JP2010116036A - Headlight device for vehicle - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a technique for accurately discriminating a road condition and an existing object over the wide range in the front of one's own vehicle. <P>SOLUTION: In a headlight device 110 for a vehicle, a front monitoring camera 16 has a zoom lens. An image processing ECU 32 fetches image information in the front of a vehicle taken by the front monitoring camera 16, and discriminates an attribute of an object existing in the front of the vehicle based on the image information. Camera control means 63 controls an optical magnification of a zoom lens based on a fetched vehicle speed information of the own vehicle. Headlight units 12R, 12L illuminate the front of the vehicle. A luminous intensity distribution control ECU 34 controls light distribution of the headlight units 12R, 12L according to the attribute of the object. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a vehicle headlamp device used in an automobile or the like.

In recent years, various attempts have been made to discriminate surrounding environments and objects based on surrounding information acquired by cameras and sensors mounted on the vehicle, and to perform vehicle control in accordance with the surroundings and objects. As such a sensor, for example, Patent Document 1 discloses a sensor that discriminates an oncoming vehicle, a preceding vehicle, a signal light, and the like based on the direction and color of light received by a light receiving element. Further, Patent Document 2 discloses an automobile glare sensor provided with a zoom lens system for identifying a distant light source.
Japanese Utility Model Publication No. 5-93981 JP-A-6-321007

  By the way, in recent years, there are various demands for in-vehicle cameras including the above-described sensors. For example, from the viewpoint of accident avoidance and preventive safety, identification of an object located further away is required. There is also a need for light distribution control by an appropriate headlamp unit according to the identified object. However, the object in the distance is apparently very small and difficult to identify. In particular, identification becomes more difficult when the number of pixels of the image sensor is small. Therefore, a technique for accurately identifying an object over a wide range in front of the vehicle is required.

  This invention is made | formed in view of such a condition, The objective is to provide the technique which discriminate | determines a road surface condition and an existing object accurately over the wide range ahead of the own vehicle.

  In order to solve the above-described problems, a vehicle headlamp device according to an aspect of the present invention acquires a video camera having a zoom lens and image information in front of the vehicle taken by the video camera, and based on the image information. Image processing means for discriminating attributes of an object existing in front of the vehicle, magnification control means for controlling the optical magnification of the zoom lens based on the acquired vehicle speed information of the own vehicle, a headlamp unit for illuminating the front of the vehicle, And a light distribution control means for controlling the light distribution of the headlamp unit according to the attribute of the object.

  According to this aspect, it is possible to accurately determine road surface conditions and existing objects over a wide range in front of the host vehicle. For example, it is possible to improve the discrimination accuracy of an object attribute by controlling the optical magnification of the video camera even for an object that exists in front of the host vehicle.

  The light distribution control means may control the light distribution of the headlamp unit so as to brighten the shooting range of the video camera in accordance with an increase in the optical magnification of the zoom lens. As a result, the shortage of the light amount in the photographing range accompanying the increase in the optical magnification is compensated, and it is suppressed that the accuracy of determining the attribute of the object in the distance is lowered.

  The magnification control means controls the zoom lens to a low magnification when the vehicle speed is lower than a predetermined first vehicle speed, and the vehicle speed is equal to or higher than a predetermined first vehicle speed and higher than a predetermined first vehicle speed. When the vehicle speed is lower than 2, the magnification of the zoom lens may be changed according to the vehicle speed, and when the vehicle speed is equal to or higher than the predetermined second vehicle speed, the zoom lens may be controlled at a high magnification. Thereby, selection of an appropriate optical magnification according to the vehicle speed can be realized by simple control.

  The magnification control means controls the zoom lens to a high magnification when the object determined by the image processing means does not include a preceding vehicle, at a speed higher than a third vehicle speed lower than the second vehicle speed. Also good. As a result, in a situation where the identified object does not include a preceding vehicle and it is not important to identify an object at a short distance from the host vehicle, the zoom lens is set to a higher magnification from a lower vehicle speed to It is possible to improve the discrimination accuracy of the attribute of an object existing far away from the object.

  The magnification control means may control the zoom lens to a low magnification regardless of the vehicle speed when the brightness in front of the vehicle is a predetermined value or more. As a result, for example, when driving in an urban area where the brightness in front of the vehicle is equal to or greater than a predetermined value, it is important to identify an object existing in a range close to the host vehicle. By using a low magnification, the shooting range of the video camera can be widened.

  The image processing means may calculate the brightness in front of the vehicle based on the G output among the outputs of the image sensor included in the video camera. Thereby, the brightness ahead of the vehicle can be calculated by a simple calculation process.

  The magnification control means may control the magnification of the zoom lens so that the field-of-view range matches the range that the vehicle reaches after a predetermined prediction time corresponding to the vehicle speed. As a result, the accuracy of determining the attributes of objects existing in the range reached by the vehicle after the prediction time is improved, and the driver's operation can be assisted from the viewpoints of accident avoidance and preventive safety.

  The light distribution control means may control the light distribution of the headlamp unit so as to irradiate a range from a road surface in the vicinity of the vehicle to a road surface reached by the vehicle after a predetermined prediction time. Thereby, it becomes easy to identify an object existing in a range where the vehicle reaches after the prediction time.

  The image processing means includes an optical flow calculation means for calculating an optical flow of an object existing in front of the vehicle as a light emitter or a light reflector based on the acquired luminance information of a captured image in front of the vehicle, and an object based on the optical flow. Attribute determination means for determining the attribute of Thereby, since an object can be discriminated based on the acquired captured image, appropriate light distribution control according to the attribute of the object can be performed without imposing a special operation burden on the driver.

  When the attribute determining unit determines that the attribute of the object is a moving object based on the optical flow, the attribute determining unit is based on the R output and B output of the imaging element corresponding to the moving object in the captured image among the imaging elements included in the video camera. It may be determined whether the vehicle is an oncoming vehicle or a preceding vehicle. Thereby, the discrimination accuracy of the attribute of the object which is a moving body can be further improved.

