JP5547160B2 - Vehicle periphery monitoring device - Google Patents

Description

  The present invention detects an object to be monitored in the vicinity of the vehicle based on an imaging signal output from an imaging device mounted on the vehicle, and issues an alarm according to an attention level of the possibility of contact with the object to be monitored. In particular, the present invention relates to a vehicle periphery monitoring device that is suitable for use when traveling at night or in a dark place.

  Conventionally, as shown in Patent Document 1, an object such as a pedestrian who has a possibility of contact with the own vehicle from an image around the own vehicle (gray scale image and its binary image) captured by an infrared camera. There is known a vehicle periphery monitoring device that detects an object and provides the detected object to a driver of the host vehicle.

  In the vehicle periphery monitoring device according to Patent Document 1, a high-temperature part is detected as an object in an image around the host vehicle captured by a pair of left and right infrared cameras (stereo cameras), and the object in the left and right images is detected. The distance to the object is calculated by obtaining the parallax, and the pedestrians and the like that are likely to affect the traveling of the vehicle (possible contact) from the moving direction of the object and the position of the object An object is detected, an image in front of the vehicle on which the object is imaged is displayed on a HUD (Head Up Display) in front of the driver's seat, and an image portion of the object detected from the image Is highlighted and an alarm is output ([0014], [0015], [0018], [0032] of Patent Document 1).

  Furthermore, an apparatus for displaying an icon for informing the HUD of the presence / absence of an operating state of a pedestrian detection function is also known {(a), (b) in FIG. 6 of Patent Document 2}.

JP 2003-284057 A JP 2004-364112 A

  In the conventional vehicle periphery monitoring device described above, an object such as a pedestrian or an animal having the highest contact possibility attention level is determined in the image and emphasized from other objects, for example, on an image portion of the object. It draws a frame and provides alert information to the driver.

  In this case, when it is determined that the pedestrian and the animal have the same level of attention, conventionally, the pedestrian is always prioritized and alerted.

  By the way, for example, in a place under a surrounding environment such as a highway, a country road, and a mountain road, even if a pedestrian is present, the speed of the traveling vehicle is relatively high. It does not take any action that crosses immediately before, and the frequency of pedestrians is low. On the other hand, there are cases where the appearance frequency of animals is high in the place under the surrounding environmental conditions described above, and because the animals behave irregularly, the attention level (risk level) of the possibility of contact is high. Become.

  However, in the case of the same level of caution, if cautionary information is provided with priority given to pedestrians as in the prior art, for example, places under surrounding environmental conditions such as expressways, country roads, and mountain roads. There is an inconvenience that oversight or discovery of animals that need more attention is delayed.

  The present invention has been made in consideration of such a problem, and when the attention level of the possibility of contact with a pedestrian and an animal is equal, it is possible to make an animal an alarm target depending on the situation. An object of the present invention is to provide a vehicle periphery monitoring device.

  The vehicle periphery monitoring device according to the present invention detects the possibility of contact with a pedestrian and an animal using an image captured by an imaging device mounted on the vehicle, and issues an alarm according to the detected attention level of the contact possibility. In the vehicle periphery monitoring device for determining whether or not to be a target, when the pedestrian and the animal have an equivalent level of attention determined as the alarm target, by default (initial setting), the walking The target of the vehicle and the surrounding environment of the vehicle for the pedestrian and the animal having the same level of caution as the target of the alarm. And an alarm target changing unit that changes the setting of the alarm target from the pedestrian to the animal in consideration of at least one of the above.

  According to this invention, for the pedestrian and the animal having the same level of caution to be determined as an alarm target, the alarm is performed in consideration of at least one of the situation of the host vehicle and the surrounding environment of the host vehicle. Since the setting of the object is changed from the pedestrian to the animal, the animal is preferentially designated as an alarm target in a situation where the appearance frequency of the animal is relatively higher than the appearance frequency of the pedestrian. Therefore, it is possible to set an alarm target according to the actual state of the risk level.

