JP6549973B2 - Outside environment recognition device - Google Patents

Description

  The present invention relates to an external environment recognition device that performs light distribution control of a headlamp according to the external environment of the vehicle.

  Conventionally, a three-dimensional object such as a vehicle located in front of the host vehicle is detected to avoid a collision with the preceding vehicle (collision avoidance control) or to control the distance between the preceding vehicle to be a safe distance Cruise control) technology is known (for example, patent document 1).

  In addition, for safe driving at night, adoption of an auto light function has been in progress that automatically turns on the headlights when the brightness outside the vehicle is not sufficient. There is also known a technique of determining the brightness of the surrounding environment based on exposure information obtained from a monitor sensor and utilizing it for auto light (for example, Patent Document 2).

Patent No. 3349060 Unexamined-Japanese-Patent No. 11-187390

  When it is desired to make the headlamp a high beam, the driver switches the dimmer switch from high beam non-enabled to high beam enabled. Therefore, the lighting device can emit the high beam when the dimmer switch is high beam enabled.

  However, in the case of using light distribution control such as HBA (High Beam Assist) or ADB (Adaptive Driving Beam), the driver may always set the main switch as an auto light and may always set the dimmer switch as a high beam. In this case, it may not be possible to simply judge that the driver desires to turn on the high beam simply by the fact that the dimmer switch is high beam enabled.

  An object of the present invention is to provide an outside environment recognition device capable of quickly determining whether a high beam is unnecessary or not in view of such a problem.

  In order to solve the above-mentioned problems, according to the outside environment recognizing device of the present invention, the computer adds and subtracts so that the lower the illuminance is, the larger the value becomes as compared with the case where the illuminance is higher. The integrated value deriving unit that derives the value and derives an integrated value obtained by integrating the addition and subtraction values, and determines that the high beam can be used when the integrated value becomes equal to or more than a predetermined first integrated threshold, and the integrated value is the first integrated threshold It functions as a necessity determination unit that determines that the high beam is unnecessary when it is smaller than the smaller second integration threshold, and the integrated value deriving unit determines the integrated value in accordance with the dimmer switch switching from high beam disabled to high beam enabled. It is characterized by switching to a value.

  The predetermined value is a value equal to or more than the second integration threshold and less than the first integration threshold, and the integration value deriving unit determines that the integration value is less than the predetermined value in response to the dimmer switch switching from high beam disable to high beam enable. If there is, the integrated value may be switched to a predetermined value, and if the integrated value is equal to or greater than the predetermined value, the integrated value may not be switched.

  According to the present invention, it is possible to quickly determine whether a high beam is unnecessary.

It is a block diagram showing the connection relation of the environment recognition system outside a car. It is a functional block diagram showing a schematic function of an outside environment recognition device. It is a flowchart which shows the flow of an environment recognition process outside a vehicle. It is an explanatory view for explaining a luminosity picture and a distance picture. It is an explanatory view showing an example of judgment processing of a necessity judgment part. It is an explanatory view for explaining a detection range. It is explanatory drawing for demonstrating the brightness | luminance image from which exposure time differs. It is explanatory drawing which showed the relationship between a tail lamp and a red reflective board. It is another explanatory view showing a relation between a tail lamp and a red reflector. It is explanatory drawing for demonstrating the point which a preceding vehicle extraction part accumulates. It is a functional block diagram for demonstrating light distribution control of a headlamp. It is an upper view surface view for demonstrating the relationship between a cutline angle and an angle of view. It is explanatory drawing for demonstrating the table of a cut line angle. It is explanatory drawing which showed operation | movement of a light distribution control part. It is an explanatory view explaining light distribution control of ADB.

  Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. The dimensions, materials, and other specific numerical values and the like shown in the embodiment are merely examples for facilitating the understanding of the invention, and do not limit the present invention unless otherwise specified. In the specification and the drawings, elements having substantially the same functions and configurations will be denoted by the same reference numerals to omit repeated description, and elements not directly related to the present invention will not be illustrated. Do.

(Outside environment recognition system 100)
FIG. 1 is a block diagram showing the connection relationship of the outside environment recognition system 100. As shown in FIG. The external environment recognition system 100 includes an imaging device 110, an external environment recognition device 120, and a vehicle control unit (ECU: Engine Control Unit) 130.

  The imaging device 110 is configured to include an imaging element such as a charge-coupled device (CCD) or a complementary metal-oxide semiconductor (CMOS), captures an environment outside the vehicle 1 and includes at least information on luminance. A luminance image (color image or monochrome image) can be generated. Moreover, the imaging devices 110 are spaced apart in the substantially horizontal direction so that the optical axes of the two imaging devices 110 are substantially parallel on the traveling direction side of the host vehicle 1. The imaging device 110 continuously generates a luminance image obtained by imaging a three-dimensional object present in a detection area in front of the host vehicle 1 every frame (60 fps) of, for example, 1/60 seconds. Here, since luminance images of different viewpoints are generated by the two imaging devices 110, it is possible to grasp the distance of a three-dimensional object. Here, the three-dimensional object recognized by the imaging device 110 is only a thing existing independently such as a vehicle (preceding vehicle, oncoming vehicle), a pedestrian, a street light, a traffic light, a road (traveling road), a road sign, a guardrail, and a building. It also includes items that can be identified as part of it.

  The outside environment recognition device 120 acquires a luminance image from each of the two imaging devices 110, and searches for a block corresponding to a block (an aggregate of a plurality of pixels) arbitrarily extracted from one luminance image from the other luminance image A so-called pattern matching is used to derive parallax (depth distance) and a screen position indicating the position of an arbitrary block in the screen, and derive a three-dimensional position of each block. Then, the external environment recognition device 120 specifies a three-dimensional object existing in the external environment, for example, a preceding vehicle traveling in the same direction, or an oncoming vehicle traveling in the opposite direction. Further, when the three-dimensional object is specified in this way, the outside environment recognition device 120 avoids the collision with the three-dimensional object (collision avoidance control), or keeps the inter-vehicle distance with the preceding vehicle at a safe distance. Control (cruise control).

  Further, the external environment recognition device 120 receives a driver's request (intention) through the illumination switch 122, and performs light distribution control of a headlamp or the like through the illumination mechanism 124 according to the external environment. As such light distribution control, for example, high beam is not to be irradiated, when there is a three-dimensional object such as a preceding vehicle or an oncoming vehicle ahead, the high beam is turned off, otherwise it is turned on Or, if the area irradiated with high beam is variable, and there is a solid object that should not be irradiated with high beam, only the area is not irradiated with high beam, and other solid objects such as streetlights, road signs, signs, and reflectors In the area where there will be, there is an ADB (Adaptive Driving Beam) that emits a high beam. In order to realize such light distribution control, for example, as a lighting switch 122, a main switch that switches the position to any one of the lighting state of the lamp, extinguishing, small lamp (position lamp), lighting (low beam), and auto light A dimmer switch is provided to switch the position to either high beam impossibility or high beam impossibility. The illumination mechanism 124 has a mechanism for switching between a low beam and a high beam in the case of the HBA, and has a mechanism for changing the area of the high beam in the case of the ADB.

  In addition, the vehicle control device 130 receives the operation input of the driver through the steering wheel 132, the accelerator pedal 134, and the brake pedal 136, and controls the vehicle 1 by transmitting it to the steering mechanism 142, the drive mechanism 144, and the braking mechanism 146. . Further, the vehicle control device 130 controls the steering mechanism 142, the drive mechanism 144, and the braking mechanism 146 in accordance with an instruction of the external environment recognition device 120.

  As described above, in the external environment recognition system 100, the leading vehicle or the oncoming vehicle can be quickly and accurately controlled to cruise control the leading vehicle or to avoid irradiating the leading vehicle or the oncoming vehicle with a high beam. Identification is required. In the external environment recognition system 100, the leading vehicle and the oncoming vehicle are identified quickly and accurately by acquiring three-dimensional position information and color information through the luminance image by the two imaging devices 110, and the headlamp is appropriately selected. The purpose is to control light distribution.

  Hereinafter, the configuration of the external environment recognition device 120 for achieving such an object will be described in detail. Here, the light distribution control of the headlamp, which is characteristic of the present embodiment, will be described in detail, and the description of the configuration irrelevant to the features of the present embodiment will be omitted.

(Exterior environment recognition device 120)
FIG. 2 is a functional block diagram showing a schematic function of the external environment recognition device 120. As shown in FIG. As shown in FIG. 2, the external environment recognition device 120 includes an I / F unit 150, a data holding unit 152, and a central control unit 154.

