JP4913101B2 - Road shape detection device for vehicles - Google Patents

Description

  The present invention relates to a vehicle road shape detection device that detects the shape of a traveling lane based on an image portion of a road segment line extracted from a captured image in front of the vehicle.

  Conventionally, an image portion of a road lane marking of a traveling lane is extracted from an image in front of the host vehicle captured by an in-vehicle camera, and an image portion of the road lane marking extracted in the right half image area (hereinafter, the right road lane marking A method of determining the gradient of the driving lane based on the degree of curve of the road lane) and the degree of curve of the road segment line extracted in the left half image area (hereinafter referred to as the image part of the left road segment line). Has been proposed (see, for example, Patent Document 1).

  In the above method, when the image portion of the right road lane marking and the image portion of the left road lane marking are both bent inward (center side of the image), it is determined that the traveling lane has a downward slope. Further, when the image portion of the right road dividing line is bent to the right side and the image portion of the left road dividing line is bent to the left side, it is determined that the traveling lane has an upward slope.

In this way, when determining the road shape based on the degree of bending of the image portion of the right road segment line and the image portion of the left road segment line, the image portion of the right road segment line and the image of the left road segment line It is assumed that the parts are extracted together. Therefore, when either the image part of the right road parting line or the image part of the left road parting line is not extracted, there is a disadvantage that the gradient of the traveling lane cannot be detected.
Japanese Patent No. 3711405

  The present invention has been made in view of the above background, and it is possible to detect a road shape for a vehicle that can detect the gradient of a traveling lane even when only one image portion of the left and right road dividing lines is extracted. An object is to provide an apparatus.

The present invention has been made in order to achieve the above object, and is an imaging unit that images the front of the host vehicle, and a road that divides the traveling lane of the host vehicle from an image in front of the host vehicle captured by the imaging unit. Road lane marking extraction means for extracting the image portion of the lane marking; and a tangent line at a location where the corresponding position in the real space is within a predetermined range from the location closest to the own vehicle with respect to the image portion of the road lane marking Road shape detecting means for detecting the gradient of the traveling lane of the host vehicle based on a deviation between the image part of the road segment line and the tangent in a direction on the image corresponding to the traveling direction of the host vehicle. The road shape detecting means extracts the image of the road parting line in the right half area when the road parting line extracting means extracts the image part of the road parting line in both the right half area and the left half area of the image. Extracted road area The deviation is calculated for the image portion of the line, and the deviation is calculated for the image portion of the road segment line extracted in the left half region, and the degree of bending of the traveling lane is determined according to the difference between the two deviations. Is detected .

  According to the present invention, as will be described in detail later, a tangent at a location within a predetermined range from a location where the corresponding position in the real space is closest to the host vehicle is obtained with respect to the image portion of the road marking line. The deviation between the image portion of the road segment line and the tangent in the direction on the image corresponding to the traveling direction of the own vehicle depends on whether the traveling lane is uphill or downhill. Invert. Further, the magnitude of the deviation changes depending on the grade of the traveling lane. Therefore, the road shape detecting means can detect the gradient of the traveling lane of the own vehicle based on the deviation.

Furthermore , according to the present invention, the deviation calculated for the image portion of the road segment line extracted in the right half region according to the degree of bending (curvature and bending direction) of the traveling lane of the host vehicle, The difference from the deviation calculated for the image portion of the road lane marking extracted in the half area changes. Therefore, the road shape detecting means can detect the degree of bending of the traveling lane according to the difference between the two deviations.

  Embodiments of the present invention will be described with reference to FIGS. Referring to FIG. 1, a vehicle road shape detection device 1 is mounted on a vehicle (not shown) and is imaged by a camera 2 (corresponding to the imaging means of the present invention) that images the front of the vehicle. The shape of the traveling lane is detected from the image of the road ahead of the vehicle.

  The vehicle road shape detection apparatus 1 is an electronic unit configured by a microcomputer or the like, and includes an image input circuit 10 that A / D converts a video signal output from the camera 2 and imports the image signal into an image memory 11.

  Then, by causing the microcomputer to execute a road shape detection program, the microcomputer functions as the road segment line extraction means 12 and the road shape detection means 13.