  According to the present invention, it is possible to accurately determine road surface conditions and existing objects over a wide range in front of the host vehicle. Moreover, appropriate light distribution control according to the attribute of the object existing in front of the vehicle is possible.

  In-vehicle image sensors are designed for functional requirements such as lane keeping, accident avoidance, and preventive safety. In addition, in order to use this sensor for light distribution control of a headlight unit (headlight), it is required that an object ahead of the vehicle can be identified at night in addition to the above-described functions. However, there are various objects in front of the vehicle, such as oncoming vehicles and preceding vehicles that require light distribution control that considers glare, and glare such as road lighting and delineators (gaze guidance signs). Some objects need only perform light distribution control that is optimal for the vehicle.

  In order to realize such light distribution control of the headlamp unit, a light-emitting body such as a preceding vehicle (an oncoming vehicle or a preceding vehicle) traveling on the front of the host vehicle or road lighting, or a light such as a delineator is used. It is preferable to use a sensor that detects the reflector. In addition, it is more preferable to have a function of specifying the attribute of the light emitter or light reflector detected as an object. Here, the attribute distinguishes, for example, whether the front light emitter or light reflector is a forward vehicle or a road facility. More specifically, if the illuminant is a vehicle, it is a preceding vehicle or an oncoming vehicle, and if the illuminant is a road ancillary facility, etc., it is road lighting, a delineator, or other light emitting facility. (For example, store lighting, advertisement, etc.) or a traffic signal.

  Therefore, first, the performance of the sensor (camera) necessary for realizing the above-described functions will be described using a vehicle-mounted camera (installation height; 1.2 m) of a passenger car as a specific example from the viewpoint of the field of view range and resolution.

  (1) Lane maintenance: In order to automatically maintain a lane in a traveling vehicle, a lane mark is detected. Such lane mark detection may be difficult to recognize, such as during night rain, and therefore needs to be recognized at a short distance. Therefore, a wide field of view that covers the oblique side is desired for the camera. FIG. 1 is a schematic diagram of a road arrangement in a straight road (one lane width 3.5 m). In the in-vehicle camera (1.2m above ground), it is necessary to detect the lane mark L1 on both sides of the own lane 5m ahead of the headlamp (about 7m from the camera). The angles from the optical axis of the camera are H (Horizontal) = Atan (1.75 / 7) ≈14 °, V (Vertical) = Atan (1.2 / 7) ≈10 °, and 14 ° to the camera. The above visual field range is required. As for the resolution of the sensor, no particular problem occurs because of the short distance.

  (2) Accident avoidance: Vehicles, pedestrians and obstacles in relatively short distances are subject to collision avoidance by handling. For this purpose, a field of view that covers them as a camera (sensor) is desired. Therefore, the vehicles ahead and obstacles are considered to exist in the range of own lane + one lane on both sides (Fig. 1). This is ± 5.25 m in the lateral range. The detection distance necessary for preventing these collisions (avoidance by handling) needs to be longer than the driver's cognitive reaction time. Generally, the cognitive reaction time is described as 1.0 second. If the traveling speed of the vehicle is 60 km / h, the distance is 16.7 m per second. Assuming that the horizontal angle range necessary for detection of this range is H, H = Atan (5.25 / (16.7 + 2)) ≈16 °.

  (3) Preventive safety: The detection distance of the preceding vehicle or obstacle is set to be equal to or greater than the stopping distance of the vehicle. This ensures safe driving. In addition, if a road shape or the like far from the stop distance can be detected, it can also support comfortable driving. From the viewpoint of comfortable driving, it is important to provide Preview Distance (PD), that is, how many seconds ahead of visual information can be obtained.

Table 1 shows the relationship between the vehicle speed, the curve radius on the road design, the stopping distance, and the prediction time (PD). FIG. 2 is a layout diagram on the road in a curved road having a curved radius of 150 m (turning radius in road design at a vehicle speed of 60 km / h).

The viewing range of 3.5s in Table 1 is the angle of the lane mark position (W = Atan (Curve _5.25 / PD _3.5 ) in the adjacent lane (5.25m side) 3.5s ahead on the curved road. )). If the driver is provided with information about a prediction time of 3.5 s ahead, the driver can be surely stopped and avoided after the warning, which is useful for preventive safety. Moreover, in order to guarantee comfortable driving, it is better that the foreseeing time is as long as possible, but 3.5 s is a satisfactory level. Therefore, the visual field range of the image sensor when the vehicle speed is 60 km / h or higher and preventive safety and comfort are taken into consideration is preferably secured at 16 ° or more.

(4) Distant Object Discrimination FIG. 3 is a diagram showing the position of a light emitting object on a night straight road viewed from an image sensor. In addition, the oncoming vehicle headlight HL and the preceding vehicle taillight TL in the figure are shown at positions 60 m ahead of the host vehicle. In such a perspective view, the road illumination 70 is located above the H line (Horizontal), and the delineator 72 is located slightly below the H line. These fixed objects move in the screen along the trajectories L2 and L3 in the figure while traveling. Comparing images after a lapse of a predetermined time, along this line, an optical flow {OF (Optical Flow); a representation of the motion of an object as a vector in a visual expression (usually a temporally continuous digital image)} Will occur.

  This optical flow amount (vector amount) is larger for an object at a short distance than for an image sensor. Also, the larger the relative speed difference, the larger. In other words, the on-going vehicle OF amount> the fixed object OF amount> the preceding vehicle OF amount of the traveling vehicle. Object attributes (preceding vehicle, oncoming vehicle, road lighting, delineator, etc.) can be determined from the OF amount and the object position on the road. In addition, since road illumination and delineators exist in a light spot sequence on an expressway or the like, the object discrimination accuracy can be further improved by taking this point into consideration. Further, since the rear light and the headlight are pair lights and the pair lights have the same amount of OF, the object discrimination accuracy can be further improved by taking this point into consideration.

