JP5568029B2 - Vehicle white line recognition device - Google Patents

Description

  The present invention relates to a vehicle white line recognition device that recognizes a white line based on an image captured by an in-vehicle camera.

  In recent years, in order to improve the safety of vehicles, driving support devices that actively support driving operations of drivers have been developed. In general, in order to realize a lane departure prevention function or the like, this driving support device recognizes a white line based on a captured image or the like ahead of the host vehicle, and estimates a driving lane or the like based on the recognized white line. . As this type of white line recognition technology, for example, in Patent Document 1, a search is made for each line extending in the horizontal direction on an image obtained by imaging the front of the host vehicle, and a threshold value set for each of the luminance value and the luminance differential value is set. Is detected as a lane candidate point, and a lane (white line) position is detected based on a plurality of lane candidate points detected by the search.

  By the way, in addition to the white line, there are many snows that cause noise, road patterns, road surface repair marks, water pools, etc., and there are many situations where it is not easy to detect only genuine white line candidate points. To do. Therefore, for example, in Patent Document 1 described above, Hough transform is performed on a candidate point group on a captured image, a straight line (Hough straight line) that fits most candidate points is obtained, and a candidate that deviates from the straight line is obtained. A technique for excluding points as noise is disclosed.

JP 2007-264955 A

  However, the road shape is not only a straight line that can be approximated, but when a curved curve section is included, it may be difficult to appropriately select white line candidate points.

  On the other hand, as a known method, there is a Hough transform method for defining a curve, a circle, or the like. For example, when approximating a white line candidate point in real space by a quadratic function, a curvature, an inclination, an intercept, etc. It is theoretically possible to statistically identify the parameters of the above using Hough transform and select white line candidate points based on the parameters.

  However, the calculation process for identifying all parameters by the Hough transform requires an enormous calculation process and is not realistic from the viewpoint of processing in real time.

  The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a vehicular white line recognition device that can accurately obtain a white line approximate line while reducing the calculation time.

A vehicle white line recognition apparatus according to an aspect of the present invention examines a luminance change in a vehicle width direction for each search line extending in a horizontal direction within a white line detection region set on an image obtained by imaging a host vehicle traveling road, Based on candidate point detection means that detects edge points that change from dark to bright over a predetermined value as white line candidate points, and higher-order approximation formulas that approximate the approximate shape of white lines in real space based on past white line detection results a plurality of high-order side provisional white approximate line for the low-order side coefficient identification means for identifying the low-order side coefficients including zero order coefficient, wherein the non-identified in the low-order side coefficient identification unit for temporary white approximate line comprising Regarding a combination of coefficients, for each of the white line candidate points projected onto the real space , a plurality of combinations of the higher-order coefficients defining a curve passing through the coordinates of the white line candidate points and the coordinates determined by the zero-order coefficient Calculate and calculate the most A higher-order coefficient identifying means for identifying the combination of the higher-order coefficients as a higher-order coefficient of the provisional white line approximation line, and candidate point selection based on the provisional white-line approximation line in which each coefficient is identified A candidate point selection means for setting an area and selecting the white line candidate point existing in the candidate point selection area from the white line candidate points detected in the white line detection area; White line approximate line calculation means for calculating a white line approximate line using the selected white line candidate point.

  According to the white line recognition device for a vehicle of the present invention, the white line approximate line can be obtained with high accuracy while reducing the calculation time.

