JP4225242B2 - Travel path recognition device - Google Patents

Description

  The present invention relates to a travel path recognition device that estimates travel path parameters in a travel path on which a vehicle travels.

Conventionally, as described in JP-A-2002-197444, as a device for estimating a travel path parameter, a plurality of lane marker candidate points are set on a captured image obtained by capturing a travel path of a vehicle, and the lane marker candidate points are set. There is known a method for calculating a parameter of a traveling road based on the coordinate value of. This device determines whether or not the captured image is an image that can detect a lane marker based on a deviation between a prediction candidate point predicted from a parameter of a traveling road and a lane marker candidate point, and determines that the image is not an image that can detect a lane marker. In this case, the parameter is not updated and the previous value is used to recognize the travel route.
JP 2002-197444 A

  However, this device has a problem that it is not possible to recognize an appropriate travel path. For example, if a threshold value serving as a reference value for determining whether or not a captured image is an image that can detect a lane marker is set strictly, it is determined that most of the captured images are images in which no lane marker is detected. Thus, the situation that the detection of the traveling path itself cannot be performed frequently occurs. On the other hand, if the threshold value is set to be loose, the number of lane markers that are erroneously detected increases, and highly accurate travel path recognition cannot be performed. For this reason, it is conceivable to set a threshold value in which the balance is adjusted. However, there is a possibility that the threshold value is too loose on a traveling road where a lane marker is easy to detect, such as an expressway. On the other hand, it is difficult to accurately detect a lane marker on a general road or a road on which multiple line lane markers are displayed, and the threshold value may be too strict, making it difficult to recognize an appropriate traveling path.

  Therefore, an object of the present invention is to provide a travel path recognition device that can recognize an appropriate travel path according to the mode of the travel path of the vehicle.

That is, the travel path recognition apparatus according to the present invention is based on an imaging unit that captures an image of the surroundings of a vehicle and a lane center in the vehicle width direction of the vehicle based on lane marker candidate points extracted from a captured image captured by the imaging unit. Parameter estimation means for estimating a parameter relating to the travel path including at least one of an offset amount of the vehicle, a yaw angle of the vehicle with respect to the lane, a lane width, a curve curvature of the lane, and a pitch angle fluctuation amount of the vehicle, and the parameter estimation means The parameter suitability determining means for determining whether or not the estimated parameter is an appropriate parameter, the traveling road aspect detecting means for detecting the aspect of the traveling road on which the vehicle is currently traveling, and the traveling road aspect detecting means. In addition, a threshold value setting for setting an appropriateness determination threshold value used for determining the appropriateness of the parameter according to the mode of the traveling road It is constituted by a stage.

  According to the present invention, the suitability judgment threshold value used for judging the suitability of parameters is set according to the mode of the running road, so that the suitability judgment is made on the running road where the lane marker is easily detected, such as a road of a single-lane lane marker or a highway. The value is set to a strict value, and the suitability determination threshold value can be set to a gentle value on a traveling road where it is difficult to detect a lane marker such as a road of a composite line lane marker or a general road. Thereby, an appropriate threshold value can be set according to the mode of the traveling road, the parameters of the traveling road can be accurately estimated, and the appropriate traveling road can be recognized.

  In the travel path recognition apparatus according to the present invention, the travel path mode detection unit detects a line type of the lane marker on the travel path as the travel path mode, and the threshold setting unit has the composite line as the lane marker. In this case, the suitability determination threshold value is set to be loose (to a value that is easily determined as an appropriate parameter) as compared with the case where the lane marker is a single line.

  In the travel path recognition apparatus according to the present invention, the travel path mode detection unit detects a line type of a lane marker on the travel path as the travel path mode, and the threshold setting unit has a broken line as the lane marker. In this case, the suitability determination threshold value is set to be loose (to a value that is easily determined as an appropriate parameter) as compared with the case where the lane marker is a solid line.