  In the embodiment described below, the above-described object attribute discrimination is realized by an apparatus having a simpler configuration, specifically, an image sensor of a monocular camera. Further, in addition to such a simple device configuration, the following method (details will be described later) is adopted to improve the object attribute discrimination accuracy. Specifically, (a) attribute determination of a moving object by using an optical flow, (b) attribute determination of an object on the road (fixed object) by using a road area setting and a light spot sequence in addition to (a), (c ) Improvement of attribute discrimination accuracy of objects on the road by utilizing the vanishing point of the road.

  In addition to the attributes of the light emitter and light reflector (hereinafter referred to as “light emitter etc.” as appropriate), the approximate distance from the vehicle to the light emitter and the brightness (brightness) of the light emitter, etc. By calculating and using the information, it is possible to further improve the object attribute discrimination accuracy. In this case, laser radar or millimeter wave radar may be used to measure the distance between the vehicle and the light emitter. An illuminometer may be used for measuring the luminance.

(Utilization of optical flow)
When the relative positional relationship between the image sensor (camera) and the object on the road changes, the image of the object flows in consecutive captured images. Such a phenomenon is referred to as an optical flow (hereinafter referred to as “OF” as appropriate). The OF increases as the relative distance between the host vehicle and the object is shorter and the relative speed difference is larger. For example, when the host vehicle is stopped, OF corresponding to the moving object is generated. In addition, when the host vehicle is traveling, OF is generated according to road fixed objects such as road lighting and a delineator, and OF is generated according to a preceding vehicle having a speed different from that of the host vehicle. Therefore, it is possible to determine whether the attribute of the object ahead of the host vehicle is a moving object or a fixed object with respect to the road based on the size of the OF.

(Utilization of road vanishing points)
FIG. 4 is a diagram schematically showing the relationship between the vanishing point and the optical flow on the straight road on the expressway. As shown in FIG. 4, a vanishing point exists on a straight road. Here, the vanishing point can be defined as a convergence point in the perspective of painting. The vanishing point is an infinite point such as a lane mark, a roadside zone, a median strip, or a road facility (road lighting, delineator) installed in a regular arrangement. If the infinity point cannot be obtained due to the road shape (for example, a curve) or the presence of a preceding vehicle, the arrangement of those objects in the foreground is extended to infinity and the intersection point is estimated on the screen. It can also be a temporary vanishing point. The vanishing point is difficult to calculate on a curved road (not only a left / right curve but also an up / down road), but can be calculated on a left / right refraction path. Further, when the host vehicle moves in the width direction of the road due to a lane change or the like, the vanishing point moves accordingly.

(Road area setting and intersection sequence)
By analyzing image data such as lane marks, roadside zones, and median strips in front of the vehicle, road areas such as the own lane and the oncoming lane can be determined. Since road ancillary facilities such as road lighting and a delineator are regularly installed on the road, the road area is determined, so that it is possible to more accurately determine what kind of object is a light emitter or the like at night. Specifically, at night, the position of a bright spot such as a illuminant or a light spot sequence with respect to a road area (for example, own lane, oncoming lane, or shoulder), and the above-mentioned OF. It is possible to determine whether the attribute of the light emitter or the like is road illumination, delineator, preceding vehicle, or oncoming vehicle.

  The present inventor has come up with the idea that the attribute of an object can be determined by using the optical flow described above. However, as a result of further investigation, the inventors have also found that the following problems may arise. That is, the visual size of an object existing on a straight road becomes smaller as it is farther away, and converges to the intersection (HV point) of the H line and V line shown in FIG. For this reason, discrimination of an object existing in the distance is a problem. Below, the performance required for a camera including a sensor is examined from the viewpoint of brightness and resolution.

  Headlight detection distance: The current illuminance standard for passing beam glare is 1 [lx] per lamp. The driver feels dazzling when receiving more illuminance. Therefore, a vehicle traveling with a high beam needs to be able to detect the preceding vehicle at a distance where the illuminance of the headlight reaching the preceding vehicle is about 1 [lx]. The current MAX light intensity per high beam is 112500 [cd]. Since the illuminance is inversely proportional to the square of the distance, the distance to the object ahead of the host vehicle having an illuminance of 1 [lx] corresponds to 335.4 m.

  In the following, in order to simplify the calculation, the detection distance of the front object required when traveling with a high beam is assumed to be 500 m, and the resolution necessary for the image sensor at that time is examined. The interval between the left and right pair lights is normally 1.2 to 1.5 m. If the distance is 1.2 m, the visual size of the pair lights is 0.1375 ° (8.25 ′). In order to identify as a pair of pair lights, a camera having a resolution of about 5 times, that is, a resolution of 0.0275 ° (1.65 ′) is preferable in consideration of image correction due to vehicle vibration during traveling.

  Also, when passing each other at a distance of 500 m and a relative speed of 200 km / h, it approaches about 55 m per second. If the processing of the image processing apparatus is 10 times / second (camera frame rate is 30 frames / second), the distance is approximately 5.5 m. In this case, the amount of OF in the lateral direction is Atan (3.5 / 494.5) −Atan (3.5 / 500) = 0.00446 °. Also for this reason, a camera having a resolution of about 5 times, that is, a resolution of 0.0009 ° (0.054 ′) is preferable for the same reason as described above.

