DEMARKATION LINE DETECTION DEVICE

Abstract

A demarcation line detection device (10) used for a vehicle in which an imaging device (21) capturing an image (40) of an area in front of the vehicle is mounted, the demarcation line detection device (10) comprising: a demarcation line detection device that detects a driving demarcation line (42) demarcating a driving lane of the vehicle based on the image (40) of the area ahead of the vehicle obtained by the imaging device (21); a demarcation line estimating means that estimates a shape of the driving demarcation line (42) in a region that cannot be recognized by the demarcation line recognizing means based on the driving demarcation line (42) recognized by the demarcation line recognizing means; a reliability level determining means for determining a reliability level of the driving demarcation line (42) detected by the demarcation line detecting means; and an estimation invalidation means for invalidating the estimation of the shape of the driving demarcation line (42) by the demarcation line estimation means based on a determination result of the reliability level of the driving demarcation line (42) by the reliability level determination means.

Description

[Technical field]

[0001] The present invention relates to a demarcation line detection device. More particularly, the present invention relates to a demarcation line detection device used for a vehicle in which an imaging device that captures an image of an area ahead of the vehicle is mounted. [State of the art]

[0002] Various types of controls, such as adaptive cruise control and lane keeping support, are known as vehicle driving support control. In the adaptive cruise control system, a vehicle traveling in the same lane as the host vehicle is selected as a leading vehicle, and the host vehicle travels following the selected leading vehicle. In lane keeping support, a vehicle's travel is controlled so that the vehicle does not deviate from the left and right driving demarcation lines (driving boundary lines).

[0003] In such a driving assistance control, a camera is mounted on the vehicle. An image of an area ahead of the vehicle is captured, and driving demarcation lines are detected. Furthermore, vehicle travel is controlled using the detected driving demarcation lines (see, for example, JP 2001-10524 A). Regarding a device described in JP 2001-10524 A, it is disclosed that a clothoid parameter of driving demarcation lines in an image captured by a camera is calculated. The clothoid parameter indicates a degree of curvature of a course. Future behavior of a vehicle on the course is predicted using the calculated clothoid parameter.

[0004] DE 10 2009 046 699 A1 discloses a driving support ECU comprising: a roadside object detecting section for detecting a roadside object in a case where a lane marking on at least one of the right side and the left side is detected by a white line detecting section and another lane marking is not detected on the other of the right side and the left side; a white line estimating section for estimating a position of a lane marking on the other of the right side and the left side based on the detected roadside object; and an information outputting section for determining a deviation from the lane in which the vehicle is traveling based on the estimated position of the lane marking on the other of the right side and the left side.

[0005] DE 11 2012 003 040 T5 discloses a lane recognition device for mounting on a vehicle, comprising: a camera that captures an image of an area including a road ahead of a vehicle, luminance detection means that detects luminances within the images, lane recognition means that recognizes a lane marking of the road in the images based on the luminances detected by the luminance detection means, a storage section in which a position of the lane marking detected by the lane recognition means is stored, and virtual lane marking setting means that sets a virtual lane marking based on an archived record of the position of a lane marking stored in the storage section in a case where the lane recognition means cannot recognize the lane marking.

[0006] DE 10 2009 045 682 A1 discloses a lane departure warning system and a lane change warning method which use a virtual lane dividing line generated using lane width information calculated using previously detected lane dividing lines and provide the warning to a driver based on the virtual lane dividing line.

[0007] DE 10 2009 028 774 A1 discloses a method for detecting a lane change of a vehicle and for providing a curvature of a target line for lane guidance of the vehicle, wherein the method comprises a step of receiving a first lane information item representing an optically detected first lane boundary next to or in front of a left-hand side of the vehicle and / or receiving a second lane information item representing an optically detected second lane boundary next to or in front of a right-hand side of the vehicle.Furthermore, the method comprises a step of detecting a lane change if the first and / or second lane information represents a lane boundary that is less than a predetermined lateral distance from a side of the vehicle, or determining a curvature of a desired target line for lane guidance of the vehicle from the first lane information weighted with a first weighting factor and the second lane information weighted with a second weighting factor, based on a correlation analysis. [Summary of the invention][Technical problem]