  According to these inventions, when the lane marker is a composite line, the suitability determination threshold is set to be lower than that when the lane marker is a single line, and when the lane marker is a broken line, the suitability determination is performed as compared with the case where the lane marker is a solid line. Set the threshold value loosely. As a result, when the lane marker is a composite line or a broken line, it is easy to be determined as an appropriate parameter, and it is possible to prevent a situation in which the parameter of the travel path cannot be acquired because it is frequently canceled as an inappropriate parameter. In addition, when the lane marker is a single line or a solid line, by setting a strict adequacy judgment threshold value, it is possible to prevent the parameter from being erroneously estimated and accurately estimate the road parameter, and to recognize the appropriate road. .

  Further, in the travel path recognition apparatus according to the present invention, the travel path mode detection unit detects a curvature in the travel path as the travel path mode, and the threshold setting unit has the travel path greater than or equal to a predetermined curvature. In this case, the suitability determination threshold value is set to be loose (to a value that is easily determined as an appropriate parameter) as compared with a case where the travel path is not equal to or more than a predetermined curvature.

  According to the present invention, the suitability determination threshold value is set to be looser when the traveling road has a predetermined curvature or more than when it does not. As a result, when the travel path is greater than or equal to a predetermined curvature, it is easy to be determined as an appropriate parameter, and it is possible to prevent a situation where the travel path parameter cannot be acquired because it is frequently canceled as an inappropriate parameter. In addition, when the road is not more than a predetermined curvature, by setting a strict judgment threshold, it is possible to accurately estimate the road parameter by preventing erroneous parameter estimation and to recognize the appropriate road. Become.

Further, in the travel path recognition apparatus according to the present invention, the travel path mode detection unit detects the type of the travel path as the travel path mode, and the threshold setting unit determines whether the travel path is an expressway or an automobile exclusive road. In this case, the suitability determination threshold value is set stricter (to a value that is less likely to be determined as an appropriate parameter) as compared with a case where the traveling road is not an expressway or a motorway.

  According to the present invention, the suitability determination threshold value is set stricter when the traveling road is an expressway or an automobile-only road than when it is not. As a result, when the road is an expressway or an automobile-only road, it is difficult to be determined as an appropriate parameter, and it is not suitable when the road condition changes such as road branching, lane reduction, or in front of the toll gate. It is possible to accurately cancel the parameter as a parameter and accurately estimate the travel route parameter. Accordingly, it is possible to recognize an appropriate traveling path.

  Further, the travel path recognition apparatus according to the present invention preferably includes a parameter setting unit that sets a parameter estimated last time when the parameter estimated by the parameter suitability determining unit is determined to be not an appropriate parameter.

  According to the present invention, it is possible to recognize an appropriate travel path according to the mode of the travel path of the vehicle.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the description of the drawings, the same elements are denoted by the same reference numerals, and redundant description is omitted.

  FIG. 1 is a schematic configuration diagram of a travel path recognition apparatus according to an embodiment of the present invention.

  As shown in FIG. 1, the travel path recognition apparatus 1 according to the present embodiment is an apparatus mounted on a vehicle 5 and estimates travel path parameters in a travel path 20 on which the vehicle 5 travels.

  The travel path recognition device 1 is provided with an imaging unit 2. The imaging unit 2 functions as an imaging unit that images the surroundings of the vehicle, and is installed, for example, so as to capture the front of the vehicle 5. For example, a C-MOS camera is used as the imaging unit 2.

  The imaging unit 2 is connected to an ECU (Electronic Control Unit) 3. The ECU 3 controls the entire apparatus, and includes, for example, a CPU, a ROM, a RAM, an input signal circuit, an output signal circuit, a power circuit, and the like. The ECU 3 inputs a captured image from the imaging unit 2 and performs image processing. The ECU 3 functions as parameter estimation means for estimating the parameters of the traveling road, parameter suitability judgment means for judging suitability of the parameters, and threshold setting means for setting suitability judgment threshold values used for parameter suitability judgment.