  In the case of an image sensor having a size of 1/3 (4.8 mm (W) × 3.6 mm (H)) inches and 2.1 million pixels (1920 W × 1080H), this device covers a field of view of ± 16 ° on the left and right. As a result, the viewing angle per pixel is 32/1920 = 0.171 ° (1 ′), which does not satisfy the above-described resolution requirement. Therefore, in order to ensure the 0.0009 ° resolution necessary for identifying the headlight ahead of 500 m, a camera having an optical zoom lens of about 20 times with 2.1 million pixels is preferable. As a result, a desired resolution can be obtained without using an image pickup device having a large number of pixels, and an image processing apparatus having a high processing capability can be eliminated, so that the cost of the entire apparatus can be expected to be reduced.

  Based on such studies, the present inventor has conceived the present invention to realize a technique for accurately discriminating road surface conditions and existing objects over a wide range in front of the host vehicle with a simple configuration.

  Hereinafter, the best mode for carrying out the present invention will be described in detail with reference to the drawings. In the description of the drawings, the same elements are denoted by the same reference numerals, and repeated descriptions are omitted as appropriate.

  FIG. 5 is a schematic view showing the appearance of a vehicle to which the vehicle headlamp device according to the present embodiment is applied. As shown in FIG. 5, the vehicle 10 according to the present embodiment includes a headlamp device 12, a control system 14 as a headlamp control device that controls light irradiation by the headlamp device 12, and the vehicle 10. Various sensors that detect information indicating the driving situation of the vehicle and output detection signals to the control system 14, a front monitoring camera 16 that monitors the front of the vehicle, and orbit signals from GPS satellites are received and output to the control system 14. And an antenna 18 to be connected.

  As various sensors, for example, a steering sensor 22 that detects the steering angle of the steering wheel 20, a vehicle speed sensor 24 that detects the vehicle speed of the vehicle 10, and an illuminance sensor 26 that detects the illuminance around the host vehicle are provided. These sensors 22, 24, 26 are connected to the control system 14 described above.

  The headlamp device applicable to the present invention is not particularly limited as long as it can change the light distribution of the irradiated light according to the attribute of the object existing in front. For example, a halogen lamp, a gas discharge headlamp, or a headlamp using an LED can be employed. In the present embodiment, a system capable of swiveling the lamp will be described as an example.

  The headlamp device 12 has a pair of left and right headlamp units 12R and 12L. The headlamp units 12R and 12L have the same structure except that the internal structure is symmetrical. The low beam lamp unit 28R and the high beam lamp unit 30R are disposed in the right lamp housing, and the low beam is disposed in the left lamp housing. A lamp unit 28L and a high beam lamp unit 30L are respectively arranged.

  The control system 14 controls the swivelable headlamp units 12R and 12L mounted on the left and right of the front portion of the vehicle, that is, the deflection direction of the irradiation direction in the left and right and up and down directions, based on the outputs of the various sensors. The headlamp device 12 capable of changing the light distribution characteristic is controlled. As such swivelable headlamp units 12R and 12L, for example, a reflector or a projector lamp provided in the headlamp unit is rotated by a driving force source such as a driving motor so as to be rotatable in a horizontal direction. Some have rotational drive means for driving. According to this type of AFS (Adaptive Front-Lighting System), when a car runs on a curved road, it becomes possible to illuminate the road ahead of the curve according to the running speed of the car. It is effective in enhancing sex.

(Vehicle headlamp device)
Next, the vehicle headlamp device according to the present embodiment will be described. FIG. 6 is a block diagram showing a schematic configuration of the vehicle headlamp device 110 according to the present embodiment. The vehicular headlamp device 110 includes headlamp units 12R and 12L, and a control system 14 that controls light irradiation by the headlamp units 12R and 12L. Then, the vehicle headlamp device 110 determines the attribute of the object existing in front of the vehicle in the control system 14, determines the light distribution control condition based on the attribute of the object, and sets the determined light distribution control condition. Based on this, the light irradiation by the headlamp units 12R and 12L is controlled.

  Therefore, the control system 14 according to the present embodiment is connected to a front monitoring camera 16 for acquiring a captured image in front of the vehicle including the driver's view target. Further, a steering sensor 22, a vehicle speed sensor 24, and an illuminance sensor 26 for detecting steering information and vehicle speed, which are referred to when determining the traveling state of the vehicle, are connected. As the illuminance sensor 26, for example, a light receiving surface installed vertically so that the vertical surface illuminance (glare amount by the artificial light source) of light received from an oncoming vehicle or an artificial light source (road lighting or store lighting) can be measured, A light receiving surface installed horizontally is used so that the horizontal illuminance of the light received from the traveling environment or from above the vehicle can be measured.

(Front surveillance camera)
The front monitoring camera 16 according to the present embodiment is a video camera having a zoom lens, and can improve the discrimination accuracy of an attribute of an object existing far away. Here, the zoom lens is a lens that can change a focal length by moving a predetermined group of lens groups. At that time, since the position of the focal plane usually changes, there are mainly the following three methods for performing the correction, and these may be appropriately selected and combined according to the desired performance.
(1) Mechanical compensation (Mechanical compensation); focal plane movement is compensated by non-linear movement of lens group by cam (2) Optical compensation (Optical compensation); focal plane even if zooming is performed by calculating the balance of refractive power (3) Electronic compensation; depends on the auto-focus function on the camera side to compensate for focal plane movement

(Control system)
The control system 14 includes an image processing ECU 32, a light distribution control ECU 34, a GPS navigation ECU 36, and an in-vehicle LAN control ECU 38. Various ECUs and various in-vehicle sensors are connected by an in-vehicle LAN bus 40 and can transmit and receive data. The image processing ECU 32 determines the attribute of the object existing ahead based on the data of the captured image acquired by the front monitoring camera 16 and various in-vehicle sensors. Note that the inside of the image processing ECU 32 is connected to each other by a high-speed bus 42. The light distribution control ECU 34 determines light distribution control conditions suitable for the traveling environment in which the vehicle is placed based on information from the image processing ECU 32 and various on-vehicle sensors, and sends the control signal to the headlamp units 12R and 12L. Output.