[0008] As a result of estimating a shape of a driving demarcation line that is farther than a detection portion of the driving demarcation line based on a shape of the driving demarcation line recognized from an image captured by an imaging device, it is possible to perform vehicle travel control using information regarding the driving demarcation line at a long distance that cannot be recognized from the image. When the recognition accuracy of the driving demarcation line recognized from the image is low, a deviation between a shape of a white line at a long distance determined by estimation and an actual shape of the white line becomes large.As a result of careful investigations by the inventors, it was found that in such cases, the controllability of the vehicle travel control may deteriorate based on the shape of the travel demarcation line.

[0009] The present invention was made in view of the above problem. It is an object of the present invention to provide a demarcation line detection device and a demarcation line detection method capable of providing, with high estimation accuracy, a result regarding a driving demarcation line outside a detection range of an image to be used for vehicle travel control. This object is achieved by a demarcation line detection device having the features of claim 1 and by a demarcation line detection method having the features of claim 5. The dependent claims are directed to advantageous developments of the invention.

[0010] A demarcation line detection device according to one aspect of the present invention is a demarcation line detection device used for a vehicle in which an imaging device that captures an image of an area ahead of the vehicle is mounted. The demarcation line detection device includes: a demarcation line detection device that detects a driving demarcation line demarcating a driving lane of the vehicle based on an image of the area ahead of the vehicle acquired by the imaging device.demarcation line estimation means that estimates, based on the driving demarcation line detected by the demarcation line detection means, a shape of the driving demarcation line in a range that cannot be detected by the demarcation line detection means; reliability level determination means that determines a reliability level of the driving demarcation line detected by the demarcation line detection means; and estimation invalidation means that invalidates the estimation of the shape of the driving demarcation line by the demarcation line estimation means based on a determination result of the reliability level of the driving demarcation line by the reliability level determination means.

[0011] The above configuration is such that the estimation of the shape of the driving demarcation line in an area that cannot be recognized from an image is invalidated based on the reliability level of the driving demarcation line recognized from the image. When the shape of the driving demarcation line in an area that cannot be recognized from an image is estimated using the driving demarcation line recognized from the image, it can be assumed that the estimation accuracy decreases when the reliability level of the recognized driving demarcation line is low, and that the controllability of the driving support control is affected. As a result of the above configuration adopted in view of this problem, execution of vehicle travel control using a result of the driving demarcation line having a low estimation accuracy can be prevented.The control capability of the vehicle's driving assistance control can thus be developed favorably. [Brief description of the drawings] Fig. 1 is a block diagram of an overall configuration of a system including a demarcation line detection apparatus according to a first embodiment; Fig. 2 is a flowchart of processing steps in a white line detection process of the demarcation line detection device; Fig. 3 shows diagrams of images in which an inter-vehicle distance to a leading vehicle is different at (a) and (b); Fig. 4 is a flowchart of processing steps in an invalidation process of the white line estimation unit; and Fig. 5 is a flowchart of processing steps in a reliability level determination process performed by the white line estimation unit. [Description of the Embodiments](First Embodiment)

[0012] Hereinafter, a demarcation line detection device according to the present embodiment will be described with reference to the drawings. The demarcation line detection device according to the present embodiment is mounted on a vehicle. The demarcation line detection device detects a white line serving as a driving demarcation line that demarcates a driving lane of the vehicle. Information (e.g., white line information) regarding the white line detected by the demarcation line detection device is used for driving support control such as adaptive cruise control and lane keeping support. In adaptive cruise control, for example, a subject vehicle travels to follow a leading vehicle, which is a vehicle traveling in the same lane as the subject vehicle, among vehicles traveling in front of the subject vehicle.In lane keeping assistance, a travel of a vehicle is controlled so that the vehicle does not cross a travel demarcation line. First, an overall configuration of the demarcation line detection device according to the present embodiment will be described with reference to FIG. Fig. 1 described.

