JP2016045015A - Road curvature detection device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a road curvature detection device in which a curvature of a curve road ahead of a vehicle is accurately calculated.SOLUTION: A road curvature detection device 1 comprises: an imaging part 12 that images an area ahead of a vehicle; an extraction unit 20 that extracts from an image captured by the imaging part 12 a first curve marker and a second curve marker which are made on a roadside object of a curve road ahead of the vehicle, arranged along the curve road, and have the same shape and size as each other; a disappearance point search unit 18 that searches for a disappearance point of the image, from the image captured by the imaging part; and a curvature calculation unit 22 that calculates a curvature of the curve road on the basis of an inter-pixel distance from the disappearance point to the first curve marker in the image, an inter-pixel distance from the disappearance point to the second curve marker in the image, and a shape variation rate of the second curve marker to the first curve marker in the image.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a road curvature detection device.

  2. Description of the Related Art Conventionally, a device that acquires the curvature of a curved road existing in front of a vehicle is known (for example, Patent Document 1). In the apparatus described in Patent Document 1, a pair of images are acquired by imaging a guard rail installed along a curved road with a pair of image sensors, and a plurality of windows including the guard rail are set in one image. Next, the area that most closely matches the images in the plurality of windows is searched from the other image, and the distance from the vehicle to the guardrail is calculated for each window from the difference between the image in the area and the image in the window. . Then, the curvature of the curved road is calculated based on the distance from the vehicle to the guard rail for each position corresponding to this window.

JP-A-4-262498

  As described above, in the apparatus described in Patent Document 1, it is necessary to search (match) from the other image the area that most closely matches the image area of the window in one image. However, since the guard rail has few characteristic shapes, it is difficult to ensure search accuracy (matching accuracy). For this reason, in the apparatus described in Patent Document 1, it may be difficult to accurately calculate the curvature of the road.

  For this reason, in this technical field, calculating | requiring the curvature of a road accurately is calculated | required.

  In the road curvature detection device according to one aspect of the present invention, an imaging unit that images the front of the vehicle and an image captured by the imaging unit are formed with respect to a roadside object on the curved road ahead of the vehicle. An extraction unit that extracts the first curve marker and the second curve marker having the same shape and the same size as each other, and an erasure that searches for the vanishing point of the image from the image captured by the imaging unit A point search unit, a pixel-to-pixel distance from the vanishing point to the first curve marker in the image, an inter-pixel distance from the vanishing point to the second curve marker in the image, and a first curve marker for the first curve marker in the image A curvature calculation unit that calculates the curvature of the road based on the shape change rate of the second curve marker.

  According to one aspect of the present invention, the curvature of a road can be calculated with high accuracy.

It is a block diagram which shows the function structure of the road curvature detection apparatus which concerns on one Embodiment. It is a flowchart which shows operation | movement of the road curvature detection apparatus which concerns on one Embodiment. It is an example of the image imaged by the imaging part. (A) is the example which extracted the edge component from the image shown in FIG. 3, (b) is a figure explaining the weighting of edge extraction. It is another example of the image imaged by the imaging part. It is a figure which shows typically the correspondence of each symbol. It is a figure explaining the azimuth | direction direction and angle direction of a curve marker. It is a figure which shows typically the correspondence of each symbol.

  Hereinafter, various embodiments will be described in detail with reference to the drawings. In the drawings, the same or corresponding parts are denoted by the same reference numerals.

  FIG. 1 is a block diagram illustrating a functional configuration of a road curvature detection device according to an embodiment. A road curvature detection apparatus 1 shown in FIG. 1 is an apparatus that is mounted on a vehicle and detects the curvature of a road (curved road) ahead of the vehicle.

  The road curvature detection device 1 includes an imaging unit 12, a vehicle speed sensor 14, a road curvature detection ECU 16, and a vehicle control ECU 24. The imaging unit 12 is a device that is mounted on a vehicle and images the front of the vehicle. The imaging unit 12 is a camera that includes an imaging element such as a complementary metal oxide semiconductor (C-MOS) or a charged coupled device (CCD) as a main component. The imaging unit 12 may be a stereo camera. The imaging unit 12 outputs the captured image to the road curvature detection ECU 16.