  The headlamp units 12R and 12L are controlled in light distribution by the control signal output from the light distribution control ECU 34 being input to the optical device driving device and the lighting control circuit of the light source. The front monitoring camera 16 is a monocular zoom camera equipped with an image sensor such as a CCD or a CMOS, and information on road alignment information necessary for driving, information on road facilities, oncoming / preceding vehicles, and information on the position and position of the vehicle from the image data. Are obtained in cooperation with other radar sensors if necessary.

  FIG. 7 is a flowchart showing a light distribution control method including object attribute discrimination processing according to the present embodiment. The discrimination of the object attribute is mainly executed by the image processing ECU 32 shown in FIG. 6, and the light distribution control is mainly executed by the light distribution control ECU 34.

  When the process is started at a predetermined timing, data output from various in-vehicle sensors (the steering sensor 22, the vehicle speed sensor 24, and the illuminance sensor 26) is acquired by the external data input unit 44 via the in-vehicle LAN bus 40 ( S10). Further, the image data of the captured image captured by the front monitoring camera 16 is acquired by the image data acquisition means 46 (S12). Note that the image data output from the front monitoring camera 16 may correspond to a monochrome image or a color image.

  The acquired image data is temporarily stored in storage means such as RAM in the image data storage means 48 (S14). The image analysis means 50 performs difference or differentiation on the density data included in the image data, and further an object having a density different from that of the surroundings (for example, a light emitter such as a headlight or road lighting, Edge processing is performed to extract the contours of light reflectors such as lane marks and delineators (S18). Note that the edge processing method is not limited thereto, and a known method may be appropriately modified and used.

  Next, the density / luminance conversion means 52 converts the density data of the image into luminance (brightness) data (S20). The conversion of the image data into luminance can be obtained in consideration of the exposure conditions if the relationship between the image density under the exposure conditions (imaging conditions) and the luminance in the range ahead of the vehicle is obtained. FIG. 8 is a diagram showing an example of a graph showing the relationship between the luminance and the image density D. As shown in FIG.

  The image analysis means 50 determines whether a lane mark (partition line) exists in the object whose contour has been extracted by edge processing (S22). If there is no lane marking, the shoulder position cannot be obtained as it is. Therefore, when it is determined that there is no lane marking (No in S22), texture analysis is performed based on the image data (S24). Texture analysis is, for example, digitizing a pattern representing regular shading changes, and different frequency characteristics are usually obtained by frequency (FFT) analysis of road surface and shoulder image data.

  FIG. 9 is a diagram schematically showing a road shoulder and a road surface pattern, and a result of texture analysis of both images. As shown in the figure, the road surface image has many fine components because of its fine texture, but the road shoulder image in which plants and road structures are present has a wide range of frequency components. Therefore, by comparing the analysis result of the predetermined area of the image data with, for example, the distribution of the frequency components of the road surface image stored in advance, the road shoulder position can be detected even when there is no lane marking. Become. That is, the position where the frequency characteristic corresponding to the road surface changes to a different frequency characteristic is estimated as the road shoulder position.

  The image analysis means 50 calculates information correlated with the road shape such as the road lane width, the shoulder position, the road shape, and the vanishing point position from the image analysis result including the lane marking detection or the texture analysis described above. The road shape is estimated (S26). FIG. 10A is a diagram showing a curved path (dotted line) that is curved to the left and right with respect to the straight road (solid line), and FIG. FIG. 10C is a schematic view of the vehicle traveling along the curved path in FIG. 10B as viewed from the side. In addition, when the lane marking cannot be detected as in rainy weather, or when the road shape is a curved road (left / right curve or up / down), the vanishing point cannot be calculated as shown in FIG. Therefore, in such a case, a virtual vanishing point may be calculated based on the steering angle data or the data from the GPS. In the present embodiment, the image analysis unit 50 functions as a road shape estimation unit.

  After the road shape is estimated, the position of a bright object such as a light emitter or a light reflector is detected. Specifically, the bright object detection means 54 extracts an object having a predetermined luminance (for example, luminance corresponding to the lane mark) or higher as the bright object by the density luminance conversion in the process of S20, and detects its position (S28). ).

  Next, the attribute of the bright object is determined by analyzing image data that is continuously acquired in time series like a moving image. First, the OF calculation means 58 acquires image characteristic data such as the position of a bright object and object candidates stored in the internal memory 56 in the previous image analysis process, and the current image characteristic data (S30). .

  The OF calculation means 58 detects whether OF has occurred in each bright object based on the image characteristic data of a plurality of captured images acquired at different times (S32). Specifically, the OF calculation unit 58 calculates the OF of the bright object existing in front of the vehicle as a light emitter or light reflector based on the acquired luminance data of the captured image in front of the vehicle. Then, the object attribute determination unit 60 determines the attribute of the object based on the OF calculated by the OF calculation unit 58 (S34). The object attribute discrimination method in S34 will be described later.

  When the attribute of the object is determined, the current image characteristic data is stored in the internal memory 56 (S36). Specifically, the position and attribute of the bright object necessary for the OF detection, the OF vector, and the like. This data is used for calculation of OF in the next image processing.

  The attribute data and road shape data of each object determined by the object attribute determining means 60 are output from the internal data output means 62 (S38). The camera control means 63 functioning as a magnification control means adjusts the focal length of the zoom lens of the front monitoring camera 16 based on data such as the attributes and road shapes of each object and vehicle speed information acquired from the vehicle speed sensor 24. The optical magnification is controlled (S39). The output data is input to the light distribution control ECU 34 via the in-vehicle LAN bus 40, and the light distribution control conditions are determined based on the attribute of the object, the traveling state of the host vehicle, the weather conditions, and the like included in the data ( S40). Then, the light distribution control ECU 34 controls the light distribution by outputting a control signal to the light sources and drive sources provided in the headlamp units 12R and 12L based on the determined light distribution control condition (S42).