  The travel path recognition device 1 is preferably used as a part of a steering control system. Based on the travel path parameters estimated by the travel path recognition device 1, automatic steering control of the vehicle 5 is performed. For example, when there is a deficiency in the driver's steering operation, steering control is performed so as to compensate for it, thereby ensuring safe vehicle travel.

  Next, the operation of the travel path recognition apparatus according to the present embodiment will be described.

  FIG. 2 is a flowchart showing the operation of the travel path recognition apparatus according to the present embodiment. 3 to 4 are explanatory diagrams of image processing in the control processing of FIG. The control process in FIG. 2 is repeatedly executed at a predetermined cycle preset by the ECU 3.

  First, as shown in S10 of FIG. 2, a captured image is read. The captured image is read by inputting the captured image captured by the imaging unit 2 to the ECU 3 and reading the captured image signal as digital image data. At that time, as shown in FIG. 3, when the vehicle travels along the travel path 20, the captured image 11 is displayed so that the travel path 20 faces in the vertical direction. The running path 20 is provided with white lines 30 and 30 that are lane markers.

  And it transfers to S12 of FIG. 2 and a lane marker detection process is performed. The lane marker detection process is a process of detecting the positions of the white lines 30 and 30 on the travel path 20 by performing image processing on the captured image 11. For example, a predetermined lane marker detection window is set in the photographed image 11, and a candidate portion of the white line 30 is extracted by performing edge processing or the like in the detection window.

  Then, as shown in FIG. 4, candidate lane marker points 31 arranged at predetermined intervals along the white lines 30 and 30 are extracted, and the position of the white line 30 is detected. At this time, five lane marker candidate points 31 are detected for each of the left and right white lines 30, for example. The position of the lane marker candidate point 31 is specified as a coordinate point of the XY axes in the captured image 11.

  Next, the process proceeds to S14 in FIG. 2, and a traveling road mode detection process is performed. This travel path mode detection process is a process for detecting the mode of the travel path 20. For example, the type of lane marker in the travel path 20 is detected. This road mode detection processing is performed by detecting the type of the white line 30 by performing image processing on the captured image 11. In the case of FIG. 4, it is determined that the detected white line 30 is not a broken line but a solid line.

  Then, the process proceeds to S16 in FIG. 2, and a threshold value setting process for determining parameter suitability is performed. This threshold value setting process is a process of setting an appropriateness determination threshold value in the parameter appropriateness determination according to the aspect of the travel path 20 detected by the travel path condition detection process. The suitability determination threshold value is a determination reference value used for determining the suitability of the estimated parameters of the travel path 20.

  In the threshold value setting process, a suitability determination threshold value is set according to the line type of the lane marker on the travel path 20. For example, when the white line 30 that is a lane marker is a composite line, the suitability determination threshold value is set looser than when the white line 30 is a single line. That is, when the white line 30 is a composite line, the suitability determination threshold is set to a value that makes it easier to determine that the white line 30 is a proper parameter than when the white line 30 is a single line.

  More specifically, as a travel path parameter based on the coordinate data of the lane marker candidate point 31, the offset amount (deviation amount) from the center of the lane (travel lane) in the vehicle width direction, the yaw angle of the vehicle with respect to the lane, the lane When five road parameters are estimated: width (lane width), lane curve curvature, and vehicle pitch angle variation, a lower threshold, upper threshold, and maximum variation for each of these parameters Appropriateness judgment thresholds for the quantity threshold are set.

  At that time, when the white line 30 that is a lane marker is a composite line, the lower threshold is set to a smaller value and the upper threshold is set to a larger value than when the white line 30 is a single line. The quantity threshold is set to a large value. As a result, the suitability determination threshold value is relaxed, and the estimated parameter is prevented from being frequently inappropriate.