Schematic configuration diagram of a vehicle driving support device Flow chart showing white line recognition routine Flow chart showing coefficient identification subroutine Explanatory drawing which shows typically an example of the captured image of the environment outside the vehicle Explanatory drawing which shows the white line candidate point detected from the image of FIG. Explanatory drawing which shows an example of the change of the brightness | luminance on a search line Explanatory drawing which shows an example of the white line candidate point projected on real space (A) is explanatory drawing which shows the movement amount of the own vehicle by driving | running | working, (b) is explanatory drawing which shows the relative movement amount with respect to the own vehicle of the white line candidate point detected by the front frame. Explanatory drawing which shows each pattern of the quadratic curve which passes between two points of a reference point and the white line candidate point to pay attention to Explanatory drawing which shows the secondary coefficient calculated based on each primary coefficient (A) is explanatory drawing which shows an example of the coefficient distribution table | surface which normalized and voted the secondary coefficient calculated based on each primary coefficient about 1 white line candidate point, (b) is a some white line candidate point Explanatory drawing which shows an example of the coefficient distribution table which normalized and voted the secondary coefficient computed based on each primary coefficient about Explanatory drawing which shows temporary white line approximate line and candidate point selection area Explanatory drawing which shows the white line approximate line calculated using the selected white line candidate point, and the white line detection area used in the next frame

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. The drawings relate to an embodiment of the present invention, FIG. 1 is a schematic configuration diagram of a vehicle driving support device, FIG. 2 is a flowchart showing a white line recognition routine, FIG. 3 is a flowchart showing a coefficient identification subroutine, and FIG. FIG. 5 is an explanatory diagram schematically illustrating an example of a captured image, FIG. 5 is an explanatory diagram illustrating white line candidate points detected from the image in FIG. 4, and FIG. 6 is an explanatory diagram illustrating an example of a change in luminance on the search line. Is an explanatory diagram showing an example of white line candidate points projected in real space, FIG. 8 (a) is an explanatory diagram showing the amount of movement of the vehicle by traveling, and (b) is a white line candidate point detected in the previous frame. FIG. 9 is an explanatory diagram showing a relative amount of movement with respect to the own vehicle, FIG. 9 is an explanatory diagram showing each pattern of a quadratic curve passing between two points of a reference point and a target white line candidate point, and FIG. 10 shows each primary coefficient FIG. 11 (a) is an explanatory diagram showing secondary coefficients calculated based on FIG. FIG. 4 is an explanatory diagram showing an example of a coefficient distribution table obtained by standardizing and voting secondary coefficients calculated based on each primary coefficient for one white line candidate point, and (b) shows each one for a plurality of white line candidate points. FIG. 12 is an explanatory diagram showing an example of a provisional white line approximation line and a candidate point selection area, and FIG. 13 is a selected white line. It is explanatory drawing which shows the white line approximated line calculated using the candidate point, and the white line detection area | region used by the following frame.

  In FIG. 1, reference numeral 1 denotes a vehicle such as an automobile (own vehicle), and a driving support device 2 is mounted on the vehicle 1. The driving support device 2 includes, for example, a stereo camera 3 and a stereo image recognition device. 4, the control unit 5 and the like are included in the main part.

  The host vehicle 1 is connected to a vehicle speed sensor 11 that detects the host vehicle speed V, a yaw rate sensor 12 that detects the yaw rate γ, a main switch 13 that performs ON / OFF switching of each function of the driving support control, and the like, and a steering wheel. A steering angle sensor 14 that detects the steering angle θst that is opposed to the steering shaft, an accelerator opening sensor 15 that detects an accelerator pedal depression amount (accelerator opening) θacc by the driver, and the like are provided.

  The stereo camera 3 is constituted by a set of CCD cameras using a solid-state imaging device such as a charge coupled device (CCD) as a stereo optical system. These left and right CCD cameras are respectively mounted at a certain distance in front of the ceiling in the vehicle interior, take a stereo image of an object outside the vehicle from different viewpoints, and output image data to the stereo image recognition device 4. In the following description, one image (for example, the right image) of the stereo images is referred to as a reference image, and the other image (for example, the left image) is referred to as a comparative image.