  The road lane marking extraction means 12 extracts an image portion of the road lane marking included in the captured image in front of the vehicle captured in the image memory 11. FIG. 2 exemplifies a captured image 20 in front of the vehicle, and shows an image portion LR of a travel division line on the right side of the travel lane (white line or yellow line, hereinafter referred to as a right travel segment line), and An image portion LL of a travel division line (white line or yellow line, hereinafter referred to as a left travel division line) is included.

  Then, the road marking line extraction means 12 performs straight line extraction processing such as Hough transform on the captured image 20, and classifies the image portion LR of the right road marking line that divides the right side of the traveling lane and the left side of the traveling lane. The image portion LL of the left road dividing line is extracted.

  It is necessary to extract both the image portion LR of the right road segment line and the image portion LL of the left road segment line when detecting the degree of curve of the travel lane, but when detecting the gradient of the travel lane, Any one of the image part LR of the road dividing line and the image part LL of the left road dividing line may be extracted.

  And when it is left-hand traffic, the right road dividing line (center line) is provided on most roads, but the left road dividing line is often not provided. In the case of right-hand traffic, the left road dividing line (center line) is provided on most roads, but the right road dividing line is often not provided.

  Therefore, in the present embodiment, since left-hand traffic is assumed, the image portion LR of the right road dividing line may be extracted in the right half of the captured image 20. As described above, by extracting only the image portion LR of the right road dividing line that is highly likely to exist, the gradient of the traveling lane can be detected efficiently. In the case of right-hand traffic, the image portion LL of the left dividing line may be extracted in the left half of the captured image 20.

Next, the road shape detection means 13 is the most vehicle-side in the Y-axis real space for the image part LR of the right road parting line and the image part LL of the left road parting line extracted by the road parting line extraction means 12. become (corresponding to a predetermined range of the present invention. for example, Y ₁₁ to Y ₁₃ is set to 10%, etc. of Y ₁₀ to Y _13.) from the Y ₁₃ in the range of Y ₁₁ straight set by Y = obtaining an intersection IR and IL and Y _12.

Then, the road shape detecting means 13 draws the tangent line TR of the image portion LR of the right road segment line at the right intersection point IR and the tangent line TL of the image portion LL of the left road segment line at the left intersection point IL. Then, the road shape detecting means 13 detects the deviation ΔYR between the right road dividing line LR and the tangent TR in the Y direction (corresponding to the traveling direction of the host vehicle) in the horizontal coordinates (X coordinates) X ₁₁ and X ₁₂ , and the left road A deviation ΔYL between the dividing line LL and the tangent line TL is calculated.

  Here, FIG. 3A shows a captured image 21 of the road when the traveling lane is uphill, and FIG. 3B shows a captured image 22 of the road when the traveling lane is downhill. . In FIG. 3A, the road parting line extraction means 12 extracts the image part LR_1 of the right road parting line and the image part LL_1 of the left side road parting line.

Further, the road shape detection means 13 draws the tangent line TR_1 of the image part LR_1 of the right road segment line and the tangent line TL_1 of the image part LL_1 of the left road segment line. The road shape detection device 13, the deviation of the Y-coordinate values of the image portion LR_1 right road lane marking in X ₁₂ and (LR_1 (X ₁₂₎ Y coordinate of the tangent Tr l and _{(TR_1 (X 12)) ΔYR_1} (= _{LR_1 (X 12) -TR_1 (X} 12)), and Y-coordinate values of the image portions LL_1 left road lane marking in _{X 11 (LL_1 (X 11)} ) and the Y coordinate value of the tangent TL_1 (TL_1 (X ₁₁₎ ) ΔL L (1) (= LL_1 (X ₁₁ ) −TL_1 (X ₁₁ )).

In the example shown in FIG. 3 (a), in X _12, image portions LR_1 right road lane marking is greater than the tangential Tr l. Further, in X _11, image portions LL_1 left road lane marking is greater than the tangential TL_1. Therefore, the deviations ΔYR_1 and ΔYL_1 are both positive (0 <ΔYR_1, 0 <ΔYL_1).

  Therefore, the road shape detection means 13 determines that the traveling lane is an uphill when the deviation ΔYR_1 is positive.

