JP6329438B2 - Outside environment recognition device - Google Patents

Description

  The present invention relates to an environment recognition apparatus outside a vehicle that identifies a floating object from a three-dimensional object existing in a detection area.

  Conventionally, a specific object such as a vehicle positioned in front of the host vehicle is detected, and a collision with a preceding vehicle is avoided (collision avoidance control), or the distance between the preceding vehicle and the preceding vehicle is controlled to be a safe distance ( (Cruise control) technology is known (for example, Patent Documents 1 and 2).

  By the way, especially in cold regions and places with high altitudes, a mass of water vapor may float above the road, or white exhaust gas may be discharged from the exhaust pipe of a preceding vehicle and may stay immediately without spreading. Therefore, it is necessary to take measures to prevent such suspended matter such as water vapor or exhaust gas from being erroneously determined as a specific object such as a vehicle or a pedestrian, and stop or deceleration control to avoid these.

  Therefore, the three-dimensional parts in the detection region in the image picked up by the image pickup device are grouped as a three-dimensional object based on the relative distance, and any of the average value, variance value, skewness, or kurtosis of the luminance with respect to the luminance histogram of the three-dimensional object. A technique for determining whether a three-dimensional object is a white floating substance based on one or a plurality of feature amounts is disclosed (for example, Patent Document 3).

Japanese Patent No. 3349060 JP 2010-224925 A JP 2012-243049 A

  However, for example, an image captured in a dark time zone has a low luminance as a whole, and a difference in luminance between a non-floating object (an object other than a floating substance) such as a vehicle or a pedestrian and a background is small, so The edge component of suspended matter may decrease. In addition, the image captured in the bright time zone has a high brightness as a whole, the number of pixels with the maximum brightness value increases, and the brightness difference between the non-floating object and the background becomes small. The edge component may decrease.

  As described above, when the edge component of the non-floating object decreases, the histogram of the three-dimensional object related to the non-floating object cannot be accurately detected. It is difficult to discriminate the object from the object, and there is a possibility that the detection accuracy of the suspended object is lowered.

  The present invention has been made in view of such problems, and an object of the present invention is to provide a vehicle environment recognition device that can accurately detect floating substances such as water vapor and exhaust gas.

  In order to solve the above-described problem, the vehicle environment recognition apparatus according to the present invention specifies a three-dimensional object based on a predetermined image and derives a rectangular region including the three-dimensional object; A luminance value on the three-dimensional object side and a luminance value on the opposite side to the three-dimensional object are derived on the basis of the boundary with the three-dimensional object in a rectangular area, and the luminance value on the three-dimensional object side is the opposite side. And a floating substance specifying unit that specifies the three-dimensional object as the floating substance when the luminance value is higher than a predetermined threshold value.

  The outside environment recognition device of the present invention specifies a solid object based on a predetermined image and derives a rectangular area including the solid object, and a plurality of the rectangular areas. The luminance value of the peak of the three-dimensional object included in each of the divided regions is derived, and the number of the divided regions in which the luminance value of the peak is equal to or greater than a predetermined luminance threshold is equal to the predetermined divided region When the number is equal to or greater than a few threshold values, a floating substance specifying unit that specifies the three-dimensional object as the floating substance is provided.

  According to the present invention, it is possible to accurately detect suspended matters such as water vapor and exhaust gas.

It is the block diagram which showed the connection relation of the environment recognition system. It is the functional block diagram which showed the schematic function of the external environment recognition apparatus. It is explanatory drawing for demonstrating a luminance image and a distance image. It is explanatory drawing which illustrated the solid object specific process. It is explanatory drawing which illustrated the low-intensity floating substance specific process. It is explanatory drawing which illustrated high-intensity floating substance specific processing. It is a figure explaining a mode that a preceding vehicle passes a floating thing, and specification of a solid thing at that time. It is a figure explaining a mode that a floating thing occurs from a preceding vehicle, and specification of a solid thing at that time. It is a figure explaining the point reset process for floating matter generation by symmetry determination. It is a flowchart which shows the flow of a vehicle exterior environment recognition process. It is a flowchart which shows the flow of a score reset process. It is a flowchart which shows the flow of the floating object rush point reset process. It is a flowchart which shows the flow of the point reset process for floating matter generation. It is a flowchart which shows the flow of a floating substance specific process. It is a flowchart which shows the flow of a low-intensity floating substance determination process. It is a flowchart which shows the flow of a high-intensity floating substance determination process.

  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 shown in the embodiment are merely examples for facilitating understanding of the invention, and do not limit the present invention unless otherwise specified. In the present specification and drawings, elements having substantially the same function and configuration are denoted by the same reference numerals, and redundant description is omitted, and elements not directly related to the present invention are not illustrated. To do.

(Environment recognition system 100)
FIG. 1 is a block diagram showing a connection relationship of the environment recognition system 100. The environment recognition system 100 includes an imaging device 110, a vehicle exterior environment recognition device 120, and a vehicle control device (ECU: Engine Control Unit) 130 provided in the host vehicle 1.

  The imaging device 110 includes an imaging element such as a charge-coupled device (CCD) or a complementary metal-oxide semiconductor (CMOS), captures an environment corresponding to the front of the host vehicle 1, and captures three hues (R ( A color image or a monochrome image composed of (red), G (green), and B (blue) can be generated. Here, a color image picked up by the image pickup device 110 is adopted as a luminance image, and is distinguished from a distance image described later.

  In addition, the imaging devices 110 are arranged in a 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 image data obtained by imaging a three-dimensional object existing in the detection area in front of the host vehicle 1, for example, every 1/60 second frame (60 fps). Here, the recognized three-dimensional object is not only a three-dimensional object that exists independently such as a vehicle, a pedestrian, a traffic light, a road (traveling path), a guardrail, or a building, but also a three-dimensional object such as a tail lamp, a blinker, or a lighting part of a traffic light. The thing which can be specified as a part of is also included. Each functional unit in the following embodiment performs each process for each frame in response to such update of the image data.

  The vehicle exterior environment recognition device 120 acquires image data from each of the two imaging devices 110, derives parallax using so-called pattern matching, and associates the derived parallax information (corresponding to a relative distance described later) with the image data. Generate a distance image. The luminance image and the distance image will be described in detail later. In addition, the vehicle environment recognition apparatus 120 uses any one of the three-dimensional objects in the detection area in front of the host vehicle 1 using the luminance based on the luminance image and the relative distance (depth distance) from the host vehicle 1 based on the distance image. Specify whether it corresponds to.