  In this case, if the vehicle speed detected by the vehicle speed sensor is equal to or higher than the threshold vehicle speed as the situation of the host vehicle, the alarm target changing unit regards the vehicle as traveling on a highway, a country road, a mountain road, etc. Under such circumstances, it is preferable to estimate that an animal is more likely to jump out than a pedestrian may jump out, and to change the setting of the alarm target from the pedestrian to the animal.

Further, the imaging device is an infrared camera, and includes a variance value calculation unit that calculates a variance value of pixels constituting the image as a surrounding environment situation of the own vehicle, and the alarm target change unit If serial partial dispersion value is equal to or less than the threshold variance value, it is preferable to change the setting of the alarm object to said animal from the pedestrian. In other words, the dispersion value of the infrared image is usually based on the knowledge that traveling on country roads and mountain roads is lower than traveling on urban roads. Therefore, it is preferable to estimate that the possibility of an animal jumping out is higher than the possibility of a pedestrian jumping out on a country road, mountain road, etc., and changing the setting of the alarm target from the pedestrian to the animal.

  Further, when the number of detections within a predetermined time of an object other than the pedestrian and the animal detected from an image captured by the imaging device as a surrounding environment situation of the host vehicle is a predetermined number of times or less, The setting of the alarm target may be changed from the pedestrian to the animal.

  Furthermore, when the total number of the pedestrians and the animals in the detected image is taken into consideration as the surrounding environment situation of the own vehicle, and the total number is less than a threshold value, Rather, it is estimated that animals are more likely to jump out than pedestrians when traveling on country roads, mountain roads, etc., and the alarm target setting is changed from the pedestrian to the animal. It is preferable to change.

  Furthermore, the navigation device includes a navigation device that detects surrounding environment information of the host vehicle, and the alarm target changing unit includes information on the navigation device, for example, road type information such as a highway, a main road, a mountain road, or an animal jumping out caution. If it is determined from the information that the surrounding environment that is currently running is an environment where there are few pedestrians, it is estimated that there is a higher possibility of animals jumping out than the possibility of pedestrians jumping out, and setting of the alarm target Is preferably changed from the pedestrian to the animal.

  In addition, when the imaging device is a visible region imaging device, the alarm target changing unit may cause an animal to jump out of the image of the visible region imaging device as the surrounding environment information of the host vehicle. When a warning sign is detected, it is presumed that the possibility that an animal will pop out is higher than the possibility that a pedestrian will pop out, and the setting of the alarm target is preferably changed from the pedestrian to the animal.

  Even if the imaging device is an infrared imaging device, the warning sign can be detected when the outline of the animal formed on the warning sign emits light with a light emitting element or the like.

  Furthermore, the alarm system further includes a duration measuring unit that counts a duration from the time when the determination is made when the setting of the alarm target is changed from the pedestrian to the animal, and the duration from the time when the judgment is established. If the determination is continued for more than a threshold time, even if the determination is interrupted for a certain period of time, at the time of interruption (ignoring the interruption), the animal after changing the setting of the alarm target It is preferable to hold it at

  According to this invention, for the pedestrian and the animal having the same level of caution to be determined as an alarm target, the alarm is performed in consideration of at least one of the situation of the host vehicle and the surrounding environment of the host vehicle. Since the setting of the object is changed from the pedestrian to the animal, the animal is preferentially designated as an alarm target in a situation where the appearance frequency of the animal is relatively higher than the appearance frequency of the pedestrian. Since it can be estimated, the effect that the alarm target can be set according to the actual state of the risk is achieved.

It is a block diagram which shows the structure of the vehicle periphery monitoring apparatus which concerns on one Embodiment of this invention. It is a schematic perspective view of the vehicle carrying the vehicle periphery monitoring apparatus shown in FIG. FIG. 3 is an interior view of the vehicle shown in FIG. 2 viewed from the driver side. FIG. 4A is a front view of the first panel portion of the general-purpose display. FIG. 4B is a front view of the second panel portion of the MID. It is a flowchart with which operation | movement description of the vehicle periphery monitoring apparatus shown in FIG. 1 is provided. It is a schematic diagram of the image of a pedestrian used for description of a pedestrian detection process. It is a schematic diagram of the image of the animal used for description of an animal detection process. It is a schematic diagram of the gray scale image at the time of city driving | running | working memorize | stored in memory. It is explanatory drawing of the example of a display of the pedestrian icon to a 2nd panel. It is explanatory drawing of the example of a display of the gray scale image shown in FIG. 8 on the 1st panel. It is a schematic diagram of the gray scale image at the time of country road driving | running | working memorize | stored in memory. It is explanatory drawing of the example of a display of the animal icon to a 2nd panel. It is explanatory drawing of the example of a display of the gray scale image shown in FIG. 11 to the 1st panel.