  The I / F unit 150 is an interface for performing two-way information exchange with the imaging device 110 and the vehicle control device 130. The data holding unit 152 is configured by a RAM, a flash memory, an HDD, and the like, and holds various information necessary for the processing of each functional unit described below.

  The central control unit 154 is constituted by a semiconductor integrated circuit including a central processing unit (CPU), a ROM in which programs and the like are stored, and a RAM as a work area, and the I / F unit 150 and the data holding unit Control 152 and the like. Further, in the present embodiment, the central control unit 154 includes the image processing unit 160, the three-dimensional position deriving unit 162, the integrated value deriving unit 164, the necessity determination unit 166, the detection range setting unit 168, the leading vehicle extraction unit 170, It also functions as a vehicle recognition unit 172, an oncoming vehicle extraction unit 174, an oncoming vehicle recognition unit 176, a streetlight extraction unit 178, a streetlight recognition unit 180, a traveling scene determination unit 182, and a light distribution control unit 184. Hereinafter, the external environment recognition processing for controlling the light distribution of the headlamp, which is characteristic of the present embodiment, will be described in detail based on the operation of each functional unit of the central control unit 154.

(Outside environment recognition processing)
FIG. 3 is a flowchart showing the flow of the outside environment recognition process. In the environment recognition process outside the vehicle, the image processing unit 160 processes the image acquired from the imaging device 110 (S200), and the three-dimensional position deriving unit 162 derives a three-dimensional position from the image (S202). If it is determined that the high beam is not necessary (YES in S204), the external environment recognition processing is terminated.

  If it is determined that the high beam is not unnecessary (NO in S204), the detection range setting unit 168 sets the detection ranges of the tail lamp, the head lamp, and the street light in the acquired image (S206). The unit 170 extracts a tail lamp from the preceding vehicle detection range (S208), the preceding vehicle recognition unit 172 recognizes the preceding vehicle (S210), and the oncoming vehicle extraction unit 174 extracts a headlamp from the oncoming vehicle detection range (S212) The oncoming vehicle recognition unit 176 recognizes the oncoming vehicle (S214), the streetlight extraction unit 178 extracts a streetlight from the streetlight detection range (S216), and the streetlight recognition unit 180 recognizes the street light (S218) And the traveling scene determination unit 182 determines whether or not the traveling scene is capable of irradiating the high beam from the position information of the streetlight and the like (S220). , The light distribution control unit 184, a preceding vehicle, an oncoming vehicle, and performs the light distribution control of the headlamp on the basis of the traveling scene (S222), and finishes the environment outside the vehicle recognition processing. The individual processes will be described in detail below.

(Image processing S200)
The image processing unit 160 acquires a luminance image from each of the two imaging devices 110, and a block corresponding to a block arbitrarily extracted from one luminance image (for example, an array of 4 horizontal pixels × 4 vertical pixels) is used as the other luminance image. Parallax is derived using so-called pattern matching, which is retrieved from. Here, "horizontal" indicates the screen horizontal direction, and "vertical" indicates the screen vertical direction.

  As this pattern matching, it is conceivable to compare the luminance (Y color difference signal) in block units indicating an arbitrary image position between two luminance images. For example, SAD (Sum of Absolute Difference) which takes a difference of luminance, SSD (Sum of Squared intensity Difference) which uses the difference as a square, NCC which takes similarity of the dispersion value which subtracted the average value from the luminance of each pixel There are methods such as (Normalized Cross Correlation). The image processing unit 160 performs such block-based parallax derivation processing on all the blocks displayed in the detection area (for example, 600 horizontal pixels × 180 vertical pixels). Here, the block is set to 4 horizontal pixels × 4 vertical pixels, but the number of pixels in the block can be set arbitrarily.

  However, although the image processing unit 160 can derive parallax for each block which is a detection resolution unit, it can not recognize what part of a three-dimensional object the block is. Therefore, the parallax is independently derived in detection resolution units (for example, block units) in the detection area, not in solid object units. Here, an image in which the parallax information (corresponding to the depth distance z described later) derived in this manner is associated with the luminance image is referred to as a distance image.

  FIG. 4 is an explanatory view for explaining the luminance image 212 and the distance image 214. As shown in FIG. For example, it is assumed that a luminance image 212 as shown in FIG. 4A is generated for the image area 216 through the two imaging devices 110. However, here, in order to facilitate understanding, only one of the two luminance images 212 generated by each of the imaging devices 110 is schematically shown. In the present embodiment, the image processing unit 160 obtains the parallax for each block from such a luminance image 212, and forms a distance image 214 as shown in FIG. 4B. Each block in the distance image 214 is associated with the parallax of that block. Here, for convenience of explanation, the block from which the parallax is derived is represented by black dots.

(Three-dimensional position derivation process S202)
Subsequently, the three-dimensional position deriving unit 162 generates parallax information for each block in the image area 216 on the basis of the distance image 214 generated by the image processing unit 160 using the so-called stereo method. Transform to a three-dimensional position in real space including y and depth distance (relative distance) z. Here, the stereo method is a method of deriving the depth distance z with respect to the imaging device 110 of the three-dimensional part from the parallax in the distance image 214 of the three-dimensional part (block consisting of pixels or plural pixels) by using triangulation. is there. At this time, the three-dimensional position deriving unit 162 is based on the depth distance z of the three-dimensional part and the detection distance on the distance image 214 between the three-dimensional part and a point on the road surface at the same depth distance z as the three-dimensional part. The height y of the three-dimensional part from the road surface is derived. Then, the derived three-dimensional position is associated with the distance image 214 again. Since various known techniques can be applied to the process of deriving the depth distance z and the process of specifying the three-dimensional position, the description thereof is omitted here.

(Necessity determination processing S204)
Next, the necessity determination unit 166 determines whether the outside of the vehicle is bright (day or night), that is, whether the high beam of the headlamp is unnecessary. Then, if the necessity determination unit 166 determines that the high beam is unnecessary (hereinafter, such a state is simply referred to as a high beam unnecessary state), the subsequent light distribution control S206 to S222 of the headlamp is omitted. Thus, the processing load can be reduced.

  By the way, in the imaging device 110, the exposure amount adjustment unit (not shown) adjusts the exposure amount based on the environment outside the vehicle. Here, the exposure amount can be calculated based on the sensitivity (gain in the present embodiment), the aperture, and the exposure time. For example, the exposure amount adjustment unit uses the luminance distribution of a partial area (for example, a road surface area) of the generated image, and if the luminance of the area is high, the gain and the exposure time are reduced and the luminance of the area is low. Gain and exposure time. That is, the gain and the exposure time are adjusted so as to obtain luminance suitable for recognition of various three-dimensional objects. The exposure adjustment unit functions as an illuminance deriving unit that derives the illuminance outside the vehicle, and indicates that the illuminance is low when the exposure amount is large, and indicates that the illuminance is high when the exposure amount is small. In the following embodiment, the illuminance outside the vehicle is represented by the exposure amount, and the threshold of the illuminance (illuminance threshold) is represented by the threshold of the exposure amount (exposure threshold). For example, the illuminance obtained from the image or the illuminance obtained from the illuminance sensor It goes without saying that it can be considered in place of.

  Therefore, the necessity determination unit 166 can grasp the external brightness by referring to the exposure amount adjusted by the exposure amount adjustment unit of the imaging device 110. For example, if the gain is small and the exposure time is short (if the amount of exposure is small) as compared with a predetermined threshold, it can be judged that the outside of the vehicle is bright (high illuminance) high beam unnecessary state. If the time is long (if the exposure amount is large), it can be determined that the outside of the vehicle is dark (illuminance is low) and a high beam is available (hereinafter, such a state is simply referred to as a high beam permission state).

  However, if the exposure amount is simply compared with the threshold value, the head lamp may be determined if the high beam unnecessary state is frequently inverted (chattering) in frame units, for example, according to the change in gain or exposure time. There is a problem that the lighting condition of the lamp is not stable. Therefore, in the present embodiment, information of a relatively long period is collected in time series, and it is determined based on the result whether or not it is a high beam unnecessary state. Specifically, the integrated value deriving unit 164 derives an addition / subtraction value (for example, a value within the range of −15 to +15) according to the exposure amount, and integrates an addition / subtraction value for each predetermined cycle (frame) For example, a value in the range of 0 to 10000 is derived. Then, based on the integrated value derived in this manner, the necessity determination unit 166 determines, for example, the high beam unnecessary state if the integrated value is small, and determines the high beam permitted state if the integrated value is large. Also, in order to avoid frequent reversal of the determination as to whether the high beam unnecessary state or not, a hysteresis characteristic is provided to the threshold for comparing the integrated values.