  Therefore, the vehicle headlamp device 110 according to the present embodiment can accurately determine road surface conditions and existing objects over a wide range in front of the host vehicle. For example, it is possible to improve the accuracy of determining the attribute of an object by controlling the optical magnification of the zoom lens of the front monitoring camera 16 even for an object that exists far in front of the host vehicle. Further, the forward distance and foreseeing time that the driver pays attention to from the viewpoints of accident avoidance and preventive safety differ depending on the vehicle speed. Therefore, the vehicle headlamp device 110 controls the optical magnification of the zoom lens in accordance with the vehicle speed, so that an object existing in the visual field range that the driver gazes without placing a special operation burden on the driver. Attribute discrimination accuracy can be improved.

  Further, the light distribution control ECU 34 controls the light distribution of the headlamp units 12R and 12L so that the photographing range of the front monitoring camera 16 is brightened according to the increase in the optical magnification of the zoom lens. As a result, the shortage of the light amount in the photographing range accompanying the increase in the optical magnification is compensated, and it is suppressed that the accuracy of determining the attribute of the object in the distance is lowered.

  Here, the camera control means 63 controls the zoom lens at a low magnification (equal magnification) when the vehicle speed is lower than 60 km / h, and the vehicle speed when the vehicle speed is 60 km / h or higher and lower than 120 km / h. The zoom lens magnification is changed according to the above, and when the vehicle speed is 120 km / h or higher, the zoom lens is controlled to a high magnification (about 10 times). Thereby, selection of an appropriate optical magnification according to the vehicle speed can be realized by simple control. In addition, when it is necessary to detect distant objects and situations from the viewpoint of accident avoidance and preventive safety, such as when driving at high speeds, it is possible to detect a preceding vehicle that is 500 meters away from the host vehicle by such control. can do.

  Further, the camera control means 63 may control the magnification of the zoom lens so that the visual field range (focus) is in a range where the vehicle reaches after a predetermined prediction time corresponding to the vehicle speed. Specifically, the prediction time PD1 at 60 km / h is 1 second (16.6 m), and the prediction time PD2 at 120 km / h is 5 seconds (166 m). The vehicle speed during that time may be determined from a linear function of the distance and the vehicle speed according to the prediction time PD. That is, when the host vehicle is traveling at 60 km / h, the camera control means 63 controls the optical magnification of the zoom lens so that the visual field range (focus) is in the vicinity of 16.6 m ahead of the host vehicle. As a result, the accuracy of determining the attributes of objects existing in the range reached by the vehicle after the prediction time is improved, and the driver's operation can be assisted from the viewpoints of accident avoidance and preventive safety.

  In lane keeping control, detection of lane marks at a short distance is the most important. In order to ensure this, it is necessary to brightly illuminate the lane mark on the lower side (near distance) in the visual field screen. Therefore, the light distribution control ECU 34 may control the light distribution of the headlamp units 12R and 12L so as to irradiate the range from the road surface near the vehicle to the road surface reached by the vehicle after a predetermined prediction time. This makes it possible to brightly illuminate a range including the lower position of the screen field of the image captured by the front monitoring camera 16, and it becomes easy to identify an object, particularly a lane mark, existing up to a range reached by the vehicle after the preview time. .

(Object attribute discrimination method)
Next, step S34 for performing object attribute discrimination processing will be described in detail with reference to FIG. As shown in FIG. 3, there are various objects ahead of the vehicle traveling on the road. Under such circumstances, there are the following objects that the front monitoring camera 16 of the vehicle traveling at night identifies as a bright object.

  First, HL (headlight) of an oncoming vehicle traveling at night and TL (taillight) of a preceding vehicle may be mentioned. Such lights normally exist as a pair of left and right pair lights and are positioned below the horizontal line H shown in FIG. In addition, the size of OF decreases when the object is a preceding vehicle, and increases when the object is an oncoming vehicle.

  Next, the road lighting 70 and the delineator 72 are mentioned. For these fixed objects, a normal installation standard is defined. For example, road illumination is 10 m or more from the road surface, that is, above the horizontal line H (see FIG. 3). On the other hand, the delineator 72 is installed at the upper end of the guardrail and is present below the horizontal line H. Further, if the object is a fixed object, the size of OF is a value corresponding to the vehicle speed of the host vehicle and the distance between the object and the host vehicle.

  In addition to the objects described above, store lighting, advertisement neon signs, curve signs, and the like can be detected as bright objects. Store lighting and advertising neon signs are fixed objects that exist at random, and there is no regularity in the positions of road lights and delineators, so road lighting etc. Can be identified as a fixed object. In addition, it is difficult to discriminate the curve sign from the delineator because the installation position is almost the same height as the delineator. However, misidentification of the curve sign as a delineator has no effect on the light distribution control. However, no special consideration is required.

  The object discrimination method according to the present embodiment is performed on the premise of the existence of the object as described above. FIG. 11 is a flowchart showing the object discrimination method according to the present embodiment. First, the object attribute discrimination means 60 calculates the position of the bright object and the OF vector based on the detected OF (S100). Next, the object attribute discriminating means 60 calculates the relative speed V1 of the object based on the OF vector and the vanishing point information (S102) and detects the current vehicle speed V2 from the vehicle speed sensor (S104). .

  If the relative speed V1 and the host vehicle speed V2 are substantially equal even if the calculation error in the image analysis and the detection error of the sensor are taken into consideration (Yes in S106), the object is determined as a fixed object (S108). The object attribute determination means 60 further determines whether or not the position where the object exists is above the horizon (S110). If the position of the object is above the horizon (Yes in S110), the object is road illumination. If it is determined (S112) and the position of the object is below the horizontal line (No in S110), it is determined that the object is a delineator (S114).