[0013] The system that Fig. 1 is mounted on a vehicle and includes a demarcation line detection device 10 according to the present embodiment. Fig. 1, the demarcation line detection device 10 is a computer including a central processing unit (CPU), a read-only memory (ROM), a random access memory (RAM), an input / output (I / O), and the like. Each function (e.g., demarcation line detection means, demarcation line estimation means, reliability level determination means, and estimation validity means) of the demarcation line detection device 10 is realized by executing programs installed in the ROM on the CPU. An imaging device 21 is mounted on the vehicle (i.e., a self-vehicle). The imaging device 21 serves as an object detection means that detects an object present in a vehicle environment.The demarcation line detection device 10 receives an image captured by the imaging device 21 and generates the white line information using the input image.

[0014] The imaging device 21 is an on-board camera. The imaging device 21 is configured by a charge-coupled device (CCD) camera, a complementary metal-oxide-semiconductor (CMOS) imaging sensor, a near-infrared camera, or the like. The imaging device 21 captures an image of surroundings including a traveling road of the own vehicle, generates image data expressing the captured image, and sequentially outputs the image data to the demarcation line detection device 10. The imaging device 21 is disposed, for example, near an upper end of a front windshield of the own vehicle. The imaging device 21 captures an image of a region spread over a range of a predetermined imaging angle δ toward a region in front of the vehicle with an imaging axis as the center.The imaging device 21 may be a single lens camera or a stereo camera.

[0015] The demarcation line detection device 10 receives the image data from the imaging device 21 and also receives respective detection signals from various sensors arranged in the vehicle. In the system shown in Fig. As shown in FIG. 1, a yaw rate sensor 22, a vehicle speed sensor 23, a steering angle sensor 24, and the like are arranged as various sensors. The yaw rate sensor 22 detects an angular velocity (e.g., yaw rate) in a turning direction of the vehicle. The vehicle speed sensor 23 detects a vehicle speed. The steering angle sensor 24 detects a steering angle. The vehicle speed sensor 23 corresponds to vehicle speed detecting means. The yaw rate sensor 22 and the steering angle sensor 24 correspond to turning detecting means.

[0016] The demarcation line detection device 10 includes a white line detection unit 11 and a white line estimation unit 12. The white line detection unit 11 detects a white line located within the image captured by the imaging device 21. The white line estimation unit 12 estimates a shape of a white line in a range that cannot be detected by the white line detection unit 11, that is, a shape of a white line farther than a white line detection range of the white line detection unit 11, using information regarding the white line detected by the white line detection unit 11.

[0017] Fig. 2 is a flowchart of processing steps in a white line detection process performed by the demarcation line detection device 10. The process is repeatedly performed by the CPU of the demarcation line detection device 10 in a predetermined control cycle. As a result of this process, respective functions of the white line detection unit 11 corresponding to the demarcation line detection means (e.g., demarcation line detection unit), the white line estimation unit 12 corresponding to the demarcation line estimation means (e.g., demarcation estimation unit), the reliability level determination means (e.g., reliability level determination unit), and the estimation validity means (e.g., estimation validity unit) are implemented.

[0018] In Fig. 2, in step S10, an image captured by the imaging device 21 is acquired. In the subsequent step S11, edge points P are extracted based on luminance information of a road image in the acquired image. In step S12, Hough transform is performed on the extracted edge points P. Here, a straight line or a curved line in which a plurality of edge points P are continuously arranged is extracted. In the subsequent step S13, feature sizes of white line candidates are calculated, with the extracted straight line or curved line serving as a white line candidate. In step S14, restriction is performed on a pair of straight lines or curved lines extending in a moving direction of the vehicle from the white line candidates using the feature sizes.

[0019] Subsequently, in step S15, bird's-eye view conversion of the edge points P is performed. Specifically, coordinate transformation is performed on the edge points P of the white line candidate to which the white line candidates have been narrowed down using a mounting position and a mounting angle of the imaging device 21, and conversion into a surface view is performed. A region where a white line is located in the acquired surface view is a white line detection region. That is, in the image captured by the imaging device 21, a white line shape can be detected from the host vehicle up to a short distance D1. A position of the detected white line that is the farthest from the host vehicle is an end portion of the white line detection region.The area image is an orthogonal coordinate system centered on the own vehicle, where an X-axis represents a vehicle width direction of the own vehicle and a Y-axis represents a direction of movement of the vehicle.