  The vehicle speed sensor 14 is a device that detects the vehicle speed of the vehicle. The vehicle speed sensor 14 may be a wheel speed sensor that is provided, for example, on a vehicle wheel and detects the rotational speed of the wheel. The vehicle speed sensor 14 outputs information regarding the detected vehicle speed to the road curvature detection ECU 16.

  The road curvature detection ECU 16 is a computer for electronic control that comprehensively controls the road curvature detection device 1, and includes a CPU (Central Processing Unit), a ROM (Read Only Memory), a RAM (Random Access Memory), and the like. It is configured. The road curvature detection ECU 16 detects the curvature of the curved road using a plurality of curve markers formed at predetermined intervals on the roadside object of the curved road. The roadside object is a structure provided on the side of the road along the road, and examples thereof include side walls, guardrails, road signs arranged at predetermined intervals, and the like. For example, the curve marker is installed side by side along a curve road so that the driver can easily recognize the curve of the curve. The plurality of curve markers have the same shape and the same size. In the present embodiment, it is sufficient that at least two curve markers are provided side by side.

  The road curvature detection ECU 16 includes a vanishing point search unit 18, an extraction unit 20, and a curvature calculation unit 22. The vanishing point searching unit 18 is a functional element that searches for the vanishing point from the image captured by the imaging unit 12. The vanishing point will be described later. The extraction unit 20 extracts a first curve marker and a second curve marker from the plurality of curve markers from the image captured by the imaging unit 12. The curvature calculation unit 22 calculates the curvature of the road based on the reduction rate in the image of the second curve marker with respect to the first curve marker. The road curvature detection ECU 16 outputs information on the detected road curvature to the vehicle control ECU 24.

  The vehicle control ECU 24 is an electronic control computer that performs various controls of the vehicle. The vehicle control ECU 24 performs driving support based on information on the curvature of the curved road acquired from the road curvature detection ECU 16. For example, the vehicle control ECU 24 sets a target speed of the vehicle based on the curvature of the road acquired from the road curvature detection ECU 16 and controls the vehicle so that the target speed is reached.

  Next, an embodiment of the road curvature detection method will be described with reference to FIG. 2 while explaining the operation of the road curvature detection device 1.

  First, in step S1, the front of the vehicle is imaged by the imaging unit 12. Next, in step S <b> 2, an image ahead of the vehicle is acquired from the imaging unit 12, and the vanishing point searching unit 18 searches for vanishing points in the image. The vanishing point is an intersection coordinate at which the edge components of the captured subject intersect each other. Here, an example in which an image in front of the vehicle as shown in FIG. In this case, when the vanishing point search unit 18 acquires an image from the imaging unit 12, the vanishing point search unit 18 extracts an area that can be regarded as a straight line from the image. Specifically, the vanishing point searching unit 18 sets a range of a road surface within 20 m from the vehicle and a range of a roadside object up to a height of 1.5 m on the side of the vehicle, and masks the region excluding the region. An image in which M is superimposed is generated. Here, the mask M is an image region having a constant luminance value.

  Next, the vanishing point searching unit 18 extracts edge components in the horizontal direction (width direction) and the vertical direction (height direction) from the image on which the mask M is superimposed. As a result, as shown in FIG. 4A, the region edge component corresponding to the position near the vehicle is extracted from the image on which the mask M is superimposed. Specifically, in FIG. 4A, edge components such as a road center line, a shoulder line, and roadside objects are extracted. As described above, the accuracy of vanishing point extraction can be improved by superimposing the mask M on a region excluding a region that can be regarded as a straight line of the image.

  As shown in FIG. 4B, the vanishing point searching unit 18 increases the weighting of the edge extraction in the horizontal direction as it approaches the horizontal center of the image, and performs the horizontal direction of the image. The weighting of edge extraction in the vertical direction may be increased as the distance from the center increases. For example, the edge extraction weight can be increased by reducing the edge extraction threshold. Next, the vanishing point searching unit 18 acquires, as the vanishing point Di, the position (coordinates) in the image at the point where the extension lines of the edge components such as the center line, the roadway outer line, and the roadside object intersect (see FIG. 3).