  First, the stereo image recognition device 4 divides the reference image into small areas of 4 × 4 pixels, for example, compares the luminance or color pattern of each small area with the comparison image, finds a corresponding area, and performs the entire reference image. Find the distance distribution over. Further, the stereo image recognition device 4 examines the luminance difference between each pixel on the reference image and an adjacent pixel, extracts those whose luminance difference exceeds a threshold value as an edge point, and extracts the extracted pixel (edge A distance image (a distribution image of edge points having distance information) is generated by giving distance information to (point). Then, the stereo image recognition device 4 performs a well-known grouping process on the generated distance image and compares it with a pre-stored three-dimensional frame (window), so that a white line and a side wall in front of the vehicle Recognize three-dimensional objects.

  Here, the white line to be recognized in the present embodiment is, for example, a single lane line or a multiple line (double line or the like) in which a line of sight guide line is provided inside the lane line, Lines that extend on the road and divide the vehicle lane are collectively referred to, and the form of each line includes a solid line, a broken line, and the like, and further includes a yellow line and the like. In the white line recognition of this embodiment, even if the white line that exists on the road is a double white line or the like, the white line is approximated and recognized by a single straight line or a curved line.

  By the way, the white line laid on the actual road has various variations as described above, and the form of the white line changes with the branching or merging of the traveling road. In addition, there are various noises such as road surface repair marks, water pools and snow on the road. Therefore, it is difficult to accurately recognize the white line in all scenes by uniform processing. Therefore, the stereo image recognition device 4 complementarily performs white line recognition using various methods without relying only on white line recognition using pattern matching based on the distance image as described above, and comprehensively recognizes these recognition results. Judgment to recognize the final white line.

  As one of such white line recognition, the stereo image recognition device 4 recognizes left and right white lines based on a change in luminance in the horizontal direction on one image (for example, a reference image shown in FIG. 4) taken in front of the host vehicle. I do. More specifically, for example, as shown in FIG. 5, the stereo image recognition device 4 sets left and right white line detection areas A (Al, Ar) on the image, and horizontally in each white line detection area A. For a plurality of search lines 1 extending, a change in luminance is examined from the inside to the outside in the vehicle width direction for each search line l. Then, the stereo image recognition device 4 detects, as white line candidate points Pd (Pdl, Pdr), on the respective search lines 1 in the respective white line detection areas A, the first edge points whose luminance changes from dark to bright by a predetermined amount or more. To do.

  Then, the stereo image recognition device 4 projects each white line candidate point Pd on the real space (see FIG. 7), and applies a higher-order approximation formula (function) to each point group of the projected left and right white line candidate points Pd. ) And a candidate point selection area As (Asl, Asr) based on the temporary white line approximate line is set (see FIG. 12).

  Then, the stereo image recognition device 4 selects the white line candidate point Pd existing in the candidate point selection area As from the white line candidate points Pd as the final white line candidate point P (Pl, Pr) (see FIG. 12). ).

  Then, the stereo image recognition device 4 calculates the final left and right white line approximate lines L (Ll, Lr) using the white line candidate points P selected in the left and right candidate point selection regions As and performs the calculation. Based on the provisional white line approximate line L, the white line detection area A in the next frame is set (see FIG. 13).

  Here, when setting the provisional white line approximate line Lt, the stereo image recognition device 4 first identifies low-order coefficients including zero-order coefficients (constants) based on past white line detection results. For example, the provisional white line approximate line Lt in the present embodiment is defined by a quadratic approximate expression. In this case, the stereo image recognition device 4 is based on the white line approximate line recognized in the previous frame. Identify the following coefficients (constants):