  When the specific object is specified, the outside-vehicle environment recognition device 120 determines whether or not there is a high possibility that the specific object and the host vehicle 1 collide while tracking the specific object (for example, a preceding vehicle). Here, when it is determined that the possibility of the collision is high, the outside environment recognition device 120 displays a warning (notification) to the driver through the display 122 installed in front of the driver, and the vehicle control device. Information indicating that is output to 130.

  The vehicle control device 130 receives a driver's operation input through the steering wheel 132, the accelerator pedal 134, and the brake pedal 136, and controls the host vehicle 1 by transmitting it to the steering mechanism 142, the drive mechanism 144, and the brake mechanism 146. In addition, the vehicle control device 130 controls the drive mechanism and the braking mechanism in accordance with instructions from the outside environment recognition device 120.

  As described above, in the environment recognition system 100, the three-dimensional object shown in the luminance image is identified as, for example, a preceding vehicle that exists in front of the host vehicle 1, and control for following the preceding vehicle or avoiding a collision with the preceding vehicle is performed. Execute control.

  Below, the structure of the external environment recognition apparatus 120 is explained in full detail. Here, the procedure for identifying the three-dimensional object characteristic of the present embodiment as a floating object will be described in detail, and the description of the configuration unrelated to the characteristics of the present embodiment will be omitted.

(Vehicle environment recognition device 120)
FIG. 2 is a functional block diagram showing a schematic function of the outside environment recognition device 120. As shown in FIG. 2, the vehicle exterior 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 bidirectional information exchange with the imaging device 110 and the vehicle control device 130. The data holding unit 152 includes a RAM, a flash memory, an HDD, and the like. The data holding unit 152 holds various pieces of information necessary for the processing of each function unit described below, and temporarily holds a luminance image received from the imaging device 110. To do.

  The central control unit 154 is configured by a semiconductor integrated circuit including a central processing unit (CPU), a ROM storing a program, a RAM as a work area, and the like, and through the system bus 156, an I / F unit 150, a data holding unit 152 and the like are controlled. In the present embodiment, the central control unit 154 also functions as the image processing unit 160, the three-dimensional position deriving unit 162, the three-dimensional object specifying unit 164, the floating object specifying unit 166, the specific object specifying unit 168, and the score resetting unit 170. To do. In the following, based on the general purpose of such functional units, image processing, three-dimensional object identification processing, suspended matter identification processing (low-intensity suspended matter identification processing, high-intensity suspended matter identification processing), specific object identification performed for each frame Detailed operations of the processing and the point reset processing will be described in order.

(Image processing)
The image processing unit 160 acquires a luminance image from each of the two imaging devices 110, and selects a block corresponding to a block arbitrarily extracted from one luminance image (for example, an array of horizontal 4 pixels × vertical 4 pixels) as the other luminance image. The parallax is derived using so-called pattern matching that is searched from the above. Here, “horizontal” indicates the horizontal direction of the screen, and “vertical” indicates the vertical direction of the screen.

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

  However, the image processing unit 160 can derive the parallax for each parallax block which is a unit of detection resolution, but cannot recognize what kind of three-dimensional object the parallax block is. Therefore, the parallax information is independently derived not in units of solid objects but in units of detection resolution (for example, units of parallax blocks) in the detection region. Here, an image in which the parallax information derived in this way (corresponding to a relative distance z described later) is associated with a luminance image is referred to as a distance image.

  FIG. 3 is an explanatory diagram for explaining the luminance image 212 and the distance image 214. For example, it is assumed that a luminance image 212 as shown in FIG. 3A is generated for the detection region 216 through the two imaging devices 110. However, here, for easy understanding, only one of the two luminance images 212 generated by each of the imaging devices 110 is schematically illustrated. In the present embodiment, the image processing unit 160 obtains a parallax for each parallax block from such a luminance image 212 and generates a distance image 214 as shown in FIG. Each parallax block in the distance image 214 is associated with the parallax of the parallax block. Here, for convenience of explanation, the parallax block from which the parallax is derived is represented by black dots.

(3D position derivation process)
Returning to FIG. 2, the three-dimensional position deriving unit 162 uses the so-called stereo method to calculate disparity information for each disparity block in the detection region 216 based on the distance image 214 generated by the image processing unit 160. It is converted into a three-dimensional position in the real space including the horizontal distance x, the height y, and the relative distance z. Here, the stereo method is a method of deriving a relative distance z to the imaging device 110 of the three-dimensional part from the parallax in the distance image 214 of the three-dimensional part (a pixel or a block composed of a plurality of pixels) by using a triangulation method. is there. At this time, the three-dimensional position deriving unit 162 is based on the relative distance z of the three-dimensional part and the detected distance on the distance image 214 between the point on the road surface and the three-dimensional part at the same relative distance z as the three-dimensional part. The height y of the three-dimensional part from the road surface is derived. Then, the position information indicating the derived three-dimensional position is associated with the distance image 214 again. Since various known techniques can be applied to the relative distance z deriving process and the three-dimensional position specifying process, the description thereof is omitted here.

(3D object identification processing)
The three-dimensional object specifying unit 164 groups three-dimensional parts having a three-dimensional position difference within a predetermined range to obtain a three-dimensional object. Specifically, the three-dimensional object specifying unit 164 includes a three-dimensional part in the distance image 214 in which a difference in horizontal distance x, a difference in height y, and a difference in relative distance z are within a predetermined range (for example, 0.1 m). They are grouped on the assumption that they correspond to the same specific object. In this way, a three-dimensional object that is a virtual three-dimensional group is generated. The above range is represented by a distance in real space, and can be set to an arbitrary value by a manufacturer or a passenger. In addition, regarding the three-dimensional part newly added by grouping, the three-dimensional object specifying unit 164 also sets the difference of the horizontal distance x, the difference of the height y, and the difference of the relative distance z within a predetermined range from the three-dimensional part. The three-dimensional sites in are further grouped. As a result, all the three-dimensional sites that can be assumed to be the same specific object are grouped.

Here, the difference in the horizontal distance x, the difference in the height y, and the difference in the relative distance z are determined independently, and only when all are included in the predetermined range, the same group is used. You can also. For example, the root mean square of the difference in horizontal distance x, the difference in height y, and the difference in relative distance z ((difference in horizontal distance x) ² + (difference in height y) ² + (difference in relative distance z) ² ) Are included in the predetermined range, the same group may be used. With this calculation, an accurate distance between the three-dimensional parts in the real space can be derived, so that the grouping accuracy can be increased.