[0020] In the subsequent step S16, a white line parameter η (e.g., white line position, white line inclination, white line width, white line curvature, or curvature change rate) is estimated. The white line parameter is a parameter that specifies the shape of the white line converted into the surface image. The estimation of the white line parameter η is performed by approximating the shape of the white line converted into the surface image with a polynomial (e.g., white line model).

[0021] Subsequently, in step S17, the shape of the white line outside the white line detection range, that is, the white line farther than the near distance D1 (hereinafter also referred to as "far white line"), is estimated by extrapolation based on the white line parameter η. The estimation of the shape of the far white line is performed based on a white line model using the white line parameter η. This is, for example, at least one of the curvature of the white line and the curvature change rate (e.g., clothoid parameter). The white line model may be approximated by a polynomial, or may be a table or the like. The white line parameter of the estimated white line is saved, and the current routine is terminated.According to the present embodiment, as a result of the processes in steps S10 to S15, the demarcation line detection device 10 realizes the function of the white line detection unit 11 corresponding to the demarcation line detection means. Furthermore, as a result of the processes in steps S16 and S17, the demarcation line detection device 10 realizes the function of the white line estimation unit 12 corresponding to the demarcation line estimation means.

[0022] According to Fig. 1, information regarding the white line detected by the white line detection unit 11 and information regarding the distant white line estimated by the white line estimation unit 12 are input to a vehicle control device 30. The vehicle control device 30 implements driving support control such as an adaptive cruise control function and a lane keeping support function.

[0023] Specifically, in the adaptive cruise control function, the vehicle speed of the own vehicle is controlled to a preset vehicle speed. Furthermore, an inter-vehicle distance between the own vehicle and a leading vehicle is controlled to a distance based on the vehicle speed of the own vehicle. Specifically, a movement trajectory of a vehicle ahead of the own vehicle is compared with the shape of the white line detected by the white line detection unit 11 and the shape of the distant white line estimated by the white line estimation unit 12. When the movement trajectory of the preceding vehicle matches the shape of the white line and the shape of the distant white line, the movement trajectory of the preceding vehicle is set as a future predicted course of the own vehicle.In addition, the leading vehicle to be followed by the own vehicle is selected based on the predicted course, and engine control and braking control are performed to follow the selected leading vehicle.

[0024] The method for predicting the future course of the own vehicle in adaptive cruise control is not limited to that described above. For example, a method may be used in which the shape of the white line detected by the white line detection unit 11 and the shape of the distant white line estimated by the white line estimation unit 12 are set as the future predicted course of the own vehicle. The information regarding the white line detected by the white line detection unit 11 and the information regarding the distant white line estimated by the white line estimation unit 12 correspond to information (e.g., white line information) regarding a white line detected by the demarcation line detection device.

[0025] Furthermore, for example, in the lane keeping assist function, a future position of the own vehicle is predicted based on the vehicle speed and yaw rate. Whether the own vehicle is deviating or will deviate from a white line is determined using the predicted future position, the shape of the white line, and the shape of the distant white line. When it is determined that the own vehicle is deviating or may deviate from the white line, a warning may be displayed on an on-board display or a message may be output via a warning sound. Furthermore, in a system in which the vehicle control device 30 provides a driving assist function, a steering force is applied to a steering wheel when it is determined that the own vehicle is deviating or may deviate from the white line.

[0026] Here, when a white line shape farther than the short distance D1 is estimated using the white line shape up to the short distance D1 detected by the imaging device 21, the accuracy of the white line shape detected by the imaging device 21 is likely to be low, so that the estimation accuracy of the white line shape farther than the short distance D1 decreases. In particular, when the shape of the far white line is estimated based on the white line model, a calculation error of the white line shape detected from the image is amplified by the error in the white line model, and the estimation accuracy is likely to decrease.In such cases, it can be assumed that the discrepancy between the estimated shape of the white line and the actual shape of the white line increases and the control performance of the driving support control deteriorates.