  When the low-order side coefficient is identified, the stereo image recognition device 4 identifies a plurality of unordered high-order coefficients. In the identification of the higher-order side coefficient, the stereo image recognition device 4 uses the coordinates of the white line candidate point Pd and the lower-order side coefficient for each white line candidate point Pd projected on the real space. A plurality of combinations of coefficients are calculated. Then, the combination of higher-order coefficients calculated the most is extracted, and the extracted higher-order coefficients are identified as the coefficients of the provisional white line approximate line Lt. For example, in the present embodiment in which the provisional white line approximate line Lt is defined by a quadratic approximation formula, the stereo image recognition device 4 substitutes the coordinates of the white line candidate point Pd for the provisional white line approximation line Lt for which the zeroth order coefficient has been identified. By doing this, a plurality of patterns of combinations of secondary coefficients and primary coefficients are calculated for each white line candidate point Pd. When a plurality of combinations of secondary coefficients and primary coefficients are calculated for all the white line candidate points Pd, the stereo image recognition device 4 selects the most from the combinations of the calculated secondary coefficients and primary coefficients. A large number of calculated combinations are extracted, and the combinations are identified as higher-order coefficients of the provisional white line approximate line Lt.

  As described above, in this embodiment, the stereo image recognition device 4 includes candidate point detection means, low-order side coefficient identification means, high-order side coefficient identification means, candidate point selection means, and white line approximate line calculation means. Realize the function.

  Note that the stereo image recognition device 4 can perform white line recognition by various other methods. Then, for example, the stereo image recognition apparatus 4 has a relationship between an approximate line of a white line recognized in each white line recognition and a white line candidate point such as an edge point used for the recognition (for example, dispersion of white line candidate points with respect to the approximate line, etc.) ) Is selected from the recognized approximate lines, and the approximate line is output to the control unit 5 as the final approximate line of the white line.

  The control unit 5 receives driving environment information in front of the host vehicle 1 recognized by the stereo image recognition device 4. Further, the vehicle speed V from the vehicle speed sensor 11, the yaw rate γ from the yaw rate sensor 12, and the like are input to the control unit 5 as travel information of the host vehicle 1, and operation input information by the driver from the main switch 13. An operation signal, a steering angle θst from the steering angle sensor 14, an accelerator opening θacc from the accelerator opening sensor 15, and the like are input.

  For example, when an execution of an ACC (Adaptive Cruise Control) function which is one of the functions of the driving support control is instructed through the operation of the main switch 13 by the driver, the control unit 5 recognizes the stereo image recognition device 4. The direction of the preceding vehicle is read, and it is identified whether or not the preceding vehicle to be followed is traveling on the own vehicle traveling path.

  As a result, when the preceding vehicle to be tracked is not detected, the constant speed traveling control for maintaining the vehicle speed V of the host vehicle 1 at the set vehicle speed set by the driver through the opening / closing control (engine output control) of the throttle valve 16. Execute.

  On the other hand, when a preceding vehicle that is a tracking target vehicle is detected and the vehicle speed of the preceding vehicle is equal to or lower than the set vehicle speed, the following traveling control is performed in which the following distance is converged to the target inter-vehicle distance. Executed. During this follow-up running control, the control unit 5 basically converges the inter-vehicle distance to the target inter-vehicle distance through the opening / closing control of the throttle valve 16 (engine output control). Furthermore, when it is determined that sufficient deceleration cannot be obtained only by controlling the throttle valve 16 due to sudden deceleration of the preceding vehicle, the control unit 5 controls the output hydraulic pressure from the active booster 17 (automatic braking intervention). Control) to converge the inter-vehicle distance to the target inter-vehicle distance.

  Further, when execution of the lane departure prevention function, which is one of the functions of the driving support control, is instructed through the operation of the main switch 13 by the driver, the control unit 5 will, for example, the left and right white lines that define the own vehicle traveling lane. A warning determination line is set based on (approximate white line recognized by the stereo image recognition device 4), and the own vehicle traveling path is estimated based on the vehicle speed V and the yaw rate γ of the own vehicle 1. Then, for example, when the control unit 5 determines that the host vehicle travel route crosses either the left or right alarm determination line within a set distance (for example, 10 to 16 [m]) ahead of the host vehicle, It is determined that the host vehicle 1 is likely to depart from the current host vehicle lane, and a lane departure warning is issued.