  In addition, the three-dimensional object specifying unit 164 includes a three-dimensional object obtained by grouping three-dimensional parts within a predetermined range, and has a rectangular shape (hereinafter simply referred to as a rectangular shape) whose outline is composed of a horizontal line and a vertical line. An area (hereinafter also referred to as a window) is set.

  Furthermore, when the three-dimensional object is specified in the current frame, the three-dimensional object specifying unit 164 specifies the position information of the three-dimensional object (position information of the three-dimensional parts grouped as a three-dimensional object) and the previous frame. Based on the position information of the three-dimensional object, the correspondence between the three-dimensional object specified in the current frame and the three-dimensional object specified in the previous frame is determined. If the three-dimensional object specifying unit 164 determines that the three-dimensional object specified in the current frame corresponds to the three-dimensional object specified in the previous frame, these three-dimensional objects are associated with each other.

  FIG. 4 is an explanatory diagram illustrating the three-dimensional object specifying process. For example, in the luminance image 212 as shown in FIG. 4A, the three-dimensional object specifying unit 164 uses the position information based on the distance image 214 as shown in FIG. , The difference in height y, and the difference in relative distance z are grouped together in a three-dimensional region of the floating object 220 within a predetermined range. At this time, the three-dimensional object specifying unit 164 sets the window 222 including the three-dimensional object together with the three-dimensional object (an assembly of three-dimensional parts). In FIG. 4B, only the relationship between the horizontal distance x and the relative distance z is shown for convenience of explanation, and the height y is omitted.

(Floating substance identification processing)
The floating substance specifying unit 166 specifies a floating substance from the three-dimensional object specified by the three-dimensional object specifying unit 164.

  By the way, in the brightness image 212 in which an object with extremely low brightness such as a black vehicle or a pedestrian in black clothes is captured in a dark environment such as night (hereinafter also referred to as a low brightness scene), the brightness image 212 is extremely bright. An object having a low is identified as a three-dimensional object. In the luminance image 212, since the luminance difference between the solid object and the background applied to the extremely low-luminance object is small, the edge component of the solid object applied to the non-floating object is reduced.

  In addition, the luminance value of the luminance image 212 captured in an overall bright environment such as noon (hereinafter also referred to as a high luminance scene) is high overall. Therefore, when a non-floating object is specified as a three-dimensional object, the luminance values of the three-dimensional object and the background in the luminance image 212 both increase, and the pixels having the maximum luminance value also increase. Since the luminance difference becomes small, the edge component of the three-dimensional object related to the non-floating object decreases.

  Thus, when the edge component of a non-floating object decreases, the luminance histogram when a non-floating object is specified as a three-dimensional object is similar to the luminance histogram when a floating object is specified as a three-dimensional object. It becomes distribution and it becomes difficult to distinguish and specify non-floating matter and floating matter using the luminance histogram.

  Therefore, in a low-luminance scene and a high-luminance scene, there may arise a problem that floating objects cannot be accurately identified using a luminance histogram. Therefore, in the present embodiment, a specific process (a low-luminance floating substance identification process) that identifies a floating object under a determination condition (first determination condition) based on the luminance value of a three-dimensional object corresponding to each of a low-luminance scene and a high-luminance scene. , A high-intensity floating substance specifying process) specifies the floating substance from the three-dimensional object specified by the three-dimensional object specifying unit 164.

(Low-intensity suspended matter identification process)
FIG. 5 is an explanatory diagram illustrating the low-intensity suspended matter specifying process. The floating substance specifying unit 166 performs the low-luminance floating substance specifying process as the floating substance specifying process with the determination condition according to the low-luminance scene. As shown in FIG. 5A, the floating substance specifying unit 166 determines the contour of the three-dimensional object in the window 222 based on the position information of the three-dimensional part derived by the three-dimensional position deriving unit 162, and The formed inner area is set as the access area 224.

  Then, as shown in FIG. 5B, the suspended matter specifying unit 166 determines the horizontal of the access region 224 at an arbitrary position in the vertical direction based on the position information of the three-dimensional part derived by the three-dimensional position deriving unit 162. A direction width (access area width) W is derived. After that, as shown in FIG. 5C, the floating substance specifying unit 166 sets the access area width W to the horizontal direction in the horizontal direction with respect to the boundary of the access area 224, that is, the outline of the three-dimensional object. A luminance detection range W / 6 that is a range of 1/6 is set. Similarly, the floating matter specifying unit 166 sets the access area width W and the luminance detection range W / 6 for each pixel corresponding to all the positions in the vertical direction.

  Then, the floating matter specifying unit 166 has an average value (average luminance value) of luminance values of pixels on the three-dimensional object side (access area 224 side) in which the luminance detection range W / 6 is set, and the luminance detection range W / 6. The average luminance value of the pixel on the opposite side (background side and outside the boundary of the access area 224) from the set three-dimensional object is derived.

  The floating object specifying unit 166 has a characteristic that the three-dimensional object seems to be a floating object when the luminance difference between the derived average luminance value on the three-dimensional object side and the average luminance value on the background side is equal to or larger than a predetermined luminance difference threshold value. Then, +1 is added (added 1) to the number of low-luminance floating object points. On the other hand, when the difference in brightness between the derived average brightness value on the three-dimensional object side and the average brightness value on the background side is less than a preset brightness difference threshold, the floating object specifying unit 166 has a feature that the three-dimensional object is likely to be a floating object. In other words, it is determined that the three-dimensional object has a characteristic like a non-floating substance, and −1 is added to the low-luminance floating point number (1 is subtracted).

  Here, the brightness difference threshold value is a background brightness value in consideration of a floating object that is white and has a relatively high brightness value, and an object (for example, a black vehicle or a pedestrian) that has a low brightness value in a low brightness scene. Is set so that a three-dimensional object that is a floating object can be identified.

  As described above, the floating object specifying unit 166 includes the average brightness value on the solid object side (accelerator area 222 side) and the average brightness value on the background side (outside the accelerator area 222), with the outline of the solid object in the window 222 as a boundary. If the luminance difference is equal to or greater than the luminance difference threshold, it is determined that the three-dimensional object has a characteristic like a floating object. Therefore, even in low-luminance scenes, it is possible to determine whether or not a solid object has characteristics that are like a floating object based on the luminance difference between the three-dimensional object and the background. Can do.

