
  Map matching processing means for correcting the determined travel position of the host vehicle on the road in the map by comparing the map information in the navigation unit in consideration of the travel distance of the host vehicle according to the vehicle speed;
  After performing the map matching process, if it is determined that the image information is correctly recognized by the runway information detection unit,Road information comparing means for comparing the road information based on the image information obtained by the road information detecting means and the road information outputted from the navigation means;
  By the runway information comparison meansIn the comparison, so that the two track information matchesVehicle position information correcting means for correcting the vehicle position information on the map of the navigation means;
  Before the vehicle position information correction on the navigation means is completed, the road information based on the image information obtained in the road information detection means is selected as control road information, and the vehicle position information correction on the navigation means is selected. Is completed, and when it is determined that the road information obtained based on the image information from the road information detection means is not correctly recognized,Runway information obtained from the navigation meansTheControl runway information selecting means for selecting as control runway information;
  It is characterized by having.
[0009]
In the invention described in claim 2, in the lane following travel control device according to claim 1,
The runway information comparison unit is a unit that performs comparison based on runway curvature information among the runway information obtained from the runway information detection unit and the navigation unit.
[0010]
According to a third aspect of the present invention, in the lane following travel control device according to the first or second aspect,
The vehicle position information correction means is a means for correcting the position in the vehicle traveling direction in the vehicle position information on the navigation means.
[0014]
  Claim 4In the described invention, claims 1 toAny one of Claim 3In the lane following travel control device described in
  In the control runway information selection means, the runway information obtained from the navigation means is selected as the control runway information, and if the recognition of the image information does not return to normal by the runway information detection means over the set time, then the lane tracking Control stop means for stopping traveling control is provided.
[0015]
  Claim 5In the described invention, claims 1 toAny one of Claim 4In the lane following travel control device described in
  The control runway information selecting means is a means for preferentially selecting the runway information obtained from the navigation means as the control runway information when information indicating that the road approach point is approached from the navigation means. Features.
[0016]
  Claim 6In the described invention, claims 1 toAny one of Claim 5In the lane following travel control device described in
  After the vehicle position information correction on the navigation means is completed, if each of the road curvature information in the road information comparison means is greatly different, the road information based on the image information is regarded as abnormal and the lane tracking driving control is prohibited. A control prohibiting means is provided.
[0017]
Operation and effect of the invention
  In the first aspect of the invention, the road information detecting means detects the road information ahead of the vehicle based on the image information, and the navigation means detects the vehicle information on the map based on the map information and the driving position information from the satellite. As the vehicle travel position is determined, the road information before and after the vehicle travel position is output,In the map matching processing means, the determined own vehicle traveling position is corrected on the road in the map by comparing the map information in the navigation means in consideration of the traveling distance of the own vehicle by the vehicle speed,In the runway information comparison means,After performing the map matching process, if it is determined that the image information is correctly recognized by the road information detection means,The road information based on the image information obtained by the road information detection means is compared with the road information output from the navigation means, and the vehicle position information correction means uses the road information comparison means.In the comparison, so that the two track information matchesThe vehicle position information on the navigation means is corrected.The AndIn the control runway information selection means,Before the vehicle position information correction on the navigation means is completed, the road information based on the image information obtained by the road information detection means is selected as the control road information, and the vehicle position information correction on the navigation means is completed. Later, if it is determined that the road information obtained based on the image information from the road information detection means is not correctly recognized,Runway information obtained from navigation meansButSelected as control track informationThe AndThen, control for causing the vehicle to follow the front lane is performed based on the selected control track information.
[0018]
  Therefore, by matching with the navigation means while the road is recognized by the image information, even if the road cannot be recognized by the image information, the high-precision road information from the navigation means is controlled. By using the lane information, it is possible to continue the lane following traveling control as it is.
  In the control runway information selection means, before the vehicle position information correction on the navigation means is completed, the runway information based on the image information obtained by the runway information detection means is selected as the control runway information. By permitting selection from two pieces of track information on the condition that matching with is complete, it is possible to obtain highly accurate control track information at any time before and after the matching is completed.