  Next, white line recognition based on the luminance change on the reference image, which is executed in the stereo image recognition device 4, will be described with reference to the white line recognition routine shown in FIG. In this routine, the same processing is performed separately for the left and right white line detection areas Al and Ar. However, to simplify the description, unless otherwise required in the following description. For example, the white line detection areas Al and Ar are collectively referred to as the white line detection area A, and the subscripts “l” and “r” indicating the left and right attributes are omitted as appropriate.

  When this routine starts, the stereo image recognition device 4 first detects a white line candidate point Pd for each search line l in the white line detection area A set in the process of step S108 for the image of the previous frame in step S101. I do. Specifically, for example, as illustrated in FIG. 6, the stereo image recognition device 4 performs edge detection on each search line l from the inner side to the outer side in the vehicle width direction, and detects pixels on the outer side in the vehicle width direction. A point (edge point) in which the luminance is relatively high with respect to the luminance of the inner pixel and the differential value of the luminance indicating the amount of change is equal to or greater than the set threshold on the plus side is detected. Then, the stereo image recognition device 4 extracts the edge point first detected as the white line candidate point Pd on each search line l in the white line detection area A (see FIG. 5). In FIG. 5, the search line l, the white line candidate point Pd, and the like are thinned out and displayed in a predetermined manner in order to simplify the description.

  In subsequent step S102, the stereo image recognition device 4 uses each white line candidate point Pd detected in step S101 as the distance information of the corresponding pixel on the distance image and the pixel position of the horizontal method of the corresponding pixel on the reference image. Based on the above, the image is projected into the real space (see FIG. 7).

When the process proceeds from step S102 to step S103, the stereo image recognition device 4 performs identification processing of the zeroth order coefficient (constant term c) of the provisional white line approximate line Lt made of the quadratic approximate expression. That is, for example, the stereo image recognition device 4 is based on the vehicle speed V of the host vehicle 1 and the yaw angle θ obtained from the yaw rate γ based on the relationship shown in FIG. The movement amounts Δx and Δz of the host vehicle 1 during the period until one frame is updated are calculated using the following equations (1) and (2).
Δx = V × sin θ (1)
Δz = V × cos θ (2)
Then, the stereo image recognition device 4 uses the movement amounts Δx, Δz of the own vehicle, and the white line candidate point detected in the previous frame (the previous value of the final white line candidate point P after selection) Ppr Then, the movement amounts dX, dZ (see FIG. 8B) on the XZ coordinate system based on the own vehicle 1 are calculated using the following equations (3) and (4).
dX = Δx × cos θ−Δz × sin θ (3)
dZ = Δx × sin θ + Δz × cos θ (4)
Then, the stereo image recognition device 4 uses each white line candidate point Ppr ′ of the previous frame moved on the XZ coordinate system using the movement amounts dX and dZ, and uses the current frame position as a reference. The white line approximate line Lpr ′ at is estimated by the least square method. Then, the obtained 0th order coefficient of the white line approximate line Lpr ′ is identified as the 0th order coefficient of the provisional white line approximate line Lt.

  That is, it is unlikely that the position of the vehicle suddenly changes in the vehicle width direction with respect to the white line except when the vehicle 1 changes lanes. Therefore, the low-order coefficient of the provisional white line approximate line Lt is identified based on the white line approximate line Lpr 'in the current frame estimated from the white line candidate point Ppr' of the previous frame based on the amount of movement of the vehicle.

  Here, in order to simplify the identification of the low-order side coefficient, for example, the low-order coefficient of the white line approximate line Lpr calculated in the previous frame can be used as it is as the low-order coefficient of the temporary white line approximate line Lt. .

  Then, when the process proceeds from step S103 to step S104, the stereo image recognition device 4 configures the white line that forms the white line approximate line Lpr ′ in the current frame estimated based on the movement amount of the vehicle from the white line candidate points recognized in the previous frame. It is checked whether or not the number of candidate points Ppr ′ is equal to or greater than a preset threshold value.