(High-intensity suspended matter identification processing)
FIG. 6 is an explanatory diagram illustrating the high-intensity suspended matter specifying process. The floating matter specifying unit 166 performs the high-intensity floating matter specifying process as the floating matter specifying process suitable for the high-intensity scene. As illustrated in FIG. 6A, the floating matter specifying unit 166 sets, as the access region 224, an area including only a three-dimensional body aggregate in the window 222 as in the low-luminance floating matter specifying process. The suspended matter specifying unit 166 sets nine divided regions 226a to 226i obtained by equally dividing the window 222 into three parts in the vertical direction and the horizontal direction.

  Then, as illustrated in FIG. 6B, the floating matter specifying unit 166 generates a luminance histogram for the luminance of the access area 224 in each of the divided areas 226a to 226i for each of the divided areas 226a to 226i. In the luminance histogram, the horizontal axis is the luminance value (0 to 255), and the vertical axis is the frequency.

  The floating substance specifying unit 166 derives the luminance value (peak luminance value) having the highest frequency for each of the divided regions 226a to 226i in the generated luminance histogram (indicated by the blackening luminance value in FIG. 6B). . In addition, the floating matter specifying unit 166 determines the number of divided areas (the number of high-luminance divided areas) in which the derived peak luminance value is equal to or greater than a predetermined luminance value threshold (for example, 166, indicated by a broken line in FIG. 6B). To derive.

  Then, the floating object specifying unit 166 determines that the three-dimensional object has a characteristic like a floating object when the number of high luminance divided areas is equal to or greater than a preset high luminance divided area number threshold (for example, “7”). Then add +1 to the number of high-intensity suspended matter. On the other hand, when the number of high-luminance divided areas is less than the preset high-luminance divided area number threshold, the floating object specifying unit 166 does not have a feature that seems to be a floating object, that is, the three-dimensional object is not It is determined that the object has a characteristic like a floating object, and -1 is added (1 is subtracted) to the number of high-luminance floating object points.

  Here, for example, when the three-dimensional object is a vehicle, the access area 224 corresponding to the rear window or tire of the vehicle has a low luminance value, so that the division including the access area 224 corresponding to the rear window or tire of the vehicle is performed. The peak luminance value of the region is less than the luminance value threshold. Therefore, when the three-dimensional object is a vehicle, there is a high possibility that the number of high-luminance divided areas will be less than a preset high-luminance divided area number threshold.

  On the other hand, when the three-dimensional object is a white floating object, the luminance value of the access area 224 corresponding to the floating object is high overall, and thus the peak luminance value of the divided area including the access area 224 corresponding to the floating object. Is greater than or equal to the luminance value threshold. Therefore, when the three-dimensional object is a floating object, there is a high possibility that the number of high-intensity divided areas is equal to or higher than a preset high-intensity divided area number threshold.

  The floating object specifying unit 166 compares the number of high-luminance divided areas with the threshold value for the number of high-luminance divided areas. Since it can be determined whether or not has a characteristic like a floating substance, the floating substance can be specified with high accuracy.

  Then, the floating matter specifying unit 166 determines whether the low-luminance floating point score or the high-luminance floating point score accumulated for each frame is larger than 0 point (first score threshold). If the number of high-intensity floating objects is greater than 0, the three-dimensional object is identified as a floating object. In addition, the floating matter specifying unit 166 specifies a three-dimensional object as a non-floating matter when the low-luminance floating point number and the high-luminance floating point number are 0 points (second score threshold) or less. Note that the floating matter specifying unit 166 does not specify that the three-dimensional object is any specific object even when the three-dimensional object is specified as a non-floating object.

  In addition, the upper and lower limits are provided for the number of low-luminance floating points and the number of high-luminance floating points after integration, for example, an upper limit of 20 points and a lower limit of -20 points. In this way, by providing upper and lower limits, the absolute value of the score does not increase even if the period during which floating objects are not detected or the period during which floating objects are detected extends for a long time. When the presence / absence of detection is switched, the presence / absence can be quickly determined.

  As described above, the floating matter specifying unit 166 performs the low-intensity floating matter specification processing and the high-intensity floating matter specification processing for each frame, and at least one of the low-intensity floating point number or the high-intensity floating point number is larger than 0. For example, a three-dimensional object is specified as a floating object.

  By the way, since the imaging device 110 is fixed to the own vehicle 1, the image is blurred in the luminance image 212 captured by the imaging device 110 due to the vertical pitching when the own vehicle 1 is traveling. Solid objects may be projected. In addition, even when a vehicle or a person crossing the front of the host vehicle 1 is moving, a three-dimensional object with a blurred image may appear in the luminance image 212 captured by the imaging device 110. The three-dimensional object with the blurred image has a small luminance difference from the background, and the edge component of the three-dimensional object in the luminance image 212 is reduced, making it difficult to accurately detect the histogram of the three-dimensional object. Therefore, in the luminance image 212 in which the three-dimensional object in which the image is blurred is projected, even if the three-dimensional object is a vehicle or a person, it may be erroneously specified as a floating object.

  Therefore, even when the floating matter specifying unit 166 identifies a three-dimensional object as a floating matter in the low-intensity floating matter identification processing and the high-luminance floating matter identification processing, the floating matter is not more than a predetermined speed in the depth direction and the horizontal direction. In consideration of not moving, if the following conditions are satisfied, the solid object once specified as a floating substance is specified (re-specified) as a non-floating substance.

  Specifically, the floating substance specifying unit 166 once specified a floating object when the three-dimensional object moved in the depth direction at ± 10 km / h or more and moved in the horizontal direction at ± 2 km / h or more. Specify (re-specify) the three-dimensional object as non-floating material.

  In addition, even when the three-dimensional object is moving in the horizontal direction at ± 15 km / h or more, the floating substance specifying unit 166 specifies (re-specifies) the three-dimensional object once identified as a floating substance as a non-floating substance. .

(Specific product identification processing)
The specific object specifying unit 168 is any specific object (vehicle, person, etc.) based on the size, shape, hue, etc. of the three-dimensional object specified as a non-floating object by the floating object specifying unit 166. To determine.

(Point reset processing)
FIG. 7 is a diagram for explaining a state in which the preceding vehicle passes through the floating object and the identification of the three-dimensional object at that time. FIG. 8 is a diagram for explaining how floating objects are generated from a preceding vehicle and specifying a three-dimensional object at that time.