[0027] In view of this problem, according to the present embodiment, a white line reliability level is determined. The white line reliability level is a reliability level (certainty) of the white line detected by the white line detection unit 11. Based on the determination result, the estimation of the white line shape by the white line estimation unit 12 is invalidated. According to the present embodiment, the white line reliability level is determined to be low when a predetermined reliability level determination condition given in advance is satisfied, and the estimation of the white line shape by the white line estimation unit 12 is then invalidated. The reliability level determination condition includes three conditions, that is, a first condition, a second condition, and a third condition, which will be described below.According to the present embodiment, the estimation of the shape of the white line by the white line estimation unit 12 is invalidated when at least one of the three conditions is satisfied.

[0028] First condition: The vehicle speed of the own vehicle is equal to or less than a predetermined low vehicle speed determination value Vth.

[0029] Second condition: The yaw rate of the own vehicle is greater than a predetermined value θth.

[0030] Third condition: A variation or change in width between white lines is present in front of your own vehicle.

[0031] Regarding the first condition, it is noted that in a vehicle providing the adaptive cruise control function, the inter-vehicle distance is controlled such that the inter-vehicle distance between the host vehicle and the leading vehicle is a distance based on the vehicle speed of the host vehicle. Specifically, the inter-vehicle distance is controlled such that the inter-vehicle distance between the host vehicle and the leading vehicle increases as the vehicle speed of the host vehicle increases. Therefore, in a situation where the host vehicle is traveling in a low vehicle speed range, for example, when the host vehicle is traveling in an urban area or a congested section of an expressway, the inter-vehicle distance to the leading vehicle can be made short.Furthermore, the visibility of the white line from the host vehicle depends on the inter-vehicle distance to the leading vehicle. The distance of the white line visible in or on the image decreases as the inter-vehicle distance to the leading vehicle decreases.

[0032] Fig. 3 shows two images 40 in which the inter-vehicle distance to a leading vehicle 43 differs. Fig. Figure 3 shows at (a) a case where the inter-vehicle distance is small, and at (b) a case where the inter-vehicle distance is large. As shown in Fig. As shown in Fig. 3, when the leading vehicle 43 traveling in a self-driving lane 41 is present in front of the own vehicle, a white line detection distance becomes short as a result of the white line being hidden by the leading vehicle 43. Furthermore, the white line detection distance at this time becomes shorter as the inter-vehicle distance to the leading vehicle 43 becomes shorter, compared to when the inter-vehicle distance becomes longer. Therefore, according to the present invention, the condition that the own vehicle is in a traveling state where the vehicle speed of the own vehicle is equal to or lower than the low-vehicle speed determination value Vth is included as a reliability level determination condition.When this first condition is satisfied, the white line reliability level of the white line detected by the white line detection unit 11 is considered to be low, and the estimation of the white line shape by the white line estimation unit 12 is invalidated.

[0033] In the configuration where it is determined based on the vehicle speed whether a situation exists where the inter-vehicle distance to the leading vehicle 43 is small, in a situation where the inter-vehicle distance to the leading vehicle 43 is likely to become short, the estimation of the white line shape by the white line estimation unit 12 is invalidated. Therefore, when a vehicle traveling in an adjacent lane changes its lane from the adjacent lane to the own lane 41, the estimation of the white line shape can be invalidated in advance in a situation where the white line reliability level decreases.

[0034] Regarding the second condition, it is noted that the host vehicle may change lanes when the yaw rate of the host vehicle is large. When the host vehicle changes lanes, the inclination of the white line becomes gentler from the viewpoint of the host vehicle, that is, a curve radius R of the white line decreases, and the edge points P of the white line become difficult to detect. In such cases, the detection accuracy of the white line by the white line detection unit 11 decreases. Therefore, according to the present embodiment, a condition that the yaw rate of the host vehicle is greater than the predetermined value θth is included as a reliability level determination condition.When this second condition is met, the white line reliability level of the white line detected by the white line detection unit 11 is considered low, and the estimation of the white line shape by the white line estimation unit 12 is invalidated. The second condition is a determination condition for determining whether the subject vehicle is in a predetermined turning state in which the vehicle is turning at an angular velocity greater than a predetermined angular velocity with respect to the white line 42.