  If it is determined in step S104 that the number of white line candidate points Ppr ′ is less than the threshold value, the stereo image recognition device 4 has a high possibility that the accuracy of the low-order coefficient identified in step S103 described above is low. Judge and exit the routine.

  On the other hand, when it is determined in step S104 that the number of white line candidate points Ppr ′ is equal to or greater than the threshold, the stereo image recognition device 4 proceeds to step S105.

  When the process proceeds from step S104 to step S105, the stereo image recognition device 4 performs identification processing of the secondary coefficient a and the primary coefficient b of the provisional white line approximate line Lt made of a quadratic approximate expression. This identification processing is executed, for example, according to the flowchart of the coefficient identification subroutine shown in FIG. In the following description, white line candidate points Pd detected in each white line detection area A are numbered n = 1 to nmax in order from the white line candidate point Pd closest to the vehicle 1. And

  When the subroutine starts, the stereo image recognition equivalence 4 first determines the white line candidate point Pd (n = 1) closest to the vehicle 1 among the white line candidate points Pd detected in the white line detection area A in step S201. The extracted white line candidate point of interest is extracted and the coordinates (X, Z) in the real space of the white line candidate point Pd are substituted into the approximate expression of the temporary white line approximate line Lt in which the constant term c is identified.

  In subsequent step S202, the stereo image recognition device 4 sets the primary coefficient b to the lowest value (−β) of a range (for example, b = −β to β) that is allowable for the coefficient b, and then in step S203. Proceed to

When the process proceeds from step S202 to step S203, the stereo image recognition device 4 uses the following equation (5) obtained by modifying the approximate expression of the temporary white line approximate line Lt using the currently set primary coefficient b. The next coefficient a is calculated (see FIG. 10).
a = (X−bZ−c) / Z ² (5)
In step S204, the primary coefficient a calculated in step S203 is digitized (normalized) into a preset Δα step value. In step S205, the corresponding part of the coefficient distribution table shown in FIG. (See FIG. 11).

  When the process proceeds from step S205 to step S206, the stereo image recognition device 4 checks whether or not the value of the primary coefficient b has reached the maximum value β that the coefficient b can take.

  If it is determined in step S206 that the primary coefficient b has not yet reached the maximum value β (that is, b ≠ β), the stereo image recognition apparatus 4 proceeds to step 207 and determines the primary coefficient b. After the value is updated to a higher value by Δβ, the process returns to step S203.

  Then, by repeatedly executing the processing of step S203 to step S207, for example, as shown in FIG. 9, a plurality of patterns of quadratic curves (coefficients a, a, which pass through the zeroth order coefficient c and the target white line candidate point Pd as shown in FIG. b) is computed.

  On the other hand, when it is determined in step S206 that the primary coefficient b has reached the maximum value β (that is, b = β), the stereo image recognition device 4 proceeds to step S208, and the white line candidate focused on this time It is examined whether or not the point Pd is the white line candidate point Pd located farthest from the own vehicle 1 (that is, whether or not n = nmax).

  If it is determined in step S208 that n ≠ nmax, the stereo image recognition apparatus 4 proceeds to step S209, and updates the white line candidate point Pd of interest to a white line candidate point Pd one distant from the current one ( After n = n + 1), the process returns to step S202.

  On the other hand, if it is determined in step S208 that n = nmax, the stereo image recognition apparatus 4 proceeds to step S210, and a coefficient distribution table (see FIG. 10) voted for all combinations of the secondary coefficient a and the primary coefficient b. 11 (b)), the most frequently voted combination of the secondary coefficient a and the primary coefficient b is extracted, and the coefficients a and b of the combination are coefficients of the approximate expression of the provisional white line approximate line Lt. And then exit the subroutine.