  If the traveling lane is not curved, ΔYR_1≈ΔYL_1. However, when the traveling lane is curved to the right, ΔYL_1> ΔYR_1 is satisfied. When the traveling lane is curved to the left, ΔYL_1 <ΔYR_1. Therefore, the road shape detection means 13 determines the bending direction of the traveling lane from the magnitude relationship between ΔYR_1 and ΔYL_1. Further, as the curvature of the traveling lane increases, the difference between ΔYR_1 and ΔYL_1 increases. Therefore, the road shape detection means 13 determines the curvature of the traveling lane based on the magnitude of the difference between ΔYR_1 and ΔYL_1.

  Next, in FIG. 3B, the road parting line extraction means 12 extracts the image part LR_2 of the right road parting line and the image part LL_2 of the left side road parting line. The road shape detecting means draws a tangent line TR_2 of the image part LR_2 of the right road segment line and a tangent line TL_2 of the image part LL_2 of the left road segment line.

The road shape detection device 13, the deviation of the Y-coordinate values of the image portion LR_2 right road lane marking in X ₁₂ and (LR_2 (X ₁₂₎₎ Y coordinate value of the tangent Tr_2 and _{(TR_2 (X 12)) ΔYR_2} ( _{= LR_2 (X 12) -TR_2 (} X 12)), and Y-coordinate values of the image portions LL_2 left road lane marking in _{X 11 (LL_2 (X 11)} ) and the Y coordinate value of the tangent TL_2 (TL_2 (X ₁₁ )) And a deviation ΔYL_2 (= LL_2 (X ₁₁ ) −TL_2 (X ₁₁ )).

In the example shown in FIG. 3 (b), in X _12, image portions LR_2 right road lane marking is smaller than the tangential Tr_2. Further, in X _11, image portions LL_2 left road lane marking is smaller than the tangential TL_2. Therefore, the deviations ΔYR_2 and ΔYL_2 are both negative (0> ΔYR_2, 0> ΔYL_2).

  Therefore, the road shape detection means 13 determines that the traveling lane is a downhill when the deviation ΔYR_2 is negative.

  If the traveling lane is not bent, the deviation is ΔYR_2≈ΔYL_2. However, when the traveling lane is turning to the right, ΔYL_2> ΔYR_2, and when the traveling lane is turning to the left, ΔYL_2 <ΔYR_2. Therefore, the road shape detection means 13 determines the bending direction of the traveling lane from the magnitude relationship between ΔYR_2 and ΔYL_2. Further, as the curvature of the traveling lane increases, the difference between ΔYR_2 and ΔYL_2 increases. Therefore, the road shape detection means 13 determines the curvature of the traveling lane based on the magnitude of the difference between ΔYR_2 and ΔYL_2.

  In this way, the data of the gradient and the degree of bending of the traveling lane detected by the vehicle road shape detection means 1 are transmitted to a control unit (ECU: Electronic Control Unit) 5 that controls the overall operation of the vehicle. . The control unit 5 performs acceleration / deceleration control according to the gradient of the traveling lane and alarm output to the driver. The control unit 5 also corrects the distance between the monitored object and the host vehicle according to the gradient of the traveling lane, performs steering control according to the degree of bending of the traveling lane, and the like.

The block diagram of the road shape detection apparatus for vehicles of this invention. Explanatory drawing of the detection process of the gradient of a driving lane, and a bending degree by the road shape detection apparatus for vehicles. Explanatory drawing of the detection process of the gradient of a driving lane, and a bending degree by the road shape detection apparatus for vehicles.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Vehicle road shape detection apparatus, 2 ... Camera, 10 ... Image input circuit, 11 ... Image memory, 12 ... Road division line extraction means, 13 ... Road shape detection means

Claims (1)

Imaging means for imaging the front of the host vehicle;
Road segment line extraction means for extracting an image portion of a road segment line that segments the traveling lane of the host vehicle from an image ahead of the host vehicle captured by the imaging unit;
A direction on the image corresponding to the traveling direction of the own vehicle by obtaining a tangent at a location within a predetermined range from a location where the corresponding position in the real space is closest to the own vehicle with respect to the image portion of the road division line Road shape detection means for detecting the gradient of the traveling lane of the host vehicle based on the deviation between the image portion of the road segment line and the tangent line ,
The road shape detection means extracts the right half area when the road part line extraction means extracts the image part of the road division line in both the right half area and the left half area of the image. The deviation is calculated for the image portion of the road lane marking that has been extracted, and the deviation is calculated for the image portion of the road lane marking that is extracted in the left half region. A road shape detecting device for a vehicle which detects a degree of lane bending .

Images