  When the relative speed V1 with respect to the object and the host vehicle speed V2 are different (No in S106), it is estimated that the object is not a fixed object but a moving body. Therefore, it is determined whether the relative speed V1 when the object approaches the host vehicle is positive and the relative speed V1 is higher than the host vehicle speed V2 (S116). The object attribute determination means 60 determines that the object is an oncoming vehicle if the relative speed V1 is greater than the host vehicle speed V2 (Yes in S116) (S118), and if the relative speed V1 is less than the host vehicle speed V2 ( In S118, it is determined that the object is a preceding vehicle (S120).

  Thereafter, the attribute of the determined object is stored for each object (S122), and it is determined whether the determination of all objects in the image data has been completed (S124). When all (all) objects have not been determined (No in S124), the process returns to S100. When all (all) objects have been determined (Yes in S124), the series of processes is once ended.

  In the above-described object discrimination processing based on the OF vector and the object position information, the road shape information calculated in step S26 and the characteristic information (highway, etc.) of the currently traveling road acquired from the GPS navigation ECU 36 are used. The road type, the number of lanes, whether it is an urban area or a suburb, etc.), the discrimination accuracy of store lighting, advertisement neon signs, curve signs, etc. is improved.

  The front monitoring camera 16 may be a color camera that can determine the color of a bright object. In such a color camera, by calculating the output ratio of the R (red) element and B (blue) element corresponding to the position of the bright object of the sensor, the preceding vehicle, oncoming vehicle, It becomes easy to distinguish traffic signals, road lighting, store lighting, and the like.

  In the image data, the output ratio R / B (TL) of the R element and B element of the sensor corresponding to the position of the taillight TL (leading vehicle, red), and the position of the headlight HL (oncoming vehicle, white) When the output ratio R / B (HL) of the sensor R element and the B element corresponding to is compared, there is a relationship of R / B (TL) >> R / B (HL). Therefore, by determining the R / B for each color light in advance through experiments and simulations, it is possible to improve the discrimination accuracy between the preceding vehicle and the oncoming vehicle. In particular, this contributes to an improvement in discrimination accuracy on a curved road where the positions of the preceding vehicle and the oncoming vehicle easily cross.

  As described above, the control system 14 according to the present embodiment can discriminate an object based on the acquired captured image. Therefore, an appropriate operation according to the attribute of the object can be performed without imposing a special operation burden on the driver. Light distribution control is possible. Based on the data acquired from the image processing ECU 32, for example, the light distribution control ECU 34 automatically sets the light distribution of the headlamp unit to a high beam when determining that there is no bright object in front of the vehicle. When it is determined that an oncoming vehicle or a preceding vehicle exists as a bright object, it is possible to perform light distribution control such that the light distribution of the headlamp unit is automatically changed to a low beam.

  Also, if the light distribution control ECU 34 determines that there are only road illuminations and delineators that could not be identified as oncoming vehicles or preceding vehicles, the light distribution control ECU 34 maintains the light distribution of the headlamp unit in a high beam state. Thus, the lighting capacity of the headlamp unit can be utilized more effectively.

  Further, since the image processing ECU 32 includes the image analysis means 50 for estimating the road shape in front of the vehicle, the attribute of the object can be determined based on the OF and the road shape of the object, and the road shape is curved. Even on a road or a curved road, it is possible to accurately determine the attribute of the object.

  The object attribute determining means 60 can determine the attribute of the object by comparing the relative speed V1 between the host vehicle and the object estimated from the OF of the object and the host vehicle speed V2. Therefore, more appropriate light distribution control can be performed.

  Specifically, by comparing the relative speed V1 and the host vehicle speed V2, it is determined whether the attribute of the object is a fixed object or a moving object with respect to the road. Thereby, for example, when the relative speed V1 with respect to the object is substantially equal to the own vehicle speed V2, the object attribute determination means 60 determines that the object is a fixed object such as road illumination or a delineator, When the relative speed V1 of the vehicle is different from the own vehicle speed V2, it can be determined that the object is a moving body such as an oncoming vehicle or a preceding vehicle.

  Further, the object attribute determination means 60 can determine whether the attribute of the object is road illumination or a delineator depending on the position of the object or where the OF is in the vertical direction of the captured image.

  Further, since the image analysis means 50 can estimate the road shoulder position of the road by texture analysis by fast Fourier transform with respect to the captured image in front of the vehicle, for example, even if the road surface has no lane mark, the road surface and the road shoulder portion. Even if the control system 14 is provided in a vehicle traveling on such a road surface, it is possible to accurately determine the attribute of the object.

  According to the control system 14 as the headlight control device and the vehicle headlight device 110 having the headlight control device according to the present embodiment, the details are particularly according to the nighttime traffic viewing environment (lighting conditions, oncoming vehicle conditions). Light distribution control is possible, contributing to the prevention of traffic accidents. In addition, the control system 14 according to the present embodiment can determine the attribute of a bright object by using a monocular camera without using a plurality of cameras or a camera having a complicated mechanism. The overall cost can be reduced.

  The vehicle headlamp device 110 according to the present embodiment can perform the following control based on the attribute of the object determined by the above-described method. For example, when the object determined by the image processing ECU 32 does not include a preceding vehicle, the camera control unit 63 sets the vehicle speed to 120 km / h (usually the first vehicle speed at which the optical magnification is set to a high magnification). ) Control the zoom lens at a high magnification when the vehicle speed is higher than a predetermined vehicle speed. As a result, in a situation where the identified object does not include a preceding vehicle and it is not important to identify an object at a short distance from the host vehicle, the zoom lens is set to a higher magnification from a lower vehicle speed to It is possible to improve the discrimination accuracy of the attribute of an object existing far away from the object.