As a result, the approximate expression shown in the following expression (6) is identified as an approximate expression indicating the provisional white line approximate line.
X = aZ ² + bZ + c (6)
In the equation (6), for convenience, the parameters are represented by the left and right common symbols a, b, and c, but these are set individually for the left and right temporary white line approximate lines Ltl and Ltr. Needless to say.

In the main routine of FIG. 2, when the process proceeds from step S105 to step S106, the stereo image recognition device 4 selects a final white line candidate point P based on the temporary approximate curve Lt in which each coefficient is identified. That is, the stereo image recognition device 4 offsets the temporary white line approximate line Lt by ΔE to the inside and the outside in the vehicle width direction, respectively, so that the candidate point selection area As defined by the following expressions (7) and (8) is obtained. Set (see FIG. 12).
X1 = aZ ² + bZ + c−ΔE (7)
X2 = aZ ² + bZ + c + ΔE (8)
Then, the stereo image recognition device 4 uses, as the final white line candidate point P, the white line candidate point Pd existing in the candidate point selection area As among all the white line candidate points Pd detected in the white line detection area A. Select (see FIG. 13). For the sake of convenience, it goes without saying that the candidate points P are also individually set as Pl and Pr respectively on the left and right temporary white lines.

In subsequent step S107, the stereo image recognition device 4 calculates the white line approximate line L using the white line candidate points P selected in the left and right candidate point selection regions As (see FIG. 13). Here, the white line approximate line L of the present embodiment is represented by the following formula (9). The white line vehicle horizontal direction distance X is determined by the parameters A, B, and C with respect to the vertical direction Z distance to the vehicle 1. Is identified.
X = AZ ² + BZ + C (9)
Here, the parameters A, B, and C of the temporary white line approximate line Lt shown in the equation (9) can be obtained by, for example, the least square method using the final white line candidate point P. B and C are determined to be values that minimize the variance of each white line candidate point P with respect to the white line approximate line L. In Equation (9), for convenience, parameters are represented by common symbols A, B, and C on the left and right, but these are set individually for the left and right white line approximate lines Ll and Lr. Needless to say.

In step S108, the stereo image recognition apparatus 4 offsets the white line approximate line L calculated in step S107 by ΔE ′ (which may be the same or different from ΔE) to the inside and outside in the vehicle width direction. Thus, after setting the white line detection area A in the next frame defined by the following equations (10) and (11) (see FIG. 12), the routine is exited.
X1 = AZ ² + BZ + C−ΔE ′ (10)
X2 = AZ ² + BZ + C + ΔE ′ (11)
According to such an embodiment, when identifying each coefficient of the temporary white line approximate line that approximates the approximate shape of the white line in real space, the low-order side coefficient is identified based on the white line approximate line recognized in the previous frame. A plurality of unidentified combinations of higher-order coefficients are calculated for each white-line candidate point Pd projected in real space using the coordinates of the white-line candidate points Pd and lower-order coefficients, By identifying the combination of the higher-order coefficients calculated most frequently as the higher-order coefficient of the provisional white line approximation line Lt, the provisional white line approximation line can be accurately obtained with a high-order approximation expression by simple computation with reduced computation time. Lt can be obtained. Then, the candidate point selection area As is set with reference to the provisional white line approximate line Lt, and the white line candidate point Pd existing in the candidate point selection area As is selected as the final white line candidate point P, and the selected white line is selected. By calculating the white line approximate line L based on the candidate point P, the white line approximate line L can be obtained with high accuracy.

  That is, when setting the provisional white line candidate point Lt, it is unlikely that the position of the own vehicle suddenly changes in the vehicle width direction with respect to the white line, except when the own vehicle 1 performs lane change or the like. By identifying the low-order coefficient based on the white line approximation line Lpr in the current frame estimated from the white line candidate point recognized in the previous frame, while leaving room for the white line curvature fluctuation or the like in the high-order coefficient, The number of coefficients (parameters) identified by fitting with each white line candidate point Pd can be efficiently reduced, and the calculation load can be significantly reduced.