  FIG. 5 is an explanatory view showing an example of the determination process of the necessity determination unit 166. As shown in FIG. The necessity determination unit 166 derives a negative addition / subtraction value if the integrated value deriving unit 164 determines that the exposure amount is less than a predetermined exposure threshold (for example, the median of the exposure amount and the median of TH3 and TH4). If the amount is equal to or greater than the exposure threshold, a positive addition / subtraction value is derived. Further, as shown in FIG. 5A, if the integrated value is less than a predetermined third integrated threshold (for example, 5000), the necessity determination unit 166 adds or subtracts an exposure amount larger than the exposure threshold than the exposure threshold. The addition / subtraction value is derived so that the absolute value is larger than the addition / subtraction value for a small exposure amount, and if the integrated value is equal to or more than the third integration threshold, the addition / subtraction value for the exposure amount larger than the exposure threshold is for the exposure amount smaller than the exposure threshold The addition / subtraction value is derived so that the absolute value is smaller than the addition / subtraction value.

  Seeing from another viewpoint, if the integrated value deriving unit 164 determines that the integrated value is smaller than the predetermined third integrated threshold if the integrated quantity is equal to or larger than the predetermined exposure threshold, the integrated value is equal to The addition / subtraction value is derived such that the absolute value is larger than the addition / subtraction value for the exposure amount in the case of the third integration threshold or more, and the integration value is less than the predetermined third integration threshold if the exposure amount is less than the exposure threshold The addition / subtraction value is derived such that the absolute value becomes smaller than the addition / subtraction value for the exposure amount when the integration value is equal to or more than the predetermined third integration threshold value. Here, TH1 to TH6 indicate respective threshold values within the exposure amount range, and have a relationship of TH1> TH2> TH3> TH4> TH5> TH6 (the larger the value, the darker the image is).

  Here, the following characteristics can be obtained by finely dividing the threshold value of the exposure amount and setting addition and subtraction values individually for the exposure amount. That is, as shown in FIG. 5 (b), in a scene (16 o'clock to 20 o'clock) in which the brightness outside the vehicle is brighter than in a bright scene (16 o'clock to 20 o'clock), as shown in FIG. In the state of less than), the integrated value approaches the third integrated threshold according to the exposure amount. At this time, since the absolute value of the addition / subtraction value is larger as the exposure amount is larger, the integrated value will more rapidly approach the third integration threshold when the exposure amount is darker. However, when the integrated value becomes equal to or greater than the third integrated threshold, for example, in a state where the exposure amount is less than TH1, the increasing speed of the integrated value decreases and stagnates near the third integrated threshold. When the outside of the vehicle becomes sufficiently dark (exposure amount 量 TH1), the integrated value becomes equal to or greater than the first integrated threshold (for example, 6000), and the necessity determination unit 166 determines that the high beam is permitted as shown in FIG. It will be done.

  With this configuration, the integrated value is 10000 in a sufficiently dark scene, and the integrated value is 0 in a sufficiently bright scene, and the integrated value is approximately 5000 at an intermediate brightness between the two. Therefore, if the exposure amount is slightly dark as in the vicinity of TH3 and TH4, it is not determined whether the high beam unnecessary state. If the addition / subtraction value is derived based only on whether the exposure amount is sufficiently bright or dark regardless of the integrated value, the integrated value does not increase until it becomes sufficiently dark in a scene from day to night (stays at 0 ) So it is necessary to wait for the integrated value to rise from 0 to 6000 after night. On the other hand, in the present embodiment, since the integrated value is increased to 5000 at the evening, it is sufficient to wait for the integrated value to increase from 5000 to 6000, and it is determined that the high beam is permitted in a shorter time. It is possible to

  On the other hand, in a scene where the brightness outside the vehicle is bright from a dark scene (integrated value = 10000), as shown in FIG. 5A, the absolute value of the addition / subtraction value becomes larger as the exposure amount becomes smaller (brighter). Since the absolute value of the addition / subtraction value is set larger as the integrated value is smaller than when the integrated value is smaller than the third integrated threshold, the integrated value quickly approaches the third integrated threshold as shown in FIG. 5 (c). It will be. However, when the integrated value is less than the third integrated threshold, for example, when the exposure amount is TH6 or more, the rate of decrease of the integrated value decreases and stagnates near the third integrated threshold. Then, when the outside of the vehicle becomes sufficiently bright (exposure amount <TH6), the integrated value becomes less than the second integrated threshold (for example, 4000), and the necessity determination unit 166 determines that the high beam is not required as shown in FIG. It will be done. Thus, even in the scene from night to day, since the integrated value has fallen to 5000 at the time of dawn, it is sufficient to wait for the integrated value to fall from 5000 to 4000, and simply regardless of the integrated value It becomes possible to determine that the high beam unnecessary state is achieved in a short time than when deriving the addition / subtraction value based only on whether the exposure amount is sufficiently bright or dark.

  In addition, the integrated value is smaller than the first integrated threshold and stagnated near the third integrated threshold larger than the second integrated threshold, and addition or subtraction is performed depending on whether the integrated value is equal to or greater than the third integrated threshold or smaller than the third integrated threshold. Due to the configuration of changing the value, the integrated value never exceeds the first integration threshold unless the dark state continues sufficiently, and the integrated value becomes less than the second integration threshold unless the bright state continues. Absent. Therefore, even when passing through shadows or tunnels in the daytime or passing through the city at night, it is possible to obtain a stable determination result without switching the determination as to whether the high beam unnecessary state is unnecessary. It becomes possible.

  In the present embodiment, since the hysteresis characteristic is provided, the necessity determination unit 166 determines that the integrated beam is in the high beam permission state when the integrated value is equal to or more than a predetermined first integrated threshold (for example, 6000). When it is less than the second integration threshold (4000) which is smaller than the 1 integration threshold, it is determined that the high beam unnecessary state. Therefore, when the integrated value becomes equal to or greater than the first integrated threshold and the necessity determination unit 166 once determines the high beam permission state, the high beam unnecessary state is determined until the integrated value becomes smaller than the second integrated threshold (determination is Will not be reversed). Similarly, when the integrated value becomes less than the second integrated threshold and the necessity determination unit 166 once determines that the high beam unnecessary state, the high beam permission state may also be determined until the integrated value becomes equal to or greater than the first integrated threshold. It disappears.

  As described above, if the necessity determination unit 166 determines that the high beam is not required, the following light distribution control S206 to S222 of the headlamp is omitted. Therefore, while the high beam unnecessary state continues, the preceding vehicle extraction process S208, the preceding vehicle recognition process S210, the oncoming vehicle extraction process S212, and the oncoming vehicle recognition process S214 are not executed. However, when the high beam unnecessary state is shifted to the high beam permitted state and the above-described processes S208 to S214 are suddenly started, the light distribution process of the headlamp may become unstable. Therefore, here, even in the high beam unnecessary state, if the integrated value is equal to or more than the predetermined value (for example, 5500), the possibility of shifting to the high beam permitted state becomes high. As preparation, preceding vehicle extraction processing S208, preceding vehicle recognition processing S210, oncoming vehicle extraction processing S212, and oncoming vehicle recognition processing S214 may be executed in advance.

  In the present embodiment, as described with reference to FIG. 5, addition / subtraction values corresponding to the exposure amount are integrated, and it is determined based on the integration value whether or not the high beam unnecessary state. In the following, (1) to (9) will be described processing of changing the integrated value and the like according to the environment outside the vehicle in order to reflect the actual environment outside the vehicle with respect to the basic technology.

  (1) At the start of driving (starting) of a car, it has not been decided which time of the day it is. Therefore, regardless of the value of the integrated value at that time, the integrated value is forcibly set to the third integrated threshold (for example, 5000), and the determination result is made the high beam unnecessary state (maintained if high beam unnecessary state, Switch to the high beam unnecessary state if it is the high beam enabled state). Thus, the necessity determination unit 166 can quickly determine whether the high beam unnecessary state is required.