  On the other hand, when the brightness in front of the vehicle is greater than or equal to a predetermined value, the camera control means 63 may control the zoom lens to a low magnification regardless of the vehicle speed. As a result, for example, when driving in an urban area where the brightness in front of the vehicle is equal to or greater than a predetermined value, it is important to identify an object existing in a range close to the host vehicle. By making the magnification low, the photographing range of the front monitoring camera 16 can be widened. At this time, the image processing ECU 32 may calculate the brightness in front of the vehicle based on the output of the G element among the outputs of the imaging element included in the front monitoring camera 16. Thereby, the brightness ahead of the vehicle can be calculated by a simple calculation process.

  As described above, the present invention has been described with reference to the above-described embodiment. However, the present invention is not limited to the above-described embodiment, and the present invention can be appropriately combined or replaced with the configuration of the embodiment. It is included in the present invention. In addition, it is possible to appropriately change the combination and processing order in the embodiment based on the knowledge of those skilled in the art and to add various modifications such as various design changes to the embodiment. The described embodiments can also be included in the scope of the present invention.

It is a schematic diagram of arrangement | positioning on a road in a straight road (one lane width 3.5m). It is a road arrangement | positioning figure in the curved road (turning radius on the road design in the vehicle speed of 60 km / h) of the curve radius 150m. It is the figure which showed the position of the light emission object in the night straight road seen from the image sensor. It is the figure which showed typically the relationship between the vanishing point in the straight road of an expressway, and an optical flow. It is the schematic which shows the external appearance of the vehicle to which the vehicle headlamp apparatus which concerns on this Embodiment is applied. It is a block diagram which shows schematic structure of the vehicle headlamp apparatus which concerns on this Embodiment. It is a flowchart which shows the light distribution control method containing the object attribute discrimination | determination process based on this Embodiment. It is a figure which shows an example of the graph which shows the relationship between a brightness | luminance and the image density D. FIG. It is a figure which shows the result of having analyzed the texture of both images typically while showing the pattern of a road shoulder and a road surface. FIG. 10A is a diagram showing a curved road (dotted line) that is curved to the left and right with respect to the straight road (solid line), and FIG. 10B is an ascending / descending direction with respect to the straight road (solid line). FIG. 10C is a schematic view of the vehicle traveling along the curved path in FIG. 10B as viewed from the side. It is a flowchart which shows the object discrimination method which concerns on this Embodiment.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 Vehicle, 12 Headlamp device, 12R Headlamp unit, 14 Control system, 16 Front monitoring camera, 18 Antenna, 20 Steering wheel, 22 Steering sensor, 24 Vehicle speed sensor, 26 Illuminance sensor, 32 Image processing ECU, 34 Arrangement Light control ECU, 36 GPS navigation ECU, 38 In-vehicle LAN control ECU, 40 In-vehicle LAN bus, 42 High-speed bus, 44 External data input means, 46 Image data acquisition means, 48 Image data storage means, 50 Image analysis means, 52 Density luminance Conversion means 54 Bright object detection means 56 Internal memory 58 OF calculation means 60 Object attribute determination means 62 Internal data output means 63 Camera control means 70 Road illumination 72 Delineator 110 Car Dual-use headlamp device.

Claims (10)

A video camera having a zoom lens;
Image processing means for acquiring image information in front of the vehicle photographed by the video camera, and determining an attribute of an object existing in front of the vehicle based on the image information;
Magnification control means for controlling the optical magnification of the zoom lens based on the acquired vehicle speed information of the host vehicle;
A headlamp unit that illuminates the front of the vehicle;
Light distribution control means for controlling the light distribution of the headlamp unit according to the attribute of the object;
A vehicle headlamp device comprising:   The light distribution control means controls the light distribution of the headlight unit so as to brighten the photographing range of the video camera in accordance with an increase in optical magnification of the zoom lens. The vehicle headlamp device described. The magnification control means includes
When the vehicle speed is lower than a predetermined first vehicle speed, the zoom lens is controlled to a low magnification,
When the vehicle speed is lower than a predetermined second vehicle speed that is higher than the predetermined first vehicle speed and higher than the predetermined first vehicle speed, the magnification of the zoom lens is changed according to the vehicle speed,
When the vehicle speed is equal to or higher than a predetermined second vehicle speed, the zoom lens is controlled at a high magnification.
The vehicle headlamp device according to claim 1 or 2, wherein   When the object determined by the image processing unit does not include a preceding vehicle, the magnification control unit sets the zoom lens to a high magnification when the vehicle speed is equal to or higher than a third vehicle speed that is lower than the second vehicle speed. The vehicle headlamp device according to claim 3, wherein the vehicle headlamp device is controlled as follows.   5. The vehicle headlamp device according to claim 3, wherein the magnification control unit controls the zoom lens to a low magnification regardless of a vehicle speed when the brightness in front of the vehicle is equal to or greater than a predetermined value. .   The vehicle headlamp device according to claim 5, wherein the image processing unit calculates a brightness in front of the vehicle based on a G output among outputs of an image sensor provided in the video camera.   7. The magnification control unit controls the magnification of the zoom lens so that a field of view range matches a range that the vehicle reaches after a predetermined prediction time corresponding to the vehicle speed. The vehicle headlamp device described in 1.   The light distribution control means controls the light distribution of the headlamp unit so as to irradiate a range from a road surface in the vicinity of the vehicle to a road surface reached by the vehicle after the predetermined prediction time. The vehicle headlamp device described in 1. The image processing means includes
Optical flow calculation means for calculating an optical flow of an object existing in front of the vehicle as a light emitter or light reflector based on the luminance information of the captured image in front of the vehicle;
Attribute determination means for determining the attribute of the object based on the optical flow;
The vehicle headlamp device according to claim 1, comprising:   When the attribute determining unit determines that the attribute of the object is a moving object based on the optical flow, the R output of the image sensor corresponding to the moving object in the captured image among the image sensors included in the video camera and The vehicle headlamp device according to claim 9, wherein it is determined whether the vehicle is an oncoming vehicle or a preceding vehicle based on the B output.

Images