  On the other hand, when the white line 30 is a single line, the suitability determination threshold is set stricter than when the white line 30 is a composite line. That is, when the white line 30 is a single line, the suitability determination threshold value is set to a value that is less likely to be determined as an appropriate parameter than when the white line 30 is a composite line. More specifically, when the white line 30 is a single line, the lower threshold value is set to a larger value and the upper threshold value is set to a smaller value than when the white line 30 is a composite line, and the maximum fluctuation amount is set. The threshold is set to a small value. Thereby, the suitability determination threshold value becomes stricter, the range in which the estimated travel path parameter is determined to be an appropriate parameter is narrowed, and erroneous recognition of the travel path parameter can be prevented.

  Here, “composite line” means a lane marker composed of a plurality of lines, for example, composed of two solid lines, composed of three solid lines, composed of one solid line and one broken line. And those having a broken line-of-sight guide line on both sides of a single solid line and those having a solid line on both sides of a single broken line. On the other hand, “single line” means a lane marker composed of a single line, for example, one composed of a single solid line, one composed of a single broken line, and the like.

  Further, in this threshold value setting process, when the aspect of the travel path 20 changes, the suitability determination threshold value is changed according to the changed aspect of the travel path 20.

  And it transfers to S18 and the estimation process of a travel path parameter is performed. This travel path parameter estimation process is a process of estimating a travel path parameter based on the coordinate data (lane marker position information) of the lane marker candidate point 31. As the travel path parameter, for example, the lane center in the vehicle width direction of the vehicle The five travel path parameters are estimated: an offset amount (deviation amount) from the vehicle, a yaw angle of the vehicle with respect to the lane, a lane width, a curve curvature of the lane, and a pitch angle fluctuation amount of the vehicle.

  As this parameter estimation method, any method may be used as long as it can estimate the road parameter. For example, an offset amount from the center of the lane in the vehicle width direction using the Kalman filter based on the coordinate data of the lane marker candidate point 31, the yaw angle of the vehicle with respect to the lane, the lane width, the curve curvature of the lane, and the pitch angle variation of the vehicle Is calculated.

  The road parameters are not limited to estimating all five of these, but when estimating a part of them, estimating other parameters in addition to these five or estimating other road parameters It may be the case.

  Then, the process proceeds to S20, and it is determined whether or not the travel path parameter estimated in S18 is appropriate. This determination process is performed using the suitability determination threshold value set in S16. For example, when the estimated road parameter is equal to or higher than the lower threshold, when it is equal to or lower than the upper threshold, and when the estimated fluctuation amount of the road parameter is equal to or lower than the maximum fluctuation threshold, It is determined that the travel path parameter is appropriate.

  On the other hand, when the estimated road parameter is lower than the lower limit threshold, when the upper limit threshold is exceeded, or when the estimated fluctuation amount of the road parameter exceeds the maximum fluctuation threshold, the road parameter Is deemed inappropriate.

  When it is determined that the travel path parameters estimated in S20 are appropriate, travel path parameter determination processing is performed (S22). This travel path parameter determination process is a process of setting the estimated travel path parameter as it is as a travel path parameter.

  On the other hand, when it is determined that the travel path parameter estimated in S20 is inappropriate, a parameter reset process is performed (S24). This parameter reset process is a process in which the estimated travel route parameter is not set as the travel route parameter as it is, but another parameter value is set as the travel route parameter. As another parameter value, for example, the previous travel route parameter is used. Moreover, you may use the other parameter value used as an appropriate value. For example, you may use the average value of the travel path parameter in the predetermined times before the last time.

  As described above, according to the travel path recognition device 1 according to the present embodiment, by setting the suitability determination threshold value used for determining the suitability of parameters according to the mode of the travel path 20, a lane marker such as a single-lane lane marker road can be set. Set the threshold value for rigorous judgment on roads that are easy to detect and easy to estimate parameters accurately, and judge whether or not lane markers are difficult to detect on roads that are difficult to detect lane markers, such as roads with complex lane markers. Set the value to a loose value. Thereby, an appropriate threshold value can be set according to the mode of the travel path, the parameters of the travel path 20 can be accurately estimated, and the appropriate travel path 20 can be recognized.