  In the control runway information selection means, if it is determined that the runway information obtained based on the image information from the runway information detection means is not correctly recognized, the runway information obtained from the navigation means is used as the control runway information. Therefore, when the road is not recognized by the image information, the lane tracking travel control can be continued as it is by using the road information from the navigation means as the road information for control.
  In the runway information comparison unit, when it is determined that the image information is correctly recognized by the runway information detection unit, the runway information comparison process is executed, so the runway information from the runway information detection unit is accurate. By comparing the running road information only in certain cases, the vehicle position information on the navigation means can be matched with high accuracy.
[0019]
In the invention described in claim 2, in the runway information comparing means, the comparison is performed based on the runway curvature information among the runway information obtained from the runway information detecting means and the navigation means.
[0020]
Therefore, the comparison of the runway information is performed based on the runway curvature information, which is common information of the runway information detecting means and the navigation means, in the runway information.
[0021]
In the invention according to claim 3, the own vehicle position information correcting means corrects the own vehicle traveling direction position in the own vehicle position information on the navigation means.
[0022]
Therefore, when performing matching with the navigation means, the own vehicle traveling direction position, which is important for improving the lane tracking performance, is corrected based on the comparison of the road information.
[0029]
  Claim 4In the described invention, in the control stop means, the runway information obtained from the navigation means is selected as the control runway information in the control runway information selection means, and then the image information is detected in the runway information detection means over a set time. If the recognition of the vehicle does not return to normal, the lane tracking travel control is stopped.
[0030]
That is, when the image information recognition does not return to normal for a long time by the lane information detection means, the position accuracy is gradually lowered by the vehicle position correction not being performed in the navigation means.
[0031]
On the other hand, when the set time elapses, the lane tracking traveling control is stopped, so that inappropriate lane tracking control can be prevented from being executed based on inaccurate information.
[0032]
  Claim 5In the described invention, when the information indicating that the road approach point has been approached is output from the navigation means in the control road information selection means, the road information obtained from the navigation means is preferentially used as the control road information. Selected.
[0033]
That is, when the road information based on the image information is still selected as the control road information when traveling on the road branch point, the wrong side lane may be recognized and followed at the road branch point.
[0034]
On the other hand, when information indicating that the vehicle is approaching the road branch point is entered in advance by the navigation means, the wrong lane at the road branch point is recognized by using the road information obtained from the navigation means as the control road information. Thus, it is possible to prevent following.
[0035]
  Claim 6In the described invention, after the vehicle position information correction on the navigation means is completed in the control prohibiting means, if the respective road curvature information in the road information comparing means is greatly different, the road information based on the image information Is regarded as abnormal, and lane following travel control is prohibited.
[0036]
That is, in the image processing, when a character display on the road, dirt, an object on the road, or the like is mistakenly recognized as a road, it follows the wrongly recognized lane.
[0037]
On the other hand, abnormalities in the road information based on the image information are judged based on whether or not the respective road curvature information in the road information comparison means is greatly different. If the road information is largely different, the road information based on the image information is regarded as abnormal and the lane is followed. By prohibiting the travel control, it is possible to prevent the vehicle from following a lane that has been mistakenly recognized.
[0038]
DETAILED DESCRIPTION OF THE INVENTION
(Embodiment 1)
First, the configuration will be described.
[0039]
FIG. 1 is an overall system diagram showing a lane following travel control apparatus according to the first embodiment. 1 is a CCD camera provided to photograph the front in the traveling direction, and 2 is track information based on forward image information from the CCD camera 1. 3 is a vehicle speed sensor provided on a drive shaft, wheels, etc. for measuring the vehicle speed, 4 is a steering actuator provided in the steering mechanism 5, 6 is a steering angle sensor for measuring the steering angle, Reference numeral 7 denotes a GPS antenna for receiving radio waves from GPS satellites (GPS is an abbreviation for Global Positioning System), 8 is a navigation unit, and 9 is a control unit.