  In this case, by identifying the low-order coefficient based on the white line approximate line Lpr estimated from the white line candidate point of the previous frame and the amount of movement of the vehicle 1 per frame, the vehicle 1 and the white line A low-order coefficient can be identified in consideration of a change in relative position.

  Further, the identification of the coefficients A, B, and C of the white line approximate line L is performed by using the least square method based on the selected point group of the white line candidate points P, so that each coefficient A, B, C can be identified.

  In the above-described embodiment, an example in which the distance information of each pixel is calculated using a stereo camera has been described. However, the present invention is not limited to this, and for example, the vehicle traveling road is a flat road. It is also possible to calculate the distance information of each white line candidate point from an image captured by a monocular camera.

  In the above-described embodiment, an example in which the white line approximate line L (and the provisional white line approximate line Lt) is approximated by a quadratic approximate expression has been described, but the present invention is not limited to this. It is also possible to approximate with the following approximation formulas.

  In this case, for example, when calculating the provisional white line approximate line Lt with a cubic approximation formula, for example, the zeroth order coefficient (constant) and the first order coefficient are used as the low order side coefficients based on the past white line detection results. It is possible to identify and identify the secondary coefficient and the cubic coefficient as higher-order coefficients by fitting with the white line candidate point Pd. Alternatively, only the 0th order coefficient (constant) is identified as the lower order coefficient based on the past white line detection result, and the first order coefficient, the second order coefficient, and the third order coefficient are used as the white line candidate point Pd as the higher order coefficient. It is possible to identify by fitting.

1 ... Vehicle (own vehicle)
2 ... Driving support device 3 ... Stereo camera 4 ... Stereo image recognition device (candidate point detection means, low order side coefficient identification means, high order side coefficient identification means, candidate point selection means, white line approximate line calculation means)
DESCRIPTION OF SYMBOLS 5 ... Control unit 11 ... Vehicle speed sensor 12 ... Yaw rate sensor 13 ... Main switch 14 ... Steering angle sensor 15 ... Accelerator opening sensor 16 ... Throttle valve 17 ... Active booster

Claims (3)

The brightness change in the vehicle width direction is examined for each search line that extends in the horizontal direction within the white line detection area set on the image of the vehicle's travel path, and the edge point where the brightness changes from dark to bright over a predetermined level is indicated by a white line. Candidate point detection means for detecting as candidate points;
Low-order coefficient identification that identifies low-order coefficients including zero-order coefficients for temporary white-line approximate lines that consist of high-order approximations that approximate the approximate shape of white lines in real space based on past white line detection results Means,
Regarding the combination of a plurality of higher-order coefficients that are not identified by the lower-order coefficient identifying means for the provisional white line approximate line, for each of the white line candidate points projected on real space, the coordinate of the white line candidate point and the zero A plurality of higher-order coefficient combinations that define a curve passing through a coordinate determined by the next coefficient are calculated, and the highest-order coefficient combination calculated most frequently is used as the higher-order coefficient of the provisional white line approximation line. Higher-order coefficient identification means for identifying as:
A candidate point selection area based on the provisional white line approximate line in which each coefficient is identified is set, and the white line candidate points detected in the white line detection area are present in the candidate point selection area. Candidate point selection means for selecting white line candidate points;
A white line recognition device for a vehicle, comprising: white line approximate line calculation means for calculating a white line approximate line using the white line candidate point selected in the candidate point selection region. 2. The vehicle according to claim 1, wherein the low-order coefficient identifying unit identifies the low-order coefficient based on a white line approximation line calculated in a previous frame and a movement amount of the vehicle per frame. White line recognition device.   3. The vehicle according to claim 1, wherein the white line approximate line calculating unit calculates the white line approximate line using a least square method based on the selected point group of the white line candidate points. White line recognition device.

Images