  Embodiments of the present invention will be described below with reference to the accompanying drawings.

  FIG. 1 is a block diagram showing a configuration of a vehicle periphery monitoring apparatus 10 according to an embodiment of the present invention. FIG. 2 is a schematic perspective view of a vehicle (also referred to as a host vehicle) 12 on which the vehicle periphery monitoring device 10 shown in FIG. 1 is mounted. FIG. 3 is an interior view of the vehicle 12 shown in FIG. 2 viewed from the driver side. In addition, this vehicle 12 has shown the condition which is drive | working the road of the country in which the motor vehicle has decided to drive on the right side, and the vehicle 12 is illustrated as a left-hand drive vehicle.

  As shown in FIGS. 1 and 2, the vehicle periphery monitoring device 10 includes an image processing unit 14 that controls the vehicle periphery monitoring device 10, and left and right infrared cameras 16R and 16L (imaging devices) connected to the image processing unit 14. ), A vehicle speed sensor 18 for detecting the vehicle speed Vs of the vehicle 12, a brake sensor 20 for detecting a brake pedal operation amount (hereinafter simply referred to as a brake operation amount Br) by the driver, and a yaw rate Yr of the vehicle 12 are detected. A yaw rate sensor 22, a speaker 24 for issuing a warning by voice, a general-purpose display 26 (first display unit; first panel unit 26p, etc.) for displaying an image based on imaging signals output from the infrared cameras 16R and 16L And MID (Multi-Information) that visualizes and displays accompanying information when driving the vehicle 12 Comprising a comprising a second panel portion 28p, etc.); Display) 28 (second display unit.. As a general-purpose display 26 that displays an image (video) based on the imaging signals output from the infrared cameras 16R and 16L, a display screen is provided on the front windshield of the vehicle 12 at a position that does not obstruct the driver's front view. May be substituted. The HUD may be provided together with the general-purpose display 26.

  The image processing unit 14 uses the infrared image around the vehicle 12 and a signal (here, the vehicle speed Vs, the brake operation amount Br, and the yaw rate Yr) indicating the traveling state (vehicle state) of the vehicle, When an object to be monitored such as an animal is detected and it is determined that the possibility of contact with the object to be monitored (attention level) is high, an alarm (for example, a beeping sound) is emitted from the speaker 24. At the same time, an icon (a pedestrian icon or an animal icon as will be described later in this embodiment) representing the monitoring object that is determined to have a high level of attention to the possibility of contact is displayed on the MID 28, and on the general-purpose display 26. The monitored object in the captured image displayed in gray scale is displayed surrounded by a conspicuous color frame such as yellow or red. In this way, the driver's attention is drawn and driving is supported.

  Here, the image processing unit 14 includes an input circuit such as an A / D conversion circuit (not shown) that converts an input analog signal into a digital signal, a CPU 14c (central processing unit) that executes various arithmetic processes, and image processing. An output circuit (not shown) that outputs a storage unit 14m for storing various data provided, a clock (clock unit) and a timer (timer unit), a driving signal for the speaker 24, and a display signal for the general-purpose display 26 and the MID 28. And further comprising.

  The storage unit 14m is a RAM (Random Access Memory) that stores imaging signals (grayscale image signals and their binarized signals) converted into digital signals, temporary data used for various arithmetic processes, or an execution program. , Tables, maps, and templates {pedestrian (human body) shape templates: right, left, front (rear) orientation, animal shape templates: deer, dogs, etc., right, left, front (rear) orientation, animals in warning signs