  (2) In addition, it is possible to change the integrated value by using an auto light function provided in a car. Here, the auto light function is a function to automatically turn on the headlamp when the detection value of the illuminance sensor is less than a predetermined illuminance threshold (when the brightness outside the vehicle is not sufficient). Under the circumstances where such automatic lighting of the headlamp is not necessary, naturally the high beam is also unnecessary. Therefore, if the detected value of the auto light illuminance sensor is equal to or greater than the predetermined illuminance threshold, that is, if the automatic lighting of the headlamp is not necessary, the integrated value is equal to or greater than the third integrated threshold (eg, 5000). Is forcibly set to the third integration threshold (for example, 5000), and the determination result is set to the high beam unnecessary state. However, when the integrated value is less than the third integrated threshold, only the determination result is set as the high beam unnecessary state without changing the integrated value. Thus, the necessity determination unit 166 can quickly determine whether the high beam unnecessary state is required.

  However, the control system of the illuminance sensor and the headlamp may be independent of the control system of the external environment recognition device 120, and the detection value of the illuminance sensor may not be directly referred to. In this case, the auto light function is effective (the main switch is positioned at the auto light) instead of determining whether the detected value of the auto light illuminance sensor is equal to or greater than the predetermined illuminance threshold, and The integrated value and the determination result can be switched on the condition that the low beam is not lit.

  (3) Also, the integrated value can be changed using not only the main switch but also the dimmer switch. For example, if the dimmer switch is high beam enabled, this can also be recognized as an indication that the driver wishes to make the headlamp a high beam. However, some drivers may wish to keep the HBA and ADB enabled at all times, and keep the dimmer switch in a high beam enabled state. Therefore, it may not be possible to simply judge that the driver desires to turn on the high beam simply by the fact that the dimmer switch is high beam enabled.

  Therefore, it is determined that the driver's intention to make the headlamp a high beam indicates that the dimmer switch is not in a state where the high beam is steadily enabled but the position is switched from the high beam disable to the high beam enable. In other words, when the dimmer switch switches from high beam disable to high beam enable, the integrated value is less than a predetermined value (here, 5500) that is equal to or greater than the third integration threshold and less than the first integration threshold. The integrated value is switched to the predetermined value (5500). If the integrated value is equal to or more than a predetermined value, the integrated value is not switched (maintained). However, the determination result as to whether the high beam is not required is not changed. In the process of (3), since hardware information such as a dimmer switch is used, the illuminance outside the vehicle is detected by an illuminance sensor or the like without using the imaging device 110 without integrating addition / subtraction values according to the exposure amount. The process of changing the integrated value can be realized by directly detecting and integrating addition / subtraction values according to the illuminance.

  Here, the reason that the predetermined value is a value equal to or more than the third integration threshold (a value slightly smaller than the first integration threshold) is to quickly reflect that the driver desires to turn on the high beam. In addition, the reason why the predetermined value is less than the first integration threshold is because, even when the dimmer switch is switched to the high beam enable in order to validate the HBA and ADB in the daytime, it is not easily determined as the high beam enable state. is there. Thus, the necessity determination unit 166 can quickly determine whether the high beam unnecessary state is required.

  (4) Also, the integrated value can be changed when the environment outside the vehicle is a special environment. For example, in the early morning after cloudy weather, the state where the exposure amount is TH6 or more (see FIG. 5A) continues for a long time, and the switching to the high beam unnecessary state may not be performed quickly. Therefore, the condition (see FIG. 5A) in which the exposure amount is less than the predetermined exposure threshold (for example, TH5) is continuous for the predetermined time threshold (for example, the number of frames corresponding to 5 to 10 minutes or equivalent) or more. Then, if the integrated value is equal to or more than the third integrated threshold, the integrated value is switched to the third integrated threshold. If the integrated value is less than the third integrated threshold, the integrated value is not switched. At this time, in any case, the determination result is set to the high beam unnecessary state.

  (5) Further, the integrated value can be changed when the outside of the vehicle is another special environment. As described above, the exposure amount adjustment unit adjusts the exposure amount using the luminance distribution of a partial area (for example, a road surface area) of the generated image. Usually, the road surface often has a color between gray and black. For example, since the road surface when there is snow is white, the luminance is relatively high. In this case, the exposure amount 十分 TH1 may be satisfied and the high beam unnecessary state may be obtained even in a sufficiently dark state where the high beam state is to be permitted.

  Therefore, the luminance is acquired by referring to, for example, an area corresponding to the upper sky (for example, 100 pixels among them) other than the road surface area to be referred to adjust the exposure amount in the image. Then, when the brightness of the area is a predetermined dark condition, that is, the number of pixels for which the brightness is less than a predetermined value (for example, 10 in 256 steps) is a predetermined value (for example 90) or more, the integrated value at that time is the third. If less than the integration threshold value, 4 is added to the integration value, and if the integration value is equal to or more than the third integration threshold value, 2 is added to the integration value. However, such processing is performed only when a predetermined condition such as the case where the exposure amount is TH4 or more is satisfied. Thus, the problem that the high beam unnecessary state continues unnecessarily for a long time even on a snowy road surface at night can be solved.

  (6) Further, it is possible to temporarily stop the integration processing of the integration value according to the environment outside the vehicle. For example, when an obstacle such as a leading vehicle or an oncoming vehicle is located immediately in front of the host vehicle 1, it is affected by the tail lamp or the brake lamp of the leading vehicle or the head lamp of the host vehicle 1 The amount of exposure may be small. Therefore, using the three-dimensional position derived in image processing S200, if the depth distance of the last three-dimensional object is less than a predetermined distance threshold (for example, 10 m), the value of the integrated value may not be updated (do not integrate) .

  If this process is adopted, the integrated value will not change under the situation where it is from day to night in a traffic jam, but if there is a preceding vehicle or obstacle in the first place, high beam is not necessary. It does not occur. In this way, it is possible to avoid an unnecessary high beam unnecessary state due to the influence of the reflection of the tail lamp and the brake lamp of the preceding vehicle or the headlamp of the own vehicle 1 or the like.

  (7) Further, integration processing of other integration values can be temporarily stopped according to the environment outside the vehicle. For example, at night and at an intersection, when the host vehicle is stopped at a leading position, the headlamp of the oncoming vehicle may be bright. In this case, the amount of exposure may be reduced due to the influence of the headlamp of the oncoming vehicle. Therefore, when it is determined that the host vehicle 1 is in the stopped state and the oncoming vehicle is present ahead, the integrated value may not be updated (not integrated). In this way, it is possible to avoid an unnecessary high beam unnecessary state due to the influence of the headlamp of the oncoming vehicle.

  (8) Also, integration processing of other integration values can be temporarily stopped. For example, even in the daytime, if the optical axis of the imaging device 110 is blocked, the exposure amount may be large regardless of the environment outside the vehicle, and the high beam permission state may be entered. In addition, the imaging device 110 is susceptible to the weather and the like, and may not recognize the leading vehicle and the like due to rainfall, fog, back light, and the like. Therefore, when it is determined that such control by the imaging device 110 is temporarily prohibited (HALT determination), the value of the integrated value may not be updated (not integrated). Thus, even in the case where the imaging device 110 can not normally recognize the environment outside the vehicle, it is possible to avoid the unnecessary high beam permission state.

  (9) Further, the threshold value of the exposure amount may be changed according to the environment outside the vehicle. For example, while the host vehicle 1 is traveling in a city area, the amount of exposure may be reduced due to the influence of streetlights. Therefore, while the current scene is determined to be a city area in the traveling scene determination process S220 of the previous frame, each threshold (TH1 to TH6) in FIG. 5A is reduced by a predetermined ratio (for example, 10%) It may be In this manner, it is possible to avoid unnecessary high beam unnecessary state due to the influence of the streetlight even in the urban area.

(Detection range setting processing S206)
When the necessity determination unit 166 determines that the high beam is permitted (NO in S204), the detection range setting unit 168 adds to the normal range for detecting a three-dimensional object in the acquired image, the preceding vehicle (tail lamp), the oncoming vehicle (Headlamp) A processing resource for detecting each of the streetlights is applied to determine a range (detection range) in which a three-dimensional object is detected in more detail and with high accuracy. In this way, in addition to the detection range of the normal three-dimensional object in the image, by limiting the detection range in detail and with high precision, the processing time for extracting the leading vehicle and the oncoming vehicle can be shortened and It is possible to prevent erroneous detection in the area where the oncoming vehicle does not originally exist. Hereinafter, the detection range for detecting the three-dimensional object in detail and with high accuracy will be described in detail.

  FIG. 6 is an explanatory view for explaining a detection range. The detection range setting unit 168 sets a leading vehicle detection range 220a indicated by a broken line rectangle at a predetermined position shown in FIG. 6 of the image area 216 for each of the preceding vehicle, the oncoming vehicle, and the streetlight to be detected. An oncoming vehicle detection range 220b indicated by a dashed-dotted line rectangle and a street lamp detection range 220c indicated by a double-dotted line rectangle are set. As can be understood with reference to FIG. 6, the preceding vehicle detection range 220a is included in the oncoming vehicle detection range 220b, and the both and the street lamp detection range 220c are exclusive.