[0035] Regarding the third condition, it is noted that when there is a change in the width between white lines in front of the own vehicle, it is assumed that the own vehicle is traveling in a section where the road width has narrowed, in a lane merging section, or the like. In this case, there is a possibility that the road shape is deformed or has changed its shape, or the road shape and the white line shape do not match. In such cases, if the shape of the white line farther than the short distance D1 is estimated using the shape of the white line up to the short distance D1 detected by the white line detection unit 11, the deviation between the estimated white line shape and the actual white line shape may be large.Therefore, according to the present invention, the condition that the width change between white lines exists in front of the subject vehicle is included as a reliability level determination condition. When this third condition is met, the white line reliability level of the white line detected by the white line detection unit 11 is considered low, and the estimation of the white line shape by the white line estimation unit 12 is invalidated.

[0036] In the following, a specific aspect according to the present embodiment will be explained with reference to the flowcharts of Fig. 4 and Fig. 5. The processes are performed by the white line estimation unit 12 in a predetermined cycle.

[0037] First, the invalidation process of the Fig. 4 described. In Fig. 4, it is determined in step S21 whether the white line reliability level is low. Here, a reliability level determination flag FA generated by the reliability level determination process of the Fig. 5 is set. The determination is made based on the acquired flag FA. According to the present embodiment, the reliability level determination flag FA is set to 0 when the white line reliability level is a low level, and is set to 1 when the white line reliability level is a high level.

[0038] If it is determined that the white line reliability level is high, the process proceeds to step S22. The estimation of the white line shape by the white line estimation unit 12 is validated. In this case, vehicle travel control is performed using the white line information detected by the white line detection unit 11 and the distant white line information estimated by the white line estimation unit 12.

[0039] If it is determined that the white line reliability level is low, the process proceeds to step S23. The estimation of the white line shape by the white line estimation unit 12 is invalidated. Here, "the estimation of the white line shape by the white line estimation unit is invalidated" includes prohibiting execution of a calculation process for estimating the shape of the removed white line by the white line estimation unit 12, deleting the result estimated by the white line estimation unit 12, not using the result estimated by the white line estimation unit 12 for driving support control, and the like. According to the present embodiment, one of these three processes is performed.Even if the estimation of the white line shape by the white line estimation unit 12 is invalidated, the use of the information concerning the white line detected by the white line detection unit 11 is permitted. According to the present embodiment, the invalidation process of steps S21 to S23 by the white line estimation unit 12 corresponds to the estimation validity means.

[0040] The following describes the reliability level determination process of the Fig. 5. In Fig.5, in step S31, it is determined whether the vehicle speed of the host vehicle is equal to or less than the predetermined low-vehicle speed determination value Vth. Here, the determination is made using the vehicle speed detected by the vehicle speed sensor 23. In step S32, it is determined whether the yaw rate of the host vehicle detected by the yaw rate sensor 22 is greater than the predetermined value θth.

[0041] In step S33, it is determined whether there is a change in the width between white lines in front of the own vehicle. Specifically, the determination is performed using the white line detection result by the white line detection unit 11. If the distance between a pair of white lines in the width direction of the vehicle in front of the own vehicle changes and the magnitude of the change is equal to or greater than a predetermined value, it is determined that there is a change in the width between white lines in front of the own vehicle.

[0042] When the result of the determination in all the processes of steps S31, S32, and S33 is negative, that is, when it is determined that the first to third conditions are all satisfied, the process proceeds to step S34. The reliability level determination flag FA is set to 1. When the result of the determination in any one of steps S31, S32, and S33 is positive, the process proceeds to step S35. The reliability level determination flag FA is set to 0. According to the present embodiment, the reliability level determination process of steps S31 to S35 by the white line estimation unit 12 corresponds to the reliability level determination means.

[0043] According to the present embodiment described above, the following advantageous effects are achieved.