[0040]
Signals from the image processing device 2, the vehicle speed sensor 3 and the steering angle sensor 6 are input to the control unit 9, and a steering command calculated based on these signals is output to the steering actuator 4. In addition, signals are input to the navigation unit 8 from the GPS antenna 7 and the vehicle speed sensor 3. Further, the control unit 9 and the navigation unit 8 are connected by digital communication, and information calculated in the units 8 and 9 can be exchanged.
[0041]
FIG. 2 is a detailed view of the steering actuator portion. The steering mechanism 5 includes a steering wheel 5a, a steering shaft 5b, and a steering gear 5c. The steering actuator 4 includes a worm gear 4a, an electromagnetic clutch 4b, and a motor 4c. Configured. The motor 4c generates a control torque according to a command current output from the control unit 9, and the electromagnetic clutch 4b is connected / released according to a command from the control unit 9. When the electromagnetic clutch 4b is released, the control torque from the motor 4c is not applied to the steering mechanism 5, so that a normal steering mechanism without any control is obtained. Lane-following travel is performed by the command current to 4c.
[0042]
  FIG. 3 shows a block diagram of the image processing apparatus 2.
The image processing apparatus 2 receives image information in front of the vehicle by the CCD camera 1, extracts a lane information from the image information, and a lane shape estimation unit 2b that estimates a lane shape based on the extracted lane information. And a position for calculating deviation information of the vehicle position with respect to the lane center based on the estimated runway shapedeviationThe calculation unit 2c is configured to output calculation results (lateral deviation, yaw angle deviation, road curvature) in the road shape estimation unit 2b and the position deviation calculation unit 2c.
[0043]
Further, the lane extraction process includes a recognition state determination unit 2d that determines whether or not the road shape can be correctly recognized by the image according to the extraction state in 2a and the estimation state in the road shape estimation unit 2b. The recognition state information is output to the control unit 9 together with the running road shape and the calculation result.
[0044]
  FIG. 4 shows a block diagram of the navigation unit 8.
The navigation unit 8 includes a latitude / longitude calculation unit 8a that calculates the latitude / longitude of the vehicle based on satellite position / time information sent from the GPS antenna 7, a map information storage unit 8b that stores map information, Latitude / longitude information from the latitude / longitude calculation unit 8a, map information from the map information storage unit 8b, from the vehicle speed sensor 3ofA map matching processing unit 8c that performs map matching based on vehicle speed information and road curvature information (running road shape information) from the image processing apparatus 2, and a screen that displays a screen based on the vehicle position on the map determined by map matching A display unit 8 d is provided, and the road shape information and the map matching state information after the map matching process are output to the control unit 9.
[0045]
FIG. 5 shows a block diagram of the control unit 9.
The control unit 9 includes steering angle information from the steering angle sensor 6, vehicle speed information from the vehicle speed sensor 3, road curvature information (running road shape information) and map matching state information from the navigation unit 8, and road curvature from the image processing device 2. (Runway shape) and recognition state information are input. And the information selection part 9a which selects track shape information based on the map matching state information from the navigation unit 8, and the recognition state information from the image processing apparatus 2, and the target rudder angle based on the selected information and the present rudder angle A target steering angle calculation unit 9b that calculates a target steering torque, a target steering torque calculation unit 9c that calculates a target steering torque based on the target steering angle value and the current steering angle, and a motor that calculates a motor control current based on the steering torque calculation value It is configured by a driver 9d.
[0046]
Next, the operation will be described.
[0047]
[Processing in image processing device]
FIG. 6 shows a processing flowchart in the image processing apparatus 2. This subroutine is called by a timer interrupt at regular time intervals.
[0048]
In step S60, an image of the forward running road photographed by the CCD camera 1 is captured. This is, for example, an image as shown in FIG.
[0049]
In step S61, an edge extraction process is performed on the travel image.
[0050]
In step S62, the extracted edge is approximated to a polynomial to obtain an estimated shape of the runway.