  The preceding vehicle detection range 220a, the oncoming vehicle detection range 220b, and the streetlight detection range 220c can be offset according to the environment outside the vehicle and the traveling route. For example, if the road is curved or has a slope, the preceding vehicle detection range 220a, the oncoming vehicle detection range 220b, and the streetlight detection range 220c are offset according to the degree. For example, if the traveling route is a left curve, the detection range setting unit 168 offsets each of the preceding vehicle detection range 220a, the oncoming vehicle detection range 220b, and the street lamp detection range 220c to the left by an amount corresponding to the traveling route. In this way, it is possible to set as the detection range the position where the possibility of the preceding vehicle, the oncoming vehicle, and the streetlight is the highest.

(Preceding vehicle extraction processing S208)
Subsequently, the preceding vehicle extraction unit 170 extracts a tail lamp from the preceding vehicle detection range 220a according to the luminance and color information and the three-dimensional position. However, the tail lamp of the leading vehicle is different in light amount from the headlamp or the street lamp of the oncoming vehicle described later. Then, in the imaging device 110, when the tail lamp is imaged at an obtainable exposure time, the brightness of the headlamp or the street lamp is saturated, and conversely, when the headlamp or the street lamp is imaged at an obtainable exposure time, the tail lamp can not be detected. I will. Therefore, the imaging device 110 generates an image with at least two long and short exposure times, with different frames.

  FIG. 7 is an explanatory diagram for explaining a luminance image 212 having different exposure times. For example, FIG. 7 (a) shows a long exposure time, and FIG. 7 (b) shows a short exposure time. Therefore, although the luminance image 212 of FIG. 7A may be used to saturate the luminance of the headlamp or the streetlight, the tail lamp can be properly extracted, and using the luminance image 212 of FIG. Although there is a possibility that it can not be extracted, it can extract headlamps and streetlights properly.

  Then, in the preceding vehicle detection range 220a, the preceding vehicle extraction unit 170 has color information (RGB or YUV) in a predetermined color range indicating red, and a three-dimensional position in a predetermined distance range (for example, 1.5 pixels) Group the pixels in). However, the leading vehicle extraction unit 170 groups the pixels into a rectangular shape including a horizontal line and a vertical line including all the pixels that satisfy this condition. The grouped tail lamp candidates have basic feature quantities such as top, bottom, left, and right coordinates of the group, the number of pixels in the group, the maximum luminance value in the group, the minimum luminance value, and the average depth distance of the group (average parallax).

  Here, in the preceding vehicle extraction unit 170, the difference (size) of the upper and lower left and right coordinates of the group is a predetermined value (for example, 2 pixels) or less, and the difference (size) of the upper and lower left and right coordinates of the group is a predetermined value (determined by the depth distance) or more If any of the exclusion conditions such that the number of pixels in the group is equal to or less than a predetermined value (for example, 2) is satisfied, it is excluded as a tail lamp candidate.

  Moreover, in addition to a tail lamp, a red reflective plate may be included in the preceding vehicle detection range 220a, and it is difficult to distinguish both by only color information. However, since the tail lamp emits light while the red reflector uses light reflection, the tail lamp has higher luminance than the red reflector at the same depth distance. Here, the tail lamp and the red reflector are distinguished by using such characteristics.

  FIG. 8 is an explanatory view showing the relationship between the tail lamp and the red reflector. In both of the tail lamp indicated by the solid line in FIG. 8 and the red reflector indicated by the broken line, the luminance decreases as the depth distance increases, but the tail lamp has higher luminance than the red reflector at any distance. Although the relationship between the distance and the luminance is shown as linear in FIG. 8 for convenience of explanation, in many cases the relationship is not linear in practice. If the preceding vehicle extraction unit 170 determines that the relationship corresponds to the red reflector in FIG. 8 based on the depth distance and the luminance of the tail lamp candidate, it excludes it from the tail lamp candidate.

  As described above, in the present embodiment, tail lamps are extracted in the preceding vehicle detection range 220a. However, the depth distance of the preceding vehicle detection range 220a is very long, such as 0 m to several hundreds m, and the exposure time must be lengthened to recognize a distant tail lamp. Then, the tail lamp located in the relatively long depth distance A region can identify the relationship between the depth distance and the brightness, for example, it can be distinguished from the red reflector, but as shown in FIG. In the tail lamp located at, the brightness is saturated at the maximum value, and the relationship between the depth distance and the brightness can not be specified. The luminance is saturated at the light emitting portion of the tail lamp, and there is a red color around it, so the tail lamp candidate itself can be extracted.

  Such a tail lamp is lower in luminance than a headlamp of an oncoming vehicle, so the difference in luminance between the tail lamp and the red reflector is small in the first place. Here, as shown in FIG. 8, if the depth distance in the B region becomes short, the difference between the tail lamp and the red reflector shrinks, and finally, the brightness of both the tail lamp and the red reflector is saturated. Thus, where the depth distance is short, the tail lamp and the red reflector can not be distinguished.

  In addition to being used as a display for prompting the driver while driving, the red reflective plate is also used as a reflector of a vehicle. In the present embodiment, it is an object of the present invention not to give glare by a high beam to a leading vehicle during traveling to the last, and it is desirable to hit a high beam to a parked vehicle, for example.

  For example, since parallel parking on a street is common in residential areas in the United States, false recognition of a red reflector and a tail lamp prevents high beam lighting at high street parking locations, or hunting of high beam and low beam occurs. Etc. problems may occur. In order to avoid such a situation, it is conceivable to use the shape and speed (≠ 0) of the tail lamp candidate, but the former often can not exclude the red reflector, and the latter is a leading vehicle that is stationary. High accuracy identification is difficult because

  Therefore, it is desirable to distinguish the tail lamp and the red reflector based on the relationship between the depth distance and the luminance. Here, in addition to the luminance image 212 of FIG. 7A having a long exposure time for extracting the tail lamp described above, the luminance of FIG. 7B for extracting headlamps and streetlights generated in different frames. The image 212 is used.

  FIG. 9 is another explanatory view showing the relationship between the tail lamp and the red reflector. The leading vehicle extraction unit 170 first uses the luminance image (first image) of FIG. 7 (a). In the luminance image 212 of FIG. 7A, the relationship between the luminance of the block and the depth distance is as shown in FIG. 9A, and the area A is distinguished based on the luminance and the depth distance of the block. And the tail lights of the preceding vehicle can be extracted.

  Then, if the depth distance of the block is less than a predetermined distance threshold (for example, 150 m), the leading vehicle extraction unit 170 selects the luminance image of FIG. 7B (b) whose exposure time is shorter than the luminance image 212 of FIG. The second image is used. In the luminance image 212 shown in FIG. 7B, the relationship between the luminance of the block and the depth distance is as shown in FIG. 9B. It is possible to distinguish between the two based on the depth distance and extract the tail lamp of the preceding vehicle.

  The preceding vehicle detection range 220a in the case of extracting the tail lamp from the luminance image 212 of FIG. 7 (b) having a short exposure time is the precedent in the case of extracting the tail lamp from the luminance image 212 of FIG. 7 (a) having a long exposure time. The vehicle detection range 220a is used as it is, or slightly expanded (for example, 1.1 times) and used. That is, the leading vehicle detection range 220a extracts the tail lamp from the range based on the leading vehicle detection range 220a already derived in the luminance image 212 of FIG. 7B having a short exposure time. Here, since the frame rate is sufficiently small and the relative speed between the leading vehicle and the host vehicle 1 is low, the position of the leading vehicle does not move much between the frames. Therefore, there is no problem even if the preceding vehicle detection range 220a is used in this way for the luminance image 212 in a different frame. In this way, it is possible to avoid the re-derivation process of the preceding vehicle detection range 220a and reduce the processing load.

  Hereinafter, specific processing will be described in which the leading vehicle extraction unit 170 extracts the tail lamp from the luminance image 212 of FIG. 7B having a short exposure time. First, the leading vehicle extraction unit 170 uses the maximum value of the red component (R of RGB) of the tail lamp candidate as the feature quantity of the tail lamp. However, it is assumed that only the pixels satisfying the condition of R ≧ G and R ≧ B are to obtain the maximum value.