[0044] The configuration is such that the estimation of a white line shape farther than the white line detection range is invalidated based on the reliability level of the white line 42 detected based on the image 40. When the white line shape in a range that cannot be detected in the image 40 is estimated using the white line 42 detected from the image 40, the estimation accuracy of the white line shape decreases if the reliability level of the detected white line 42 is low. As a result of the above-described configuration, in view of the above problem, execution of vehicle travel control using white line information having low estimation accuracy can be prevented. This can favorably affect the controllability of a drive support control of the vehicle.

[0045] The configuration is such that the condition that the vehicle speed of the host vehicle is equal to or less than the predetermined low-vehicle speed determination value Vth (first condition) is included as a reliability level determination condition. When the vehicle speed is equal to or less than the low-vehicle speed determination value Vth, the estimation of the shape of the white line by the white line estimation unit 12 is invalidated. As a result of such a configuration, when the estimation accuracy of the white line by the white line estimation unit 12 is low because the white line detection distance in the image is short, the execution of vehicle travel control using the estimation result of this white line can be prohibited.

[0046] The configuration is such that the condition that the yaw rate of the host vehicle is greater than the predetermined value θth (i.e., the second condition) is included as a reliability level determination condition. When the yaw rate of the host vehicle is greater than the predetermined value θth, the estimation of the shape of the white line by the white line estimation unit 12 is invalidated. As a result of this configuration, when the estimation accuracy of the white line by the white line estimation unit 12 is low due to the lower detection accuracy of the edge points P of the white line, the execution of the vehicle travel control using the estimation result of this white line can be prohibited.

[0047] The configuration is such that a condition that a change in the width between white lines exists in front of the own vehicle (i.e., the third condition) is included as a reliability level determination condition. When a change in the width between white lines exists in front of the own vehicle, the estimation of the white line shape by the white line estimation unit 12 is invalidated. As a result of such a configuration, when the estimation accuracy of the white line by the white line estimation unit 12 is low due to the road shape not being a fixed shape or the road shape and the white line shape not matching, the execution of the vehicle travel control using the estimation result of this white line may be prohibited. (Other embodiments)

[0048] The present invention is not limited to the above-described embodiment and can be embodied by various modifications. For example, the present invention can be embodied in the following manner.

[0049] According to the embodiment described above, the configuration is such that the first to third conditions are included as reliability level determination conditions. However, the configuration may also be such that one or two of the first to third conditions are included as reliability level determination conditions. The estimation of the white line shape by the white line estimation unit 12 is invalidated when at least one condition is satisfied.

[0050] According to the embodiment described above, the yaw rate sensor 22 is used as turning detection means that detects the turning state of the own vehicle. The estimation of the white line shape by the white line estimation unit 12 is invalidated when the yaw rate of the own vehicle is greater than the predetermined value θth. However, the turning detection means is not limited to the above. For example, the steering angle sensor 24 can be used as turning detection means. When the steering angle of the own vehicle is greater than a predetermined value, the own vehicle can be assumed to be in a predetermined turning state, and the estimation of the white line shape by the white line estimation unit 12 can be invalidated. Alternatively, the imaging device 21 can be used as turning detection means.It can be determined based on image data that the vehicle is in a predetermined turning state.

[0051] According to the above-described embodiment, based on the detection result of the inter-vehicle distance to the leading vehicle 43 of the object detection device, it can be determined whether the vehicle speed of the host vehicle is equal to or less than the predetermined low vehicle speed determination value Vth (i.e., the first condition). In this case, the configuration is such that it is determined that the first condition is satisfied when the detected inter-vehicle distance is less than a determination value, and the estimation of the white line shape by the white line estimation unit 12 is invalidated.

[0052] Conditions other than the first to third conditions described above may be included as reliability level determination conditions. For example, it may be assumed that the detection accuracy of the white line decreases at night or in rainy or snowy conditions. Therefore, a reliability level determination condition based on the environment, especially the environment at night, rainy or snowy conditions, may be included.

[0053] According to the embodiment described above, the configuration includes the imaging device as the object detecting means. However, the present invention can be applied to a system that includes a radar device or a sonar device in addition to the imaging device.