[0051]
In step S63, the quality of the recognition state is determined according to the number of edges extracted in the edge extraction process and the change state of the estimated shape of the runway.
[0052]
In step S64, data is transmitted to the control unit 9.
[0053]
For example, in the edge extraction process for the road image shown in FIG. 7, when sufficient edges are detected to estimate the road shape as shown in FIG. 8, and the road shape is estimated as shown in FIG. It is conceivable that the number of edges is insufficient.
[0054]
As for the estimated road shape, as shown in FIG. 10, there is little change in the road shape and it is considered that continuity with the previous estimated shape is maintained, and as shown in FIG. It may be considered that the change in shape is large and continuity with the previous estimated shape is not maintained.
[0055]
In step S63 (recognition state determination unit 2d), a comprehensive determination of the recognition state is performed based on these states. FIG. 12 shows an example in which the determination state is divided into three stages. Here, the three-stage classification is
State A: Can be used for both control and navigation correction.
State B: Can be used for control. Cannot be used for navigation correction.
State C: Cannot be used for both control and navigation correction.
Is expressed.
[0056]
In step S <b> 64, the road curvature based on the estimated shape of the running path estimated in the above process, the left and right of the vehicle with respect to the lane position, the yaw angle deviation, and the recognition state information are transmitted to the control unit 9.
[0057]
[Processing in the navigation unit]
FIG. 13 shows a process flowchart in the navigation unit 8. This subroutine is called by a timer interrupt at regular time intervals.
[0058]
In step S130, the vehicle speed is captured, and in step S131, a signal sent from a GPS satellite is received. In the next step S132, the road curvature information calculated in the image processing apparatus 2 and the recognition state information in the image processing apparatus 2 are received by the control unit 9. Receive via.
[0059]
In the latitude / longitude calculation processing in step S133, the distance between the satellite and the own vehicle is determined by comparing the transmission time information sent from the GPS satellite and the current time information (not shown) in the navigation unit 8. The vehicle position is calculated by measuring and integrating distance information from a plurality of satellites.
[0060]
In the map matching process in step S134, the calculated vehicle position is corrected on the road in the map by comparing the map information in the navigation unit 8 in consideration of the traveling distance of the vehicle according to the vehicle speed. Prevents the location from being off-road and increases the accuracy of the vehicle location information.
[0061]
In step S135, after performing the map matching process, it is determined whether or not the correction process based on the curvature is performed based on the received recognition state information in the image processing apparatus 2. If the recognition state is the above-described state A, the image information can be used for navigation correction. Therefore, the process proceeds to the next step S136, and correction processing based on curvature information is performed. Does not perform correction processing based on curvature. If the latter state continues for a predetermined time, the correction completion flag is cleared to 0 in step S140.
[0062]
In the correction process based on the curvature information in step S136, position correction using the road curvature information calculated in the image processing device 2 is performed. This is shown in FIG. In FIG. 14, after the map matching process, the vehicle position is recognized as a position on the road in the map. The road shape (a) before and after the vehicle in the map is compared with the road curvature information (b) obtained by image processing, and the vehicle position is corrected so as to match (c).
[0063]
In this correction process, when a matching vehicle position is found in the vicinity of the vehicle position after the map matching, the process proceeds to step S138, the correction completion flag is set to 1, and a state where no correction is found continues for a predetermined time In step S139, the correction completion flag is cleared to zero. If it is not found, the position is not corrected.
[0064]
Based on the vehicle position obtained in the above procedure, road shape information ahead of the vehicle is acquired from the map, and in step S141, this subroutine is completed by transmitting it to the control unit 9 together with the correction completion flag.
[0065]
[Processing in the control unit]
FIG. 15 shows a process flowchart in the control unit 9. This subroutine is called by a timer interrupt at regular time intervals.
[0066]
In step S150, the steering angle is read. In step S151, communication processing with the image processing apparatus 2 and the navigation unit 8 is performed. In communication with the image processing device 2, road curvature, left and right of the vehicle with respect to the lane position, yaw angle deviation, and recognition state information are received. In communication with the navigation unit 8, road curvature information received from the image processing device 2, While transmitting recognition state information, it receives road shape information and a correction completion flag.