  In addition, when the lane marker is a composite line, the suitability determination threshold value is set to be lower than that when the lane marker is a single line, so that it is easy to be determined as an appropriate parameter and frequently canceled as an inappropriate parameter. It is possible to prevent a situation in which parameters cannot be acquired. In addition, when the lane marker is a single line, by setting a strict adequacy judgment threshold value, it is possible to prevent erroneous parameter estimation and to accurately estimate the road parameter and to recognize the appropriate road 20.

  Here, FIGS. 5 and 6 show the detection distribution of the lane width of the lane marker when the lane width is estimated as the travel path parameter. FIG. 5 is a distribution graph of lane width values detected for the single line lane marker, and FIG. 6 is a distribution graph of lane width values detected for the composite line lane marker. 7 and 8 show the detection transition of the lane width of the lane marker when the lane width is estimated as the travel path parameter. FIG. 5 is a transition graph of the lane width value detected for the single line lane marker, and FIG. 6 is a transition graph of the lane width value detected for the composite line lane marker.

  As shown in FIG. 5, in the case of a single line lane marker, the lane width values are concentrated on one value. On the other hand, as shown in FIG. 6, in the case of the composite line lane marker, the lane width value is dispersed into two values. This is considered to be due to the result that the lane marker candidate points 31 are respectively set on the outer line and the inner line in the composite line.

  Also, as shown in FIG. 7, in the case of a single track lane marker, the lane width value changes in the vicinity of a certain value. On the other hand, as shown in FIG. 8, in the case of a composite line lane marker, the lane width value moves with time and takes two values. This is considered to be due to the result that the lane marker candidate point 31 was initially set on the outer line in the compound line, but the lane marker candidate point 31 was set on the inner line after that.

  In this way, in the case of a composite line, by setting the appropriateness determination threshold value loosely, it is determined as an appropriate parameter, and it is frequently canceled as an inappropriate parameter to prevent a situation where the parameters of the travel path 20 cannot be acquired. can do.

  In each embodiment described above, the line type of the lane marker is detected as the aspect of the traveling road 20 in the traveling road aspect detection process, and the lane marker is a single line when the lane marker is a composite line in the threshold setting process. Although the case where the suitability determination threshold value is set loosely as compared with the case has been described, the travel path recognition apparatus according to the present invention is not limited to such a case.

  For example, when the lane marker of the traveling road 20 is a solid line or a broken line as the aspect of the traveling road 20 in the traveling road aspect detection process, and the lane marker of the traveling road 20 is a broken line as the threshold setting process The suitability determination threshold value may be set looser (to a value that can be easily determined as an appropriate parameter) than when the lane marker is a solid line. Even in this case, the same effect as the travel path recognition apparatus according to the above-described embodiment can be obtained.

  Moreover, the curve curvature in the traveling road 20 is detected as the aspect of the traveling road 20 in the traveling road aspect detection process, and when the traveling road 20 has a predetermined curvature or more as the threshold setting process, the traveling road has a predetermined curvature or more. Compared with the case where it is not, the threshold value for suitability determination may be set loosely.

  In this case, when the travel path 20 has a predetermined curvature or more, the suitability determination threshold is set to be lower than when the travel path 20 does not. Therefore, when the travel path 20 has a predetermined curvature or more, an appropriate parameter is set. Therefore, it is possible to prevent a situation in which the parameters of the travel path 20 cannot be acquired because the parameters are frequently canceled as inappropriate parameters. In addition, when the travel path 20 is not equal to or more than a predetermined curvature, by setting a strict adequacy judgment threshold, it is possible to accurately estimate the travel path parameters by preventing erroneous parameter estimation, and to recognize the appropriate travel path 20. Can be done. The curve curvature may be detected based on the image processing of the captured image 11 or may be detected based on the map information of the navigation system.