  As shown in FIG. 7A, when the preceding vehicle 300 enters the floating object 310, if the preceding vehicle 300 is located closer to the own vehicle 1 than the floating object 310, FIG. As shown, the preceding vehicle 300 located on the own vehicle 1 side is specified as a three-dimensional object, and a window 222 including the three-dimensional object (preceding vehicle 300) is set. And since the solid object in the window 222 has the characteristic like a non-floating substance, in the above-mentioned floating substance identification process, a solid object will be identified as a non-floating substance (vehicle). At this time, there is a high possibility that the low-luminance floating point score and the high-luminance floating point number are -20 points that are lower limit values or values close thereto.

  In this state, as shown in FIG. 7C, when the preceding vehicle 300 enters the floating object 310, the floating object 310 is positioned closer to the own vehicle 1 than the preceding vehicle 300. As shown in d), the floating object 310 located on the own vehicle 1 side is specified as a three-dimensional object, and a window 222 including the three-dimensional object (floating object 310) is set. However, in the above-described floating matter specifying process, even if it is detected that there is a characteristic like a floating matter, the low-luminance floating point number and the high-luminance floating point number are only updated by +1 every frame. Therefore, until at least one of the low-intensity floating point number and the high-intensity floating point number is greater than 0 (for example, a period of 20 frames) It will be identified as floating.

  Further, as shown in FIG. 8 (a), when the suspended matter 310 such as exhaust gas is generated from the preceding vehicle 300, when the suspended matter 310 is generated, as shown in FIG. A window 222 that is specified as a three-dimensional object and includes the three-dimensional object (preceding vehicle 300) is set. Since the three-dimensional object in the window 222 has a characteristic like a non-floating object, the three-dimensional object (preceding vehicle 300) is specified as a non-floating object (vehicle) in the above-described floating object specifying process.

  In such a state, as shown in FIG. 8 (c), when the time has elapsed since the floating object 310 is generated, the preceding vehicle 300 moves forward, but the floating object 310 is moved from the absolute position where it was generated. There is almost no movement (relative distance to the host vehicle 1 is shortened), and the floating object 310 is positioned closer to the host vehicle 1 than the preceding vehicle 300. In such a case, as shown in FIG. 8D, the floating object 310 is specified as a three-dimensional object, and a window 222 including the three-dimensional object (floating object 310) is set. Even in such a case, even if the three-dimensional object is actually the floating object 310, it is specified as a non-floating object.

  As described above, when a considerable period is required from the state in which the non-floating object has been specified until the floating object is specified, the outside environment recognition device 120 avoids or stops so as not to collide with the non-floating object. Information indicating that is output to the vehicle control device 130. Therefore, the vehicle control device 130 performs control to avoid or stop the host vehicle 1. Such control will cause the driver to take an unexpected action, such as avoiding floating substances that can actually travel as it is, and will give the driver a sense of incongruity.

  Therefore, the score resetting unit 170 determines the three-dimensional object based on the second determination condition different from the first determination condition when the three-dimensional object is specified as a non-floating object before the above-described floating object specifying process. If it is determined that the object has characteristics of floating objects, resetting the number of low-intensity floating points and high-intensity floating points to 0 Execute point reset processing. Hereinafter, the score reset process will be specifically described.

(Floating object entry point reset processing)
The score resetting unit 170 relates the positional information of the three-dimensional object (an assembly of three-dimensional parts) in the luminance image 212 captured in the previous frame for the window 222 including the three-dimensional object identified as a non-floating object. Based on the position information of the three-dimensional object, the acceleration of the three-dimensional object and the jerk that is a derivative of the acceleration are derived.

  In addition, the score reset unit 170 determines whether or not the brake lamp is lit in the window 222 identified as a non-floating object. As a method of determining the lighting of the brake lamp, for example, the lighting of the brake lamp may be detected based on a luminance change or an area change of a pixel region that satisfies a predetermined range of R component in the window 222. A known brake lamp detection method may be used.

  If the derived acceleration is less than -0.2G, the jerk is 0.01G or more, and the brake lamp is turned off, the score resetting unit 170 can obtain a low-intensity floating point number and a high-intensity floating point number. Is reset to 0. Here, as shown in FIG. 7, when the three-dimensional object in the window 222 is switched from a non-floating object to a floating object, the absolute position of the floating object 310 does not change, and thus a negative acceleration within a predetermined range is detected. In addition, a positive jerk within a predetermined range is detected. On the other hand, even when the preceding vehicle 300 brakes, a negative acceleration within a predetermined range is detected, and a positive jerk within a predetermined range is detected. However, when the three-dimensional object in the window 222 is switched from a non-floating object to a floating object, the detected negative acceleration and positive jerk values are larger than when the preceding vehicle 300 is braked. Therefore, when a negative acceleration (less than −0.2 G) in a predetermined range is detected and a positive jerk (0.01 G or more) in the predetermined range is detected, the three-dimensional object in the window 222 is not It can be determined that the suspended matter has switched to the suspended matter.

  Further, when the preceding vehicle 300 applies a brake, the lighting of the brake lamp of the preceding vehicle 30 is detected. Therefore, in addition to the negative acceleration and the positive jerk, the presence or absence of the lighting of the brake lamp can be detected. More precisely, it can be determined that the three-dimensional object in the window 222 has been switched from a non-floating object to a floating object.

  As described above, based on the acceleration, jerk, and whether or not the brake lamp is lit, it is determined whether or not the three-dimensional object has been switched from the non-floating substance to the floating substance when the non-floating substance enters the floating substance. be able to. Note that it may be determined whether the three-dimensional object has been switched from a non-floating object to a floating object based on the acceleration and jerk.

(Floating object generation point reset processing)
The score resetting unit 170 applies, for example, a 3 × 3 pixel spatial differential filter to the entire luminance image 212 and creates an edge image by the spatial differential filter. Then, the point reset unit 170 calculates the total number of edges (the total number of edges) included in the window 222 in the edge image.

  Here, when the three-dimensional object in the window 222 is a floating object, the total number of edges is small, and when the three-dimensional object in the window 222 is a non-floating object, the total number of edges is large. Therefore, when the solid object in the window 222 continues to be a non-floating object, the rate of change in the total number of edges is small, but when the solid object in the window 222 is switched from a non-floating object to a floating object, The rate of change of the total number of edges increases.