  Next, in step S3, a first curve marker is extracted from the plurality of curve markers provided on the roadside object. As shown in FIG. 5, a plurality of curve markers CM having a substantially arrow shape may be provided in a curve section on the road in order to improve the visibility of the curve. These curve markers CM are provided at predetermined intervals along the road curve and have the same shape. Below, the example which extracts the curve marker CM drawn with respect to the side wall provided in the roadside as a curve marker is demonstrated. In step S3, the extraction unit 20 of the road curvature detection ECU 16 extracts, for example, the curve marker farthest from the vanishing point Di of the image, that is, the curve marker CM1 closest to the vehicle as the first curve marker.

  Next, in step S4, a distance Y in the vehicle width direction between the vehicle and the curve marker CM1 is calculated. This distance Y is calculated based on the change in the appearance of the curve marker CM1 that changes as the vehicle travels. Specifically, the distance Y is calculated by the following formula (1), for example.

  In Expression (1), ΔX is the distance traveled by the vehicle from time t to time (t + Δt), and δ is the focal length of the imaging unit 12. ΔX can be calculated based on the vehicle speed output from the vehicle speed sensor 14 and the time difference Δt. When the vehicle position at time t is X and the vehicle position at time (t + Δt) is X ′, y1 represents the inter-pixel distance from the vanishing point Di in the image captured at the position X to the curve marker CM1. Y2 represents the inter-pixel distance from the vanishing point Di in the image captured at the position X ′ to the curve marker CM1. FIG. 6 is a diagram schematically illustrating a correspondence relationship between X, X ′, ΔX, y1, and y2. In FIG. 6, O indicates the center position of the imaging unit 12. The distance Y in the vehicle width direction between the vehicle and the curve marker CM1 may be measured using a stereo camera, or may be measured using a range finder such as a laser radar.

  Next, in step S5, a second curve marker different from the first curve marker is extracted, and a reduction rate (shape change rate) in the image of the second curve marker with respect to the first curve marker is calculated. Is done. The shape change rate is a change rate (enlargement rate / reduction rate) of the shape of the second curve marker with reference to the shape of the first curve marker in the image. Hereinafter, an example in which the curve marker CM1 is used as the first curve marker and the curve marker CM2 adjacent to the vanishing point Di side of the curve marker CM1 is extracted as the second curve marker will be described (see FIG. 5). Further, the shape change rate will be described as a reduction rate.

  Since the curve marker CM2 is provided at a position farther from the vehicle than the curve marker CM1, the curve marker CM2 is displayed smaller than the curve marker CM1 in the image captured by the imaging unit 12. That is, the curve marker CM2 is reduced and displayed with respect to the curve marker CM1. Using this principle, in step S5, the extraction unit 20 reduces the image corresponding to the curve marker CM1 in the azimuth direction and the angular direction, and an object similar to the reduced curve marker CM1 disappears from the curve marker CM1. Search along the direction toward the point Di. Here, as shown in FIG. 7, the azimuth direction of the curve marker CM is defined as the direction from the curve marker CM toward the vanishing point Di (that is, the traveling direction of the vehicle), and the angular direction of the curve marker CM is the azimuth direction. Is defined as the direction perpendicular to the vehicle (ie, the vehicle height direction). When the extraction unit 20 finds a subject similar to the reduced curve marker CM1, the extraction unit 20 extracts the subject as the curve marker CM2, and corresponds the reduction rate in the azimuth direction and the angle direction when the subject is found to the curve marker CM2. Add and remember.

  Note that the reduction ratio of the curve marker CM2 with respect to the curve marker CM1 may be calculated by a method different from the method described above. For example, based on the ratio of the inter-pixel distance D1 from the vanishing point Di to the curve marker CM1 and the inter-pixel distance D2 from the vanishing point Di to the curve marker CM2, the reduction ratio in the angular direction of the curve marker CM2 with respect to the curve marker CM1 May be calculated. Further, the reduction ratio in the azimuth direction of the curve marker CM2 with respect to the curve marker CM1 is calculated based on the inclination (angle) of the curve marker CM1 with respect to the traveling direction of the vehicle and the inclination (angle) of the curve marker CM2 with respect to the traveling direction of the vehicle. May be. When the positions of the curve marker CM1 and the curve marker CM2 are opposite, the enlargement rate is calculated as the shape change rate.