  Further, since the brightness of the tail lamp varies depending on the type of vehicle and the environment, a value accumulated by a scoring method between a plurality of frames is used. Specifically, a tail lamp point indicating "tail lamp likeness" with an initial value of 0 and a non-tail lamp point indicating "non-tail lamp likeness" are accumulated for each frame.

  FIG. 10 is an explanatory diagram for explaining points accumulated by the preceding vehicle extraction unit 170. As shown in FIG. In FIG. 10, four regions ((a) to (d)) are set based on the relationship between the depth distance between the tail lamp and the red reflector in FIG. 9 (b) and the luminance. Here, it is determined that the area of (b) indicated by hatching corresponding to the relationship between the depth distance of the tail lamp and the brightness is likely to be a tail lamp, and indicated by cross hatching corresponding to the relationship between the depth distance of the red reflector and the brightness. It is determined that the area of (d) is not likely to be a tail lamp, and it is determined that neither of the areas of (c) is in between, and the luminance is higher than that of the area of (b). It is determined that the area of (a) is likely to be a tail lamp. In FIG. 10, for convenience of explanation, the region is shown linearly with respect to the depth distance, but the depth distance is divided into a plurality of stages, and discretized (stepwise) for each depth distance. It is also good.

  When the preceding vehicle extraction unit 170 determines that the relationship between the depth distance of the block and the luminance is included in the area of (a), it adds five points to the tail lamp point and determines that it is included in the area of (b) If one point is added to the tail lamp point and it is judged that it is included in the area of (c), no processing is performed, and if it is judged that it is included in the area of (d), one point is added to the non tail lamp point. The tail lamp point and the non-tail lamp point obtained in this manner are used in preceding vehicle recognition processing S210 described later. Specifically, after the preceding vehicle recognition unit 172 determines that the preceding vehicle candidate is the preceding vehicle, the tail lamp point and the non-tail lamp point are used to perform the correction. The correction will be described later in detail in preceding vehicle recognition processing S210.

  By the way, although the tail lamp and the head lamp to be described later (which will be described for the tail lamp to be briefly described below) are extracted based on the feature amount (brightness and color information) in the image, the feature amount is near the threshold If it fluctuates, the tail lamp extraction itself may become unstable. For example, the situation where the tail lamp is determined to be a tail lamp in an arbitrary frame but is determined not to be a tail lamp in another frame is repeated for a preceding vehicle actually present. Thus, when the tail lamp is extracted unstably, hunting occurs in which light distribution control of the headlamp of the vehicle 1 repeats high beam and low beam.

  In addition, a three-dimensional object that does not emit light, such as a reflector plate located in the vicinity of the traveling path, will have different image feature amounts depending on how the headlights of the host vehicle 1 hit, and tail lamps (preceding vehicles) and headlamps (oncoming vehicles It becomes easy to misunderstand as). For example, when the host vehicle 1 is set to a high beam, the reflection thereof causes a false recognition as a tail lamp (preceding vehicle), and the high beam switches to a low beam so that the leading vehicle is not irradiated. However, switching to the low beam eliminates the reflection of the high beam, and if the tail lamp is not recognized, hunting may occur such as becoming the high beam again.

  For such hunting, it is possible to execute countermeasure processing in the light distribution control after recognition of the preceding vehicle and the oncoming vehicle, but when the recognition itself that is the source of the light distribution control is unstable, the light distribution control itself The result is that it loses its robustness.

  So, in this embodiment, at the extraction time of a tail lamp etc., the process which incorporated the hysteresis characteristic in recognition processing is performed. Specifically, the threshold for comparing the feature amounts is made different depending on whether or not the object to be extracted is hit by the high beam.

  For example, in the area where the high beam is striking, the threshold value is set higher (strictly) than the area where the high beam is not striking to prevent false detection of the tail lamp or the like. Alternatively, in the area where the high beam is not hit, the threshold is made lower (more relaxed) than the area where the high beam is hit, so that it is easy to extract the tail lamp or the like. Thus, tail lamps and the like can be properly extracted to prevent hunting.

  As described above, in order to change the threshold depending on whether or not the high beam is hit, it is necessary to first determine the region where the high beam is hit.

  FIG. 11 is a functional block diagram for explaining light distribution control of a headlamp. In the present embodiment, tail lamps and headlamps are extracted by the preceding vehicle extraction unit 170 and the oncoming vehicle extraction unit 174. Then, the preceding vehicle recognition unit 172 and the oncoming vehicle recognition unit 176 recognize the preceding vehicle and the oncoming vehicle based on the extracted tail lamp and headlamp. Then, based on such information, the light distribution control unit 184 determines the light distribution of the headlamp, that is, the irradiation range (or the irradiation presence or absence) of the high beam, and CAN (Controller Area) indicated by the broken line by the illumination mechanism 124. Network) Irradiate the high beam based on the irradiation range of the high beam acquired through communication.

  Here, as shown by the broken line in FIG. 11, the leading vehicle extraction unit 170 acquires the resultant information on which irradiation area the illumination mechanism 124 has irradiated the high beam through CAN communication, as described above. Suppose that the threshold is changed. In this case, the CAN communication may be delayed as follows.

  That is, when the preceding vehicle is recognized, the light distribution control unit 184 sends an instruction to switch the high beam of the area to the low beam. The illumination mechanism 124 switches the high beam to the low beam according to the instruction. However, due to the delay of CAN communication from the light distribution control unit 184 to the lighting mechanism 124 and from the lighting mechanism 124 to the preceding vehicle extraction unit 170, when the light distribution control unit 184 instructs switching to the low beam, The preceding vehicle extraction unit 170 and the oncoming vehicle extraction unit 174 can obtain only past information that the illumination mechanism 124 is emitting a high beam.

  Therefore, the leading vehicle extraction unit 170 extracts the tail lamp in the next frame while the threshold value of the area is high, based on the past information that the headlamp is the high beam, and the tail lamp extracted next time There is a risk of losing sight of the frame.

  Therefore, as shown by the broken line in FIG. 11, the leading vehicle extraction unit 170 acquires the resultant information on which irradiation area the illumination mechanism 124 has irradiated with the high beam through CAN communication, as indicated by the one-dot chain line. Also, information on which irradiation range the light distribution control unit 184 instructed to irradiate the high beam is acquired. Then, the leading vehicle extraction unit 170 changes the threshold by using both pieces of information.

  Specifically, in the preceding vehicle extraction unit 170, the resulting information (the information executed by the illumination mechanism 124) to which illumination range the illumination mechanism 124 has emitted the high beam and the illumination range to which the light distribution control unit 184 belongs The area where the low beam is irradiated (the high beam is not irradiated) if any one of the information indicating that the high beam is to be irradiated (the information instructed by the light distribution control unit 184) indicates the low beam (not the high beam). Recognize as For example, in the case of HBA, if any indicates a low beam, it is determined that a low beam is being irradiated, and in the case of ADB, if any indicates a low beam for each angle range (area), a low beam It is determined that the

  With such a configuration, in a scene where the high beam is switched to the low beam, light distribution control is performed even if the resultant information on which irradiation area the illumination mechanism 124 has irradiated the high beam due to the CAN communication delay still indicates the high beam. Since the information indicating which irradiation area the part 184 instructed to irradiate the high beam is the low beam (not the high beam), a low (loose) value is used as the threshold value to stabilize the preceding vehicle (oncoming vehicle) It is possible to continue detecting. In this way, it is possible to reduce the possibility of hitting the high beam on the preceding vehicle (oncoming vehicle).

  Further, in the case of using ADB, information on which irradiation area the illumination mechanism 124 has irradiated with the high beam is obtained from cut line angle information indicating the irradiation angle of the high beam transmitted from the illumination mechanism 124. However, such cut line angle information alone can not simply identify which range of the luminance image 212 the high beam hits.

  FIG. 12 is a top view for explaining the relationship between the cut line angle and the angle of view. For example, in FIG. 12, it is assumed that A and B exist at the same arbitrary angle of view. Referring to the cut line angle, A and B are equal in horizontal position in the luminance image 212 generated by the imaging device 110, but A is hit by a high beam as shown by hatching, and B is hit by a high beam. Not. This is based on the difference in the position of the imaging device 110 and the headlamp (illumination mechanism 124) like FIG. Therefore, as long as the depth distance of the block can be grasped, the leading vehicle extraction unit 170 determines which of the luminance image 212 is geometrically calculated as shown in FIG. 12 based on the angle of view and depth distance of the block and the cutline angle. It can be determined whether the range (block) is hit by the high beam.