[0067]
In the next step S152, it is determined whether or not the control start SW (not shown) is set to 0N. If it is set to 0N, a clutch connection process is performed, and then a subsequent control command calculation is performed. If not, the process proceeds to step S161 to perform control stop processing. In the control stop process, if the steering torque is suddenly reduced to 0 when the control is currently being performed, the driver feels uncomfortable. Therefore, the steering torque is gradually decreased and the electromagnetic clutch 4b is released after this becomes zero. Stop control. Further, when the control is not being executed and the electromagnetic clutch 4b is released, nothing is performed in this process.
[0068]
Hereinafter, processing when the control start SW is 0N will be described.
In the clutch connection process in step S153, when the control start SW changes from 0FF to 0N, a clutch connection command is output. If the control has already started, the electromagnetic clutch 4b is connected, so nothing is done here.
[0069]
In the next step S154, a determination based on the recognition state information in the image processing apparatus 2 is performed. If the state is the state A or the state B, the next target rudder angle calculation process is performed as it is. Proceeding to step S156, the correction completion flag is determined.
[0070]
In the target rudder angle calculation process in step S155, the target rudder angle δ0 is calculated based on the rudder angle δ, the lateral position deviation Y calculated by the image processing device 2, the yaw angle deviation ψ, and the curvature ρ of the front road surface. The target rudder angle is calculated by a method of giving corrections of amounts proportional to the lateral position deviation and its differential value, yaw angle deviation, and road surface curvature with respect to the current rudder angle δ.
[0071]
δ0 = δ + K1 × Y + K2 × (dY / dt) + K3 × ψ + K4 × ρ
Here, K1 to K4 are proportional constants, respectively.
[0072]
In step S159, after calculating the target steering angle, an appropriate target steering torque T for making the steering angle coincide with the target value is calculated by the following equation.
[0073]
T = K5 × δ0 + K6 × δ
Here, K5 and K6 are proportional constants, respectively.
[0074]
In step S160, the calculated target steering torque is output as a control command, and the process ends.
[0075]
If it is determined in step S154 that the recognition state in the image processing apparatus 2 is state C, the process proceeds to step S156, and determination based on the correction completion flag in the navigation unit 8 is performed. Here, when the correction completion flag is 1, the process proceeds to step S157, and the target steering angle δ0 is calculated by the following formula based on the road shape information transmitted from the navigation unit 8.
[0076]
δ0 = δ + K5 × ρ
Here, K5 is a proportionality constant.
[0077]
Thereafter, similarly to the case where the lateral position deviation or the like in the image processing apparatus 2 is used, the process proceeds to step S159, the target steering torque is calculated, the process proceeds to step S160, and this is output as a control command.
[0078]
On the other hand, if the correction completion flag is 0 in the determination in step S156, the accuracy of the road shape information in the navigation unit 8 is not sufficient, so the process proceeds to step S158 and the control stop process is performed.
[0079]
In this control stop process, if the steering torque is suddenly reduced to 0 when the control is currently being performed, the driver feels uncomfortable, so the steering torque is gradually reduced and the electromagnetic clutch 4b is released after this becomes zero. To stop control.
[0080]
As described above, according to the processing of the first embodiment, the vehicle position information in the navigation unit 8 is corrected using the road curvature information obtained by the image processing device 2, and highly accurate vehicle position information is obtained. Therefore, when the recognition state in the image processing device 2 is not good, it becomes possible to perform lane tracking using the road shape information obtained from the navigation unit 8 and obtain a control effect under a wider range of conditions. Is possible.
[0081]
Also, if the recognition state in the image processing device 2 does not return well even after a predetermined time has elapsed, the vehicle position correction in the navigation unit 8 is not performed, so the position accuracy gradually decreases. Since the control completion process is executed in the control unit 9 by clearing the correction completion flag to 0, inappropriate control based on inaccurate information can be prevented.
[0082]
In the first embodiment, an example in which the recognition state has three levels has been described. However, this can be classified from a minimum of two levels to more levels according to the determination method. Even in such a case, when it is determined that the image processing apparatus 2 is surely recognizing, the curvature information is used for the vehicle position correction in the navigation unit 8.
[0083]
Further, the recognition state in the image processing apparatus 2 includes the number of extracted edges and the change state of the road shape. However, the determination method is not limited to this, and the strength of the edge and the color information of the white line position. It is also possible to use the degree of coincidence in the left and right white line directions.
[0084]
Further, it is possible to further improve the determination accuracy by determining whether the road shape change is appropriate while predicting the traveling direction of the own vehicle using the vehicle speed and the steering angle information.
[0085]
Furthermore, in the first embodiment, the example in which the steering actuator 4 provided with the motor 4c and the electromagnetic clutch 4b is added to the steering mechanism 5 as the actuator is shown. However, electric power steering can also be used as the actuator. is there.
[0086]
(Embodiment 2)
FIG. 16 shows a process flowchart in the control unit 9 of the second embodiment.
[0087]
In the second embodiment, in the communication process between the navigation unit 8 and the control unit 9, in addition to the information in the first embodiment, information that the road junction is approaching is transmitted from the navigation unit 8 to the control unit 9. This is used for judgment in the arithmetic processing in the control unit 9.
[0088]
In the second embodiment, the clutch connection process is performed in the same procedure as in the first embodiment, and then a determination is made based on the correction completion flag and the branch point approach information.
[0089]
That is, if it is determined in step S162 that the correction completion flag = 1 and the branch point is approaching, the process proceeds to step S163, and is transmitted from the navigation unit 8 regardless of the recognition state of the image processing apparatus. Based on the road shape information, the target rudder angle δ0 is calculated by the following equation.
[0090]
δ0 = δ + K5 × ρ
Here, K5 is a proportionality constant.
[0091]
On the other hand, if the above determination is not established in step S162, the processing is performed based on the procedure of proceeding to step S154, as in the first embodiment.
[0092]
As described above, according to the second embodiment, the target rudder is unconditionally based on the road shape information when the vehicle position in the navigation unit 8 is corrected and the branch point is approaching. Since the corner is calculated, the wrong side lane is not recognized at the branching point and is not followed.
[0093]
(Embodiment 3)
FIG. 17 shows a processing flowchart in the image processing apparatus 2 according to the third embodiment.
[0094]
In the third embodiment, in the processing of the image processing device 2, road shape information ahead of the host vehicle and the correction completion flag are received from the navigation unit 8, and the road shape information and the road shape information based on the image processing are greatly different. In addition, it is determined that the image processing information is inappropriate.
[0095]
In the third embodiment, as in the case of the first embodiment, the process proceeds from step S60 to step S61 to step S62 in FIG.
[0096]
Thereafter, in step S65, the road shape information is received from the navigation unit 8 before the recognition state is determined. In step S66, it is determined whether the correction completion flag is 1. If the correction completion flag is 1, Proceeding to step S67, it is determined whether or not the road shape obtained from the navigation unit 8 matches the road information based on the image processing. If the road information is largely different, the process proceeds to step S68, where state C is determined.
[0097]
On the other hand, if the correction completion flag is 0 in step S66, and if the road shape obtained from the navigation unit 8 in step S67 and the road information based on the image processing substantially match, as in the first embodiment, Processing after the recognition state determination in step S63 is performed.
[0098]
In the edge extraction process of step S61, in addition to the road white line, character display on the road, dirt, roadside objects, and the like may be erroneously extracted to estimate a road shape different from the actual road shape. is there.
[0099]
However, according to the third embodiment, when the vehicle position in the navigation unit 8 is corrected, the road shape obtained from the navigation unit 8 and the road shape based on the image processing are compared and greatly different. Is determined to be an image that is not used for control, and in such a case, there is no false tracking.
[Brief description of the drawings]
FIG. 1 is an overall system diagram showing a lane following travel control device in a first embodiment.
FIG. 2 is a detailed view showing a steering actuator in the first embodiment.
3 is a block diagram illustrating an image processing apparatus according to Embodiment 1. FIG.
FIG. 4 is a block diagram showing a navigation unit in the first embodiment.
FIG. 5 is a block diagram showing a control unit in the first embodiment.
FIG. 6 is a flowchart illustrating a process in the image processing apparatus according to the first embodiment.
FIG. 7 is an example of a road image.
FIG. 8 is an example diagram illustrating an edge extraction state in a road image.
FIG. 9 is another example diagram showing an edge extraction state in a road image.
FIG. 10 is an example of a recognized road shape.
FIG. 11 is another example of a recognized road shape.
FIG. 12 is an example of recognition state determination.
FIG. 13 is a diagram showing a processing flowchart in the navigation unit of the first embodiment.
FIG. 14 is an example of position correction based on curvature information.
FIG. 15 is a diagram illustrating a processing flowchart in the control unit according to the first embodiment;
FIG. 16 is a diagram showing a processing flowchart in the control unit of the second embodiment.
FIG. 17 is a diagram illustrating a processing flowchart in the image processing apparatus according to the third embodiment.
[Explanation of symbols]
1 CCD camera
2 Image processing device
3 Vehicle speed sensor
4 Steering actuator
5 Steering mechanism
6 Rudder angle sensor
7 GPS antenna
8 Navigation unit
9 Control unit

Claims (6)

In a lane tracking travel control device that performs control for tracking the vehicle in the front lane based on the control road information,
Road information detection means for detecting road information ahead of the vehicle based on the image information;
Navigation means for determining the own vehicle traveling position on the map based on the map information and the traveling position information from the satellite, and outputting the traveling road information before and after the own vehicle traveling position;
Map matching processing means for correcting the determined travel position of the host vehicle on the road in the map by comparing the map information in the navigation unit in consideration of the travel distance of the host vehicle according to the vehicle speed;
After performing the map matching process, if it is determined that the image information is correctly recognized by the road information detection means, the road information based on the image information obtained by the road information detection means, A road information comparison means for comparing the road information output from the navigation means;
In the comparison by the road information comparison means , the vehicle position information correction means for correcting the vehicle position information on the map of the navigation means so that two pieces of road information match ,
Before the vehicle position information correction on the navigation means is completed, the road information based on the image information obtained in the road information detection means is selected as control road information, and the vehicle position information correction on the navigation means is selected. Is completed, and when it is determined that the road information obtained based on the image information from the road information detection means is not correctly recognized, the road information obtained from the navigation means is used as the control road. Control road information selection means to select as information,
A lane-following travel control device comprising: In the lane following travel control device according to claim 1,
A lane tracking travel control device characterized in that the travel path information comparison means is a means for performing comparison based on travel path curvature information among the travel path information obtained from the travel path information detection means and the navigation means. In the lane following travel control device according to claim 1 or 2,
The vehicle position information correction means is a means for correcting the position in the vehicle traveling direction in the vehicle position information on the navigation means. In the lane following travel control device according to any one of claims 1 to 3,
In the control runway information selection means, the runway information obtained from the navigation means is selected as the control runway information, and if the recognition of the image information does not return to normal by the runway information detection means over the set time, then the lane tracking A lane tracking travel control device, characterized in that a control stop means for stopping travel control is provided . In the lane following travel control device according to any one of claims 1 to 4,
The control runway information selecting means is a means for preferentially selecting the runway information obtained from the navigation means as the control runway information when information indicating that the road approach point is approached from the navigation means. A lane following travel control device. In the lane following travel control device according to any one of claims 1 to 5,
After the vehicle position information correction on the navigation means is completed, if each of the road curvature information in the road information comparison means is greatly different, the road information based on the image information is regarded as abnormal and the lane tracking driving control is prohibited. A lane following travel control device, characterized in that a control prohibiting means is provided .

Images