[0054] According to the above-described embodiment, in the demarcation line detection device 10, a program is stored in the ROM corresponding to a non-volatile recording medium. Each function of the demarcation line detection device 10 is realized by the CPU corresponding to a processor of a computer on which the program runs. However, the configuration may be such that a program is stored in a non-volatile recording medium (for example, a non-volatile memory other than the ROM), and a processor such as the CPU executes the program.In this case, the configuration may be such that, in the demarcation line detection device 10, as a result of the processor executing the program stored in the non-volatile recording medium, a process corresponding to the program (for example, a demarcation line detection process) is performed.

[0055] In addition, some or all of the functions performed by the demarcation line detection device 10 may be implemented using hardware, for example, a single or multiple integrated circuits (i.e., ICs). Furthermore, each device (e.g., demarcation line detection device corresponding to the white line detection unit 11, demarcation line estimation device corresponding to the white line estimation unit 12, reliability level determination device, and estimation validity device) provided by the demarcation line detection device 10 may be implemented using software recorded in a non-volatile recording medium such as a non-volatile memory and a computer running the software, using only hardware, or using a combination of these.

Claims

[1] A demarcation line detection device (10) used for a vehicle in which an imaging device (21) capturing an image (40) of an area in front of the vehicle is mounted, the demarcation line detection device (10) comprising: a demarcation line detection device that detects a driving demarcation line (42) demarcating a driving lane of the vehicle based on the image (40) of the area ahead of the vehicle obtained by the imaging device (21); a demarcation line estimating means that estimates a shape of the driving demarcation line (42) in a region that cannot be recognized by the demarcation line recognizing means based on the driving demarcation line (42) recognized by the demarcation line recognizing means; a reliability level determining means for determining a reliability level of the driving demarcation line (42) detected by the demarcation line detecting means; and an estimation invalidation means for invalidating the estimation of the shape of the driving demarcation line (42) by the demarcation line estimation means based on a determination result of the reliability level of the driving demarcation line (42) by the reliability level determination means. [2] Demarcation line detection device (10) according to claim 1, further comprising: a vehicle speed detecting device (23) which detects a speed of the vehicle, wherein the reliability level determining means determines as the reliability level of the driving demarcation line (42) whether the speed detected by the vehicle speed detecting means (23) is equal to or less than a predetermined low vehicle speed determination value, and the estimation validity means invalidates the estimation of the driving demarcation line (42) by the demarcation line estimation means when the reliability level determination means determines that the speed of the vehicle is equal to or less than the predetermined low vehicle speed determination value. [3] Demarcation line detection device (10) according to claim 1 or 2, further comprising: a rotation detecting device (22, 24) which detects a rotation state of the vehicle, wherein the reliability level determining means determines, as the reliability level of the driving demarcation line (42), whether the vehicle is in a predetermined turning state in which the vehicle rotates at an angular velocity greater than a predetermined angular velocity with respect to the driving demarcation line (42) by the turning detecting means (22, 24), and the estimation validity means invalidates the estimation of the driving demarcation line (42) by the demarcation line estimation means when the reliability level determination means determines that the vehicle is in the predetermined turning state. [4] Demarcation line detection device (10) according to one of claims 1 to 3, wherein the reliability level determining means determines as the reliability level of the driving demarcation line (42) whether there is a change in width between driving demarcation lines (42) in front of the vehicle; and the estimation validity means invalidates the estimation of the driving demarcation line (42) by the demarcation line estimation means when the reliability level determination means determines that there is a change in width between driving demarcation lines (42). [5] Demarcation line detection method comprising: by means of a demarcation line detection device (10) mounted on a vehicle, detecting a driving demarcation line (42) demarcating a driving lane of the vehicle based on an image (40) of an area ahead of the vehicle obtained from an imaging device (21) mounted on the vehicle; by means of the demarcation line recognition device (10), estimating a shape of the driving demarcation line (42) in an area that cannot be recognized in the image (40) on the basis of the driving demarcation line (42) recognized in the image (40); by means of the demarcation line detection device (10) determining a reliability level of the driving demarcation line (42) detected in the image (40); and invalidating, by means of the demarcation line detection device (10), the estimation of the shape of the driving demarcation line (42) based on a determination result of the reliability level of the driving demarcation line (42).

Images