  Therefore, the score resetting unit 170 sets the number of edges in the associated window 222 set for the luminance image 212 in the previous frame to the number in the window 222 set for the luminance image 212 in the current frame. The rate of change of the total number of edges is derived. Then, the score resetting unit 170 determines that the three-dimensional object in the window 222 has been switched from a non-floating object to a floating object when the change rate of the derived edge total number is 80% (edge total number threshold) or more.

  FIG. 9 is a diagram for explaining the float resetting point reset processing based on the symmetry determination. Further, as shown in FIG. 9, the score resetting unit 170 sets a center line CL extending in the vertical direction that equally divides the window 222 in the horizontal direction with respect to the window 222 set for the luminance image 212. Then, the score resetting unit 170 is configured such that, for each pixel in the vertical direction on the center line CL, the pixels separated by the separation position IP set according to the relative distance in each of the left and right directions with reference to the set center line CL. Both are determined to be pixels corresponding to a three-dimensional object. Then, the score reset unit 170 derives the total number of symmetry determined that the pixels separated by the separation positions IP in the left and right directions with respect to the center line CL are both pixels corresponding to the three-dimensional object. Note that the separation position IP is set such that a pixel in the window 222 is selected.

  Here, as shown in FIG. 9A, when the three-dimensional object in the window 222 is a floating object, since the symmetry in the left-right direction with respect to the center line CL is low, the total number of symmetry decreases. On the other hand, as shown in FIG. 9B, when the three-dimensional object in the window 222 is a non-floating object (for example, a vehicle), the symmetry in the left-right direction with respect to the center line CL is high. Will increase. Therefore, when the three-dimensional object in the window 222 continues as a non-floating object, the rate of change in the total symmetry is small, but when the three-dimensional object in the window 222 is switched from a non-floating object to a floating object. The rate of change of the total number of symmetry is large.

  Therefore, the score resetting unit 170 sets the total symmetry of the associated window 222 set for the current luminance image 212 with respect to the total symmetry of the window 222 set for the previous luminance image 212. The rate of change is derived. Then, the score resetting unit 170 determines that the three-dimensional object in the window 222 has been switched from a non-floating object to a floating object when the derived rate of change is 80% (symmetry threshold) or more.

  Then, when at least one of the change rate of the total number of edges and the change rate of the total number of symmetry is 80% or more, the point reset unit 170 resets the low-luminance floating point number and the high-luminance floating point number to 0. Thereby, the suspended | floating matter specific | specification part 166 can identify a suspended | floating matter in the period from 0 point to 1 point which is shorter than the period required from -20 points to 1 point which is a lower limit, for example. A three-dimensional object can be specified as a floating substance.

(External vehicle environment recognition process)
Next, the flow of the outside environment recognition process including the above-described image processing, three-dimensional object specifying process, floating object specifying process, specified object specifying process, and point reset process executed by the central control unit 154 will be described.

  FIG. 10 is a flowchart showing the flow of the external environment recognition process. FIG. 11 is a flowchart showing the flow of the score reset process. FIG. 12 is a flowchart showing the flow of the float entry point reset process. FIG. 13 is a flowchart showing the flow of the float generation point reset process. FIG. 14 is a flowchart showing the flow of the suspended matter specifying process. FIG. 15 is a flowchart showing the flow of the low-luminance floating substance determination process. FIG. 16 is a flowchart showing the flow of the high-intensity floating substance determination process. The flowchart of the point reset process shown in FIG. 11 and the flowchart of the floating substance specifying process shown in FIG. 14 are subroutines of the flowchart of the outside environment recognition process shown in FIG. Further, the flowchart of the floating object entry point reset process shown in FIG. 12 and the flowchart of the floating substance generation point reset process shown in FIG. 13 are subroutines of the flowchart of the point reset process shown in FIG. Further, the flowchart of the low-intensity floating substance determination process shown in FIG. 15 and the flowchart of the high-intensity floating substance determination process shown in FIG. 16 are subroutines of the flowchart of the floating substance identification process shown in FIG.

  As shown in FIG. 10, first, the image processing unit 160 processes the image acquired from the imaging device 110 to generate a distance image 214 (S400), and the three-dimensional position deriving unit 162 determines the three-dimensional part from the image. The dimensional position information is derived (S402), and the three-dimensional object specifying unit 164 specifies three-dimensional objects grouped based on the three-dimensional position (S404). When the point reset unit 170 determines that the three-dimensional object in the window 222 has been switched from a non-floating object to a floating object, a point reset process is performed to reset the low-luminance floating object score and the high-luminance floating object score to zero. (S500). Thereafter, the floating matter specifying unit 166 executes a floating matter specifying process for specifying whether the three-dimensional object in the window 222 is a floating matter or a non-floating matter (S600). A specified object specifying process for specifying which one of the three-dimensional objects specified as a floating object is executed (S406), and the vehicle exterior environment recognition process is terminated. Hereinafter, the point reset process S500 and the suspended matter specifying process S600 will be described in detail.

(Point reset processing S500)
As shown in FIG. 11, the point reset unit 170 has one or more three-dimensional objects specified in the three-dimensional object specifying process S404, and in the point reset process S500, the non-floating object is specified in the previous floating object specifying process S600. It is determined whether there is a three-dimensional object that has been selected yet (S502). If there is a solid object that has not yet been selected (YES in S502), the point reset unit 170 selects one solid object that has not yet been selected (S504).

  Next, the point reset unit 170 executes the floating object entry point reset process (S506), and then executes the floating object generation point reset process (S508). The float entry point reset processing S506 and the float generation point reset processing S508 will be described later.

  The point reset unit 170 is a floating object for resetting the low-intensity floating point number and the high-intensity floating point number in at least one of the floating point entry point reset process S506 and the floating point generation point reset process S508, which will be described in detail later. It is determined whether or not the point reset flag is turned on (S510). If the floating point number reset flag is not turned on (NO in S510), the process returns to S502.

  On the other hand, if the floating point reset flag is on (YES in S510), the point reset unit 170 resets the low-intensity floating point and the high-intensity floating point to 0 (S512), and the floating point reset flag Is turned off (S514), and the process returns to S502.

  In the process of S502, when there is no solid object that has not yet been selected (NO in S502), the score reset process S500 is terminated.

(Floating object entry point reset processing S506)
As illustrated in FIG. 12, the point reset unit 170 performs acceleration and jerk of the three-dimensional object based on the position information of the three-dimensional object derived in the three-dimensional position derivation process S <b> 402 and the previous position information of the three-dimensional object. Is derived (S506-1). In addition, the score reset unit 170 detects whether or not the brake lamp is lit in the window 222 including the three-dimensional object (S506-2).