  Next, in step S6, the curvature calculator 22 calculates the curvature of the road based on the reduction ratio in the image of the curve marker CM2 with respect to the curve marker CM1 acquired in step S5. In the following, as shown in FIG. 8, the center position of the imaging unit 12 is O, the vehicle vicinity side end of the curve marker CM1 is R1, the vehicle far side end of the curve marker CM1 is S1, and the vehicle marker side of the curve marker CM2 The description will be made assuming that the end is R2, and the far end of the curve marker CM2 is S2.

In step S6, according to the following formulas (2), (3), and (4), the width L along the azimuth direction of the curve marker CM1 (that is, the vehicle vicinity side end R1 and the vehicle far side end of the curve marker CM1). Distance between S1 and the distance between the vehicle vicinity side end R2 and the vehicle far side end S2 of the curve marker CM2), the center position O of the imaging unit 12 and the vehicle vicinity side end of the curve marker CM1. The distance D ₀ between R1 and the distance D between the center position O of the imaging unit 12 and the vehicle vicinity side end R2 of the curve marker CM2 are calculated. Incidentally, formula (2) in ~ (4), _{a 0} indicates a width along the angular direction of the curve markers CM1 in the image (height direction), _{b 0} is the azimuth direction (the vehicle curve markers CM1 in the image The width along the traveling direction) is shown. Further, a is indicated the width along the angular direction of the curve marker CM2 in the image (height direction), y ₀ represents the inter-pixel distance from the vanishing point Di to the curve marker CM1. Note that a ₀ / a shown in the following formula (4) is the reciprocal of the reduction ratio in the angular direction of the curve marker CM2 with respect to the curve marker CM1. That is, it can be said that the distance D is calculated based on the reduction ratio in the angular direction of the curve marker CM2 with respect to the curve marker CM1.

Next, the curvature calculation unit 22 calculates the angles ψ ₁ , φ ₁ , and θ ₁ by the following formulas (5), (6), and (7). The angle ψ ₁ is an angle formed by a straight line connecting the center position O of the imaging unit 12 and the vehicle far end S1 and a straight line C along the traveling direction of the vehicle. Angle phi ₁ is a straight line connecting the center position O and the vehicle near side end portion R1 of the imaging section 12, the angle which the straight line forms connecting the center position O and the vehicle distal end portion S1 of the imaging unit 12. Angle theta ₁ is an angle formed between the line connecting the vehicle near the side end portion R1 and the vehicle distal end portion S1, and a straight line connecting the center position O and the vehicle distal end portion S1 of the imaging unit 12.

Further, the curvature calculation unit 22 calculates the angles ψ ₂ , φ ₂ , and θ ₂ by the following formulas (8), (9), and (10). The angle ψ ₂ is an angle formed by a straight line connecting the center position O of the imaging unit 12 and the vehicle far side end S2 and a straight line C along the traveling direction of the vehicle. Angle phi ₂ is a straight line connecting the center position O and the vehicle near side end portion R2 of the imaging unit 12 is the angle which the straight line forms connecting the center position O and the vehicle distal end portion S2 of the imaging unit 12. Angle theta ₂ is the angle between a straight line connecting the straight line connecting the vehicle near the side end portion R2 and the vehicle distal end portion S2, and a center position O and the vehicle distal end portion S2 of the imaging unit 12. In Expressions (8) to (10), b represents the width along the azimuth direction (vehicle traveling direction) of the curve marker CM2 in the image, and y represents the inter-pixel distance from the vanishing point Di to the curve marker CM2. Show.

Next, the curvature calculation unit 22 calculates the coordinates (p ₁ , q ₁ ) of the vehicle vicinity side end R1 of the curve marker CM1 and the vehicle vicinity side end of the curve marker CM2 according to the following formulas (11) and (12). The coordinates (p ₂ , q ₂ ) of R2 are calculated. These coordinates indicate the positions of the vehicle vicinity side end portion R1 and the vehicle vicinity side end portion R2 in an orthogonal coordinate system having the center position O of the imaging unit 12 as an origin and the vehicle width direction and the traveling direction of the vehicle as coordinate axes. Yes.