  However, if it is calculated based on the angle of view and depth distance of the block in FIG. 12 and the cut line angle whether or not the high beam is hit, the processing load such as trigonometric function or division increases. Therefore, in the present embodiment, the depth distance is divided into a plurality of distance ranges, and the processing load is reduced by using a table for each distance range.

  FIG. 13 is an explanatory view for explaining a table of cut line angles. Here, the depth distance is divided into five steps (0 m to 10 m, 10 m to 30 m, 30 m to 70 m, 70 m to 200 m, 200 m), and one cut line angle (resolution 0.1 °) is associated in each distance range . Here, the cut line angle at a depth distance of 10 m and the cut line angle at a depth distance of 30 m, both of which are shown by broken lines in FIG. 13A, are compared. Then, it can be understood that the cut line angle is wider as the depth distance is shorter.

  Therefore, as shown in FIG. 13B, as the cut line angle on the image with a distance range of 10 m to 30 m, a 10 m cut line angle (here 24.9 °) at which the depth distance shown by the broken line is shortest is used. To be. Therefore, the cut line angle is made wider than the true cut line angle at all depth distances including the distance range of 10 m to 30 m, and the possibility of applying the high beam to the leading vehicle can be reduced. It becomes.

  As shown in FIG. 13 (b), the five-step distance range is changed not linearly but nonlinearly. This is because the difference between the cut line angle on the image and the true cut line angle rapidly increases as the distance decreases. By changing the distance range nonlinearly in this manner, it is possible to set an appropriate cut line angle according to the depth distance.

  In this manner, the preceding vehicle extraction unit 170 determines whether or not the high beam strikes the block based on the angle of view and depth distance of the block of the image and the cut line angle of the high beam, and The threshold for determining whether the lamp is a tail lamp is changed, the tail lamp of the leading vehicle is extracted, and based on that, the leading vehicle recognition unit 172 recognizes the leading vehicle as described later.

  However, since the leading vehicle has a certain width in the horizontal direction, for example, a situation occurs in which the high beam strikes on the horizontal half of the leading vehicle and does not hit the other half. In this case, with respect to the preceding vehicle candidate, the ratio at which the high beam is hit is compared with the ratio at which the high beam is not hit, and it is determined that the ratio is larger. Therefore, if the ratio at which the high beam hits is large, it is assumed that the high beam is hitting the entire preceding vehicle candidate, and if the ratio at which the high beam is not hitting is large, the high beam does not hit the entire preceding vehicle candidate.

  Although there is an idea that it is judged that the suspect is not hit, as mentioned above, the cut line angle is wider than the true cut line angle by using the table, and it is shifted in the direction where the high beam is not hit. Here, the high beam success or failure is simply determined according to the ratio so that the high beam does not have to be hit excessively.

  Thus, it is possible to prevent hunting of the high beam and the low beam and to reduce the possibility of the high beam being applied to the preceding vehicle.

(Preceding vehicle recognition processing S210)
Subsequently, the preceding vehicle recognition unit 172 groups the tail lamps extracted by the preceding vehicle extraction unit 170, and recognizes the preceding vehicle in the preceding vehicle detection range 220a.

  Specifically, the distance between the tail lamps (groups) on the image is within the distance range included in the same car (longer than the predetermined distance range of the preceding vehicle extraction processing S208) or the difference of the average depth distance (average parallax) When all the conditions of whether the ratio of the maximum luminance value is in the predetermined range or the distance range included in the same automobile is satisfied, the tail lamps are grouped to be the leading vehicle candidate.

  The leading vehicle candidates grouped in this way inherit the basic feature quantities of the tail lamp before grouping. For example, the top, bottom, left, and right coordinates of the leading vehicle candidate inherit the top, bottom, left, and right coordinates of the tail lamp corresponding to the outside of the leading vehicle candidate, and the top and bottom luminance values of the leading vehicle candidate are the maximum and minimum luminance values of the tail lamp. The average depth distance of the leading vehicle candidate inherits the shorter one of the average depth distance of the tail lamp (the larger one of the average parallax). Then, the number of tail lamps included in the preceding vehicle candidate is also counted.

  Further, the preceding vehicle recognition unit 172 determines whether the presence of the preceding vehicle has been confirmed at the same three-dimensional position in the past frame, and counts the number of times of the confirmation. The number of times of existence affects the reliability of the feeling of the preceding vehicle. Then, the preceding vehicle recognition unit 172 determines whether the condition indicating that there is reliability as the preceding vehicle is satisfied, and whether the condition that indicates that there is no reliability as the preceding vehicle is satisfied, according to the result. The preceding vehicle candidate is specified as a preceding vehicle or excluded from the preceding vehicle candidates. Since such preceding vehicle recognition processing S210 can employ various existing technologies such as Japanese Patent Application No. 2014-232431, the detailed description thereof will be omitted here.

  Subsequently, the preceding vehicle recognition unit 172 corrects the identified preceding vehicle based on the tail lamp point accumulated by the preceding vehicle extraction unit 170 and the non-tail lamp point. Specifically, if the non-tail lamp point is equal to or more than a predetermined value (for example, 3), the preceding vehicle recognition unit 172 excludes the specified preceding vehicle from the preceding vehicle candidates. Thereafter, if the tail lamp point is equal to or more than the predetermined value, the preceding vehicle recognition unit 172 newly specifies the excluded preceding vehicle as the preceding vehicle. Thus, the tail lamp point indicating "tail lamp likeness" is reflected more strongly than the non-tail lamp point indicating "not like tail lamp". However, the correction that is excluded from the preceding vehicle targets only the leading vehicle candidates existing in the area where the high beam is considered to be hit. This is to stabilize the recognition of the preceding vehicle.

  Here, the predetermined value to be compared with the tail lamp point differs depending on the number of tail lamps. For example, 8 is adopted if there is one tail lamp, 5 etc. is adopted if it is 2 or more. This is to make it difficult for a single reflector present on the road shoulder to be misrecognized as a tail lamp.

(Oncoming vehicle extraction processing S212)
Subsequently, the oncoming vehicle extraction unit 174 extracts a headlamp from the oncoming vehicle detection range 220b in accordance with the luminance and color information and the three-dimensional position. However, as described above, since the headlamp of the oncoming vehicle has a light amount different from that of the tail lamp of the preceding vehicle, the luminance image 212 with a short exposure time shown in FIG. 7B is used.

  Next, the oncoming vehicle extraction unit 174 determines that the brightness is equal to or higher than a predetermined brightness threshold (for example, 5 to 10 in 256 steps) in the oncoming vehicle detection range 220b, and the three-dimensional position is a predetermined distance range (for example, 1.5) Groups pixels together). However, the oncoming vehicle extraction unit 174 groups the pixels into a rectangular shape including a horizontal line and a vertical line including all the pixels satisfying the condition. The grouped headlamp candidates have basic feature quantities such as upper and lower left and right coordinates of the group, the number of pixels in the group, the maximum luminance value in the group, the minimum luminance value, and the average depth distance (average parallax) of the group.

  Here, in the oncoming vehicle extraction unit 174, the difference (size) of the upper and lower left and right coordinates of the group is a predetermined value (for example, 2 pixels) or less, the difference (size) of the upper and lower left and right coordinates of the group is a predetermined value (determined by the depth distance) or more If any of the exclusion conditions such that the number of pixels in the group is equal to or less than a predetermined value (for example, 2) is satisfied, it may be excluded as a headlamp candidate. Here, the predetermined value to be compared with the luminance is adjusted based on the predetermined value in the previous frame. Since such oncoming vehicle extraction processing S212 can employ various existing technologies such as Japanese Patent Application No. 2014-232430, the detailed description thereof will be omitted here.

  Further, in the oncoming vehicle extraction process S212, the technology of the preceding vehicle extraction process S208 described above can be used as it is. That is, the oncoming vehicle extraction unit 174 obtains information as to which irradiation range the illumination mechanism 124 has irradiated with the high beam, and information as to which irradiation range the light distribution control unit 184 has instructed to irradiate the high beam. If any one of them indicates a low beam (not a high beam), it is recognized as an area to be irradiated with a low beam (not irradiated with a high beam). For example, in the case of HBA, if any one indicates a low beam, it is determined that a low beam is irradiated, and in the case of ADB, if any indicates a low beam for each angle (region), a low beam Judge as irradiating.

  With such a configuration, in a scene where the high beam is switched to the low beam, light distribution control is performed even if the resultant information on which irradiation area the illumination mechanism 124 has irradiated the high beam due to the CAN communication delay still indicates the high beam. Since the information indicating which irradiation area the part 184 instructed to irradiate the high beam is the low beam (not the high beam), a low (loose) value is used as the threshold value to stably detect the oncoming vehicle. It is possible to continue. In this way, it is possible to reduce the possibility of hitting the high beam to the oncoming vehicle.