  And the score reset part 170 judges whether the acceleration derived | led-out by S506-1 is less than -0.2G (S506-3), and if an acceleration is less than -0.2G (in S506-3) YES), it is determined whether or not the jerk derived in S506-1 is 0.01 G or more (S506-4). If the jerk is 0.01G or more (YES in S506-4), the score reset unit 170 determines whether or not the brake lamp is turned off (S506-5), and if the brake lamp is turned off. (YES in S506-5), the floating point reset flag is turned on (S506-6), and the floating point entry point reset processing S506 is terminated.

  On the other hand, if the acceleration is not less than −0.2 G (NO in S506-3), the jerk is not 0.01 G or more (NO in S506-4), or the brake lamp is not turned off (NO in S506-5). ), The point reset unit 170 ends the floating point entry point reset processing S506.

(Floating object generation point reset processing S508)
As shown in FIG. 13, the score resetting unit 170 applies a spatial filter to each pixel in the window 222 to derive the edges in the window 222 and derive the total number of derived edges (S508-1). . In addition, the score resetting unit 170 sets a center line CL that equally divides the window 222 in the horizontal direction, and the pixels that are separated from the center line CL by the separation position IP in the left-right direction are both pixels that correspond to solid objects. The total number of sexes is derived (S508-2).

  The score resetting unit 170 determines whether or not the change rate of the total number of edges derived in the current S508-1 is 80% or more with respect to the total number of edges derived in the previous S508-1 (S508-3). . If the rate of change in the total number of edges is not 80% or more (NO in S508-3), the score resetting unit 170 performs the symmetry derived in the current S508-2 with respect to the total symmetry derived in the previous S508-2. It is determined whether the change rate of the total number of sex is 80% or more (S508-4).

  If the change rate of the total number of edges is 80% or more (YES in S508-3) or if the change rate of the total number of symmetry is 80% or more (YES in S508-4), the score resetting unit 170 The floating substance point reset flag is turned on (S508-5), and the floating substance generation point reset process S508 is terminated.

  If the change rate of the total number of edges is not 80% or more (NO in S508-3) and the change rate of the total number of symmetry is not 80% or more (NO in S508-4), the score resetting unit 170 The floating matter generation point reset processing S508 is terminated.

(Floating substance identification processing S600)
As illustrated in FIG. 14, the suspended matter specifying unit 166 determines whether or not there is a solid object that has one or more solid objects specified in the solid object specifying process S404 and has not yet been selected (S602). If there is a solid object that has not yet been selected (YES in S602), the floating matter specifying unit 166 selects one solid object that has not yet been selected (S604).

  Next, the floating matter specifying unit 166 executes a low brightness floating matter specifying process (S606), and then executes a high brightness floating matter specifying process (S608). Note that the low-intensity floating matter specifying process S606 and the high-intensity floating matter specifying process S608 will be described later.

  The floating matter specifying unit 166 determines whether or not the number of low-luminance floating matter points derived in the low-luminance floating matter specification processing S606 described later in detail is greater than zero (S610). If the low-brightness floating point score is not greater than 0 (NO in S610), the floating-point specifying unit 166 determines that the high-brightness floating point number derived in the high-brightness floating point specifying process S608, which will be described in detail later, is 0 point. It is determined whether it is larger (S612).

  If the low-intensity floating point score is 0 or less (NO in S610) and the high-intensity floating point number is 0 or less (NO in S612), the floating matter specifying unit 166 converts the three-dimensional object into a non-floating matter. (S622), and the process returns to S602.

  On the other hand, if the low-intensity floating point score is larger than 0 (YES in S610) or the high-intensity floating point number is larger than 0 (YES in S612), the floating matter specifying unit 166 is the position derived in S402. Based on the information, the depth speed and the horizontal speed of the three-dimensional object are derived (S614). Then, the suspended matter specifying unit 166 determines whether the depth speed is ± 10 km / h or more and the horizontal speed is ± 2 km / h or more (S616), and the depth speed is less than ± 10 km / h or If the horizontal speed is less than ± 2 km / h (NO in S616), it is determined whether the horizontal speed is more than ± 15 km / h (S618). If the absolute value of the horizontal velocity is less than ± 15 km / h (NO in S618), the floating substance specifying unit 166 specifies the three-dimensional object as a floating substance (S620), and returns to the process of S602.

  On the other hand, if the depth speed is ± 10 km / h or more and the horizontal speed is ± 2 km / h or more (YES in S616), or if the horizontal speed is ± 15 km / h or more (YES in S618), floating The object specifying unit 166 specifies the three-dimensional object as a non-floating object (S622), and returns to the process of S602.

(Low-intensity suspended matter identification process S606)
As illustrated in FIG. 15, the floating matter specifying unit 166 sets, as the access region 224, an area of only a three-dimensional part aggregate in the window 222 based on the position information of the three-dimensional part derived by the three-dimensional position deriving unit 162. Set (S606-1).

  Then, the floating substance specifying unit 166 sets the luminance detection range W / 6 in the horizontal direction in the horizontal direction with reference to the access area width W of the set access area 224 and the boundary of the access area 224. Thereafter, the floating object specifying unit 166 derives the average brightness value on the three-dimensional object side where the brightness detection range W / 6 is set and the average brightness value on the background side where the brightness detection range W / 6 is set (S606). -2). Then, the floating object specifying unit 166 derives a luminance difference between the average luminance value on the three-dimensional object side where the luminance detection range W / 6 is set and the average luminance value on the background side where the luminance detection range W / 6 is set. (S606-3).

  Then, the floating matter specifying unit 166 determines whether or not the luminance difference is larger than the luminance difference threshold (S606-4). If the luminance difference is larger than the luminance difference threshold (YES in S606-4), the low-luminance floating matter is determined. It is determined whether or not the score is 20 or more (S606-5). If the low-luminance floating point score is 20 or more (YES in S606-5), the floating-point identification unit 166 ends the low-luminance floating point identification processing S606, and the low-luminance floating point number is less than 20 points. If so (NO in S606-5), +1 is added to the number of low-luminance floating objects (S606-6), and the low-luminance floating substance identification processing S606 is terminated.

  On the other hand, if the luminance difference is equal to or smaller than the luminance difference threshold value (NO in S606-4), the floating matter specifying unit 166 determines whether or not the low-luminance floating point number is -20 points or less (S606-7). . Then, if the low-luminance floating point score is -20 or less (YES in S606-7), the floating-point identification unit 166 ends the low-luminance floating point identification processing S606, and the low-luminance floating point number is -20. If it is larger than the point (NO in S606-7), -1 is added to the low-luminance floating point number (S606-8), and the low-luminance floating matter specifying process S606 is terminated.

(High-intensity suspended matter identification process S608)
As shown in FIG. 16, the floating substance specifying unit 166 sets an access area 224 in the window 222 and sets nine divided areas 226 a to 226 i obtained by dividing the window 222 into three in the vertical direction and the horizontal direction ( S608-1). Then, the floating matter specifying unit 166 generates a luminance histogram for the luminance of the access area 224 in each of the divided areas 226a to 226i for each of the divided areas 226a to 226i (S608-2). Thereafter, the floating matter specifying unit 166 derives the peak luminance value for each of the divided regions 226a to 226i for the generated luminance histogram and derives the number of high luminance divided regions whose peak luminance value is equal to or greater than the luminance value threshold (S608-). 3).

  Then, the floating matter specifying unit 166 determines whether or not the number of high-luminance divided areas is equal to or greater than the threshold value for high-luminance divided areas (S608-4), and the number of high-luminance divided areas is equal to or greater than the threshold value for high-luminance divided areas. If there is (YES in S608-4), it is determined whether or not the number of high-intensity suspended matter points is 20 or more (S608-5). If the high-intensity floating point score is 20 or more (YES in S608-5), the floating-point specifying unit 166 ends the high-intensity floating point specifying process S608 and the high-intensity floating point number is less than 20 points. If so (NO in S608-5), +1 is added to the number of high-intensity suspended matter (S608-6), and the high-intensity suspended matter specifying process S608 is terminated.

  On the other hand, if the number of high-intensity divided regions is less than the high-intensity divided region number threshold (NO in S608-4), the floating matter specifying unit 166 determines whether or not the high-intensity floating point number is −20 points or less. (S608-7). Then, if the high-intensity floating point score is -20 or less (YES in S608-7), the floating-point specifying unit 166 ends the high-intensity floating point specifying process S608, and the high-intensity floating point number is -20. If it is larger than the point (NO in S608-7), -1 is added to the number of high-intensity floating matter points (S608-8), and the high-intensity floating matter specifying process S608 is terminated.

  As mentioned above, although preferred embodiment of this invention was described referring an accompanying drawing, it cannot be overemphasized that this invention is not limited to this embodiment. It will be apparent to those skilled in the art that various changes and modifications can be made within the scope of the claims, and these are naturally within the technical scope of the present invention. Is done.

  Also provided are a program that causes a computer to function as the vehicle exterior environment recognition device 120 and a computer-readable storage medium that stores the program, such as a flexible disk, magneto-optical disk, ROM, CD, DVD, and BD. Here, the program refers to data processing means described in an arbitrary language or description method.

  For example, in the above-described embodiment, an example in which a suspended matter is specified by a low-intensity floating matter specifying process and a high-intensity floating matter specifying process is described. A technique for identifying a floating object according to variation (dispersion) with respect to an average value of a distance of a three-dimensional part, or an average value, a dispersion value, and a distortion value for a luminance histogram of a three-dimensional object as disclosed in Japanese Patent Application Laid-Open No. 2012-243049. The floating substance may be specified by the overall evaluation using the technique of specifying the floating substance based on one or a plurality of feature amounts of degree or kurtosis and the technique of this embodiment. In this way, it is possible to increase the accuracy of identifying the suspended matter.

  In the above embodiment, when the low-luminance floating object score is larger than 0 (first score threshold), the solid object is specified as a floating object, and when the low-luminance floating object score is 0 or less, The three-dimensional object was specified as a non-floating object. However, for example, when the number of low-intensity floating objects is greater than 10 (first score threshold), the three-dimensional object is identified as a floating object, and the number of low-intensity suspended objects is -10 points (second score threshold) or less. In addition, the three-dimensional object may be specified as a non-floating object. Similarly, for example, when the number of high-intensity floating objects is larger than 10 (first score threshold), the three-dimensional object is specified as a floating object, and the number of high-intensity floating objects is -10 points (second score threshold) or less. In this case, the three-dimensional object may be specified as a non-floating object.

  In the above embodiment, the score resetting unit 170 has a feature that the three-dimensional object seems to be a floating object based on the second determination condition when the three-dimensional object is specified as a non-floating object. When judged, the low-intensity floating point number and the high-intensity floating point number are reset to zero. However, for example, when the low-intensity floating point score and the high-intensity floating point number are larger than 10 points (first score threshold), the three-dimensional object is specified as a floating point, and the low-intensity floating point number and the high-intensity floating point number are When the three-dimensional object is specified as a non-floating object in the case of −10 points (second score threshold) or less, the low-luminance floating object score and the high-luminance floating object score are 10 points (first score threshold) and −10 It may be reset to any value between points (second score threshold).

  Note that each step of the vehicle environment recognition method of the present specification does not necessarily have to be processed in time series in the order described in the flowchart, and may include processing in parallel or a subroutine.

  INDUSTRIAL APPLICABILITY The present invention can be used for an external environment recognition device that specifies which specific object corresponds to a three-dimensional object existing in a detection region.

120 Outside environment recognition device 164 Solid object specifying unit 166 Floating object specifying unit 170 Point reset unit

Claims (2)

A three-dimensional object specifying unit for specifying a three-dimensional object based on a predetermined image and deriving a rectangular region including the three-dimensional object;
The luminance value on the solid object side and the luminance value on the opposite side of the solid object are derived based on the boundary with the solid object in the rectangular area, and the luminance value on the solid object side is the opposite. If the threshold value is higher than the predetermined threshold value by a predetermined threshold or more, a floating matter specifying unit that identifies the three-dimensional object as the floating matter,
A vehicle exterior environment recognition device comprising: A three-dimensional object specifying unit for specifying a three-dimensional object based on a predetermined image and deriving a rectangular region including the three-dimensional object;
Dividing the rectangular area into a plurality of divided areas, deriving the luminance value of the peak of the solid object included in each of the divided areas, the luminance value of the peak being equal to or greater than a predetermined luminance threshold If the number is equal to or greater than a predetermined threshold value for the number of divided areas, a floating substance specifying unit that specifies the three-dimensional object as the floating substance;
A vehicle exterior environment recognition device comprising:

Images