Next, the curvature calculation unit 22 calculates the coordinates (p ₀ , q ₀ ) of the curvature center CC of the curved path. The coordinates (p ₀ , q ₀ ) are a straight line perpendicular to the curve marker CM1 starting from the vehicle vicinity side end R1 and a straight line perpendicular to the curve marker CM2 starting from the vehicle vicinity side end R2. It is obtained by calculating the intersection with. Specifically, the curvature calculation unit 22 calculates the coordinates (p ₀ , q ₀ ) of the curvature center CC by the following equation (13). In Equation (13), Θ ₁ represents θ ₁ + ψ ₁ , and Θ ₂ represents θ ₂ + ψ ₂ .

  Next, the curvature calculation unit 22 calculates the curvature radius R of the curved path according to the following equation (14). Further, the curvature k of the curved road is calculated by the following equation (15).

  Returning to FIG. 2, in step S <b> 7, driving assistance corresponding to the curvature of the road is executed by the vehicle control ECU 24. For example, the vehicle control ECU 24 sets a target speed of the vehicle based on the curvature k of the road calculated in step S6, and controls the vehicle so that the target speed is reached. In step S7, when driving assistance is executed, a series of road curvature detection ends.

  As described above, the road curvature detection device 1 according to an embodiment calculates the curvature of a curved road based on the reduction rate in the image of the second curve marker with respect to the first curve marker. In this road curvature detection device 1, since the curvature of a curved road is detected from a plurality of curve markers in a single image, the curvature detection is performed as compared with the conventional technique for searching for a matching region between a pair of images. The accuracy can be improved. Further, it is possible to detect the curvature of a curved road without mounting a stereo camera that has a greater restriction for mounting on a vehicle than a monocular camera. Further, in the above embodiment, the curve marker CM drawn on the roadside object standing on the side of the curved road is extracted. Since the curve marker CM2 drawn on the roadside object can be easily recognized even from a distance, the curvature of the curve can be calculated even from a position away from the curve road. In addition, since the curve marker drawn on the roadside object is less likely to be shielded by other vehicles than when drawn on the road, the curvature detection rate can be improved.

As mentioned above, although the road curvature detection apparatus of one embodiment has been described, various modifications can be made without being limited to the above-described embodiment. For example, as long as the curvature calculation unit 22 calculates the curvature based on the reduction ratio of the first curve marker and the second curve marker, any curvature calculation procedure can be adopted. Specifically, the curvature calculation unit 22 calculates the curvature radii R _A and R _B of the road by the following equations (16) and (17), and the average of the curvature radii R _A and R _B is used as the curvature radius of the curvature. May be obtained. Alternatively, when η ₁ and η ₂ are π / 4 or more, the curvature radius of the curved path is calculated using the following formula (17), and when η ₁ or η ₂ is less than π / 4, You may calculate the curvature radius of a curved path using Formula (16). Thereby, the calculation accuracy of the curved path radius can be improved. Here, η ₁ is θ ₁ −ψ ₁ and η ₂ is θ ₂ −ψ ₂ .

  Moreover, in the said embodiment, although the curve marker CM drawn on the roadside thing was used, the curve marker does not necessarily need to be drawn on the roadside thing, for example, the structure which has the above-mentioned substantially arrow shape is used as a roadside thing. It may be attached.

  DESCRIPTION OF SYMBOLS 1 ... Road curvature detection apparatus, 12 ... Imaging part, 14 ... Vehicle speed sensor, 16 ... Road curvature detection ECU, 18 ... Vanishing point search part, 20 ... Extraction part, 22 ... Curvature calculation part, 24 ... Vehicle control ECU, CM ... Curve marker, Di ... Vanishing point.

Claims (1)

An imaging unit for imaging the front of the vehicle;
A first curve marker formed on the roadside object on the curved road ahead of the vehicle from the image taken by the imaging unit, arranged along the curved road, and having the same shape and the same size as each other And an extraction unit for extracting the second curve marker;
A vanishing point searching unit for searching for a vanishing point of the image from an image captured by the imaging unit;
The inter-pixel distance from the vanishing point in the image to the first curve marker, the inter-pixel distance from the vanishing point in the image to the second curve marker, and the first curve in the image A curvature calculator that calculates a curvature of the curved road based on a shape change rate of the second curve marker with respect to a marker;
A road curvature detection device comprising:

Images