  Further, the oncoming vehicle extraction unit 174 can specify which range of the luminance image 212 is hit by calculation based on the angle of view and depth of the block and the cut line angle. Thus, it is possible to prevent hunting of the high beam and the low beam and to reduce the possibility of the high beam being applied to the oncoming vehicle.

(Oncoming vehicle recognition processing S214)
Subsequently, the oncoming vehicle recognition unit 176 groups headlamps extracted by the oncoming vehicle extraction unit 174 and recognizes oncoming vehicles in the oncoming vehicle detection range 220b.

  Specifically, the distance between the headlamps (groups) on the image is within the distance range included in the same car (longer than the predetermined distance range of the oncoming vehicle extraction processing S212), or the difference in average depth distance is the same. When all the conditions of whether it is in the distance range included in the car or whether the ratio of the maximum luminance value is in the predetermined range are satisfied, the headlamps are grouped into an oncoming vehicle candidate.

  The oncoming vehicle candidates grouped in this way inherit the basic feature quantities of the headlamps before grouping. For example, the top, bottom, left, and right coordinates of the oncoming vehicle candidate inherit the top, bottom, left, and right coordinates of the headlamp corresponding to the outside of the oncoming vehicle candidate. The maximum luminance value and the minimum luminance value of the oncoming vehicle candidate are the maximum luminance value and the minimum of the headlamps. The average parallax (depth distance) of the oncoming vehicle candidate inherits the smaller one of the average depth distances of the headlamps (the larger one of the average parallaxes). Then, the number of headlamps included in the oncoming vehicle candidate is also counted.

  Further, the oncoming vehicle recognition unit 176 determines whether the presence of the oncoming vehicle can be confirmed at the same three-dimensional position in the past frame, and counts the number of times of the confirmed presence. The number of times of occurrence affects the reliability of the oncoming vehicle likeness. Then, the oncoming vehicle recognition unit 176 determines whether the condition indicating the reliability as the oncoming vehicle is satisfied or the condition indicating the absence of the reliability as the oncoming vehicle is satisfied, according to the result. The oncoming vehicle candidate is specified as an oncoming vehicle or excluded from the oncoming vehicle candidate. Since such oncoming vehicle recognition processing S214 can employ various existing technologies, such as Japanese Patent Application No. 2014-231301, for example, the detailed description thereof is omitted here.

(Street lamp extraction process S216)
Subsequently, the street lamp extraction unit 178 extracts a street lamp from the street lamp detection range 220 c in accordance with the luminance and color information and the three-dimensional position by the same process as the oncoming vehicle extraction process S 212.

(Streetlight recognition processing S218)
The street lamp recognition unit 180 recognizes the street lamp extracted by the street lamp extraction unit 178. Here, the streetlight is not a three-dimensional object that should not be irradiated with a high beam, but it is used in a traveling scene determination process S220 at a later stage.

(Traveling scene determination processing S220)
The traveling scene determination unit 182 determines whether or not the traveling scene is capable of emitting a high beam. For example, when the vehicle speed is equal to or less than a predetermined value (for example, 20 km / h), the traveling scene determination unit 182 determines that the high beam is unnecessary. In addition, the traveling scene determination unit 182 determines that the high beam is unnecessary when the host vehicle 1 turns left and right. Furthermore, when the number of streetlights is equal to or greater than a predetermined number (for example, 3), the traveling scene determination unit 182 determines that the outside of the vehicle is sufficiently bright and the high beam is unnecessary. As such traveling scene determination processing S220 can adopt various existing technologies such as Japanese Patent Application No. 2014-232408 and Japanese Patent Application No. 2014-232409, the detailed description thereof is omitted here.

(Light distribution control process S222)
Finally, the light distribution control unit 184 executes light distribution control of the headlamps of the host vehicle 1 based on the preceding vehicle, the oncoming vehicle, and the traveling scene.

  FIG. 14 is an explanatory view showing the operation of the light distribution control unit 184. As shown in FIG. As shown in FIG. 14, when the traveling scene determination unit 182 determines that the high beam is unnecessary, whether it is an HBA or an ADB, and a solid object that should not be irradiated (preceding vehicle, oncoming vehicle) Irradiate the high beam regardless of the number of In addition, when the traveling scene determination unit 182 determines that the scene permits the use of the high beam, in the case of the HBA, if there are one or more three-dimensional objects that should not be illuminated, the high beam is not illuminated and should be illuminated. If there are no three-dimensional objects, high beam irradiation is performed. Also, in the case of ADB, if there is one or more three-dimensional objects that should not be irradiated, high-beam irradiation is performed in part of the range while avoiding irradiation to the three-dimensional objects, and three-dimensional objects that should not be irradiated. For example, high beam irradiation is performed over the entire range.

  FIG. 15 is an explanatory view for explaining light distribution control of ADB. Here, in ADB, if there are one or more three-dimensional objects that should not be irradiated, the maximum width W in the horizontal direction for all three-dimensional objects is calculated as shown in FIG. I do. However, if it is possible to split the high beam of ADB and irradiate the central part as well, as shown in FIG. 15 (b), the maximum widths W1 and W2 in the horizontal direction of the preceding vehicle and the oncoming vehicle which should not be irradiated Calculate and irradiate the central and outer high beams.

  In the case of ADB, when the oncoming vehicle becomes very close to the host vehicle 1, the oncoming vehicle may be out of the oncoming vehicle detection range 220b. Therefore, when the oncoming vehicle approaches a certain depth distance (for example, 50 m), irradiation in a direction in which passing is expected is not performed for a certain period (for example, 1 sec).

  With such an outside environment recognition device 120, it is possible to appropriately execute the light distribution control of the headlamp.

  In addition, a program that causes a computer to function as the external environment recognition device 120, and a computer readable storage medium such as a flexible disk, a magneto-optical disk, a ROM, a CD, a DVD, a BD, etc., storing the program are also provided. Here, the program refers to data processing means described in any language or description method.

  Although the preferred embodiments of the present invention have been described above with reference to the accompanying drawings, it goes without saying that the present invention is not limited to such embodiments. It is obvious that those skilled in the art can conceive of various changes or modifications within the scope of the claims, and it is naturally understood that they are also within the technical scope of the present invention. Be done.

  For example, in the above-described embodiment, the light distribution control by the HBA or ADB is described each time, but all the processes can be applied to any light distribution control.

  In the embodiment described above, various threshold values are set, but such values can be changed as appropriate, and can also be set according to experience values or experimental values.

  Note that each process of the outside environment recognition process of the present specification does not necessarily have to be processed in chronological order according to the order described as the flowchart, and may include processes in parallel or by a subroutine.

  The present invention can be used for an outside environment recognition device that performs light distribution control of a headlamp according to the outside environment of the vehicle.

120 Vehicle outside environment recognition device 124 lighting mechanism 164 integrated value derivation unit 166 necessity determination unit 168 detection range setting unit 170 leading vehicle extraction unit 172 preceding vehicle recognition unit 174 oncoming vehicle extraction unit 176 oncoming vehicle recognition unit 178 streetlight extraction unit 180 streetlight recognition Unit 182 Driving scene determination unit 184 Light distribution control unit 220a Preceding vehicle detection range 220b Oncoming vehicle detection range 220c Street light detection range

Claims (2)

The computer is
An illuminance deriving unit that derives the illuminance outside the vehicle;
An integrated value deriving unit that derives an addition / subtraction value so that the value becomes larger as the illuminance is lower than when the illuminance is high, and derives an integrated value obtained by integrating the addition / subtraction value;
It is determined that the high beam is available when the integrated value is equal to or more than a predetermined first integrated threshold, and it is necessary to determine that the high beam is unnecessary when the integrated value is less than a second integrated threshold smaller than the first integrated threshold. Judgment unit,
Act as
The outside environment recognition system according to claim 1, wherein the integrated value deriving unit switches the integrated value to a predetermined value in response to switching of the dimmer switch from high beam disabled to high beam enabled. The predetermined value is a value greater than or equal to the second integration threshold and less than the first integration threshold,
The integrated value deriving unit switches the integrated value to a predetermined value if the integrated value is less than the predetermined value in response to the dimmer switch switching from high beam not available to high beam enabled, and the integrated value is the same as the integrated value. The outside environment recognition device according to claim 1, wherein the integrated value is not switched if the value is equal to or more than a predetermined value.

Images