  Further, the type of the traveling road 20 is detected as the aspect of the traveling road 20 in the traveling road aspect detection process, and when the traveling road 20 is an expressway or an automobile-only road as the threshold setting process, the travel path is an expressway or The suitability determination threshold value may be set stricter (to a value that is less likely to be determined as an appropriate parameter) as compared to a case where the road is not a car-only road.

  In this case, when the traveling road 20 is an expressway or an automobile-only road, when the traveling road 20 is an expressway or an automobile-only road, by setting a threshold value for judging appropriateness in comparison with the case where it is not so, However, if the road conditions change, such as road branching, lane reduction, or in front of the toll booth, the parameters will be canceled appropriately as inappropriate parameters and the road parameters will be accurately determined. Can be estimated. Therefore, it is possible to recognize an appropriate travel path 20. The type of the travel path 20 may be detected based on the image processing of the captured image 11 or may be detected based on the map information of the navigation system.

1 is a schematic configuration diagram of a travel path recognition apparatus according to an embodiment of the present invention. It is a flowchart which shows operation | movement of the traveling path recognition apparatus of FIG. It is explanatory drawing of the picked-up image in the control processing of the flowchart of FIG. It is explanatory drawing of the control processing content of the flowchart of FIG. It is explanatory drawing of lane width value distribution detected by the travel path recognition apparatus of FIG. It is explanatory drawing of lane width value distribution detected by the travel path recognition apparatus of FIG. It is explanatory drawing of lane width value transition detected by the travel path recognition apparatus of FIG. It is explanatory drawing of lane width value transition detected by the travel path recognition apparatus of FIG.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Traveling path recognition apparatus, 2 ... Imaging part, 3 ... ECU, 5 ... Vehicle, 11 ... Captured image, 20 ... Traveling path, 30 ... White line, 31 ... Lane marker candidate point.

Claims (6)

Imaging means for imaging the surroundings of the vehicle;
Based on the lane marker candidate points extracted from the captured image captured by the imaging means, the offset amount from the lane center in the vehicle width direction of the vehicle, the yaw angle of the vehicle with respect to the lane, the lane width, the curve curvature of the lane, and the vehicle pitch Parameter estimating means for estimating a parameter relating to the travel path including at least one of the angular fluctuation amounts ;
Parameter suitability judging means for judging whether or not the parameter estimated by the parameter estimating means is an appropriate parameter;
A road mode detection means for detecting the mode of the road on which the vehicle currently travels;
Threshold value setting means for setting a suitability determination threshold value used in the parameter suitability determination according to the travel path aspect detected by the travel path condition detection means;
A travel path recognition device. The travel path mode detection means detects a line type of a lane marker in the travel path as the travel path mode,
The threshold setting means, when the lane marker is a composite line, to set the suitability determination threshold looser than when the lane marker is a single line;
The travel path recognition apparatus according to claim 1. The travel path mode detection means detects a line type of a lane marker in the travel path as the travel path mode,
The threshold setting means, when the lane marker is a broken line, to set the suitability determination threshold looser than when the lane marker is a solid line;
The travel path recognition apparatus according to claim 1. The travel path mode detection means detects a curvature in the travel path as the travel path mode,
The threshold value setting means, when the travel path is equal to or greater than a predetermined curvature, to set the suitability determination threshold looser than when the travel path is not equal to or greater than a predetermined curvature;
The travel path recognition apparatus according to claim 1. The travel path mode detection means detects the type of the travel path as the mode of the travel path,
The threshold setting means, when the road is an expressway or an automobile-only road, strictly setting the suitability determination threshold as compared to a case where the road is not an expressway or an automobile-only road;
The travel path recognition apparatus according to claim 1. 6. The apparatus according to claim 1, further comprising parameter setting means for setting a parameter estimated last time when it is determined that the parameter estimated by the parameter suitability determining means is not an appropriate parameter. The travel path recognition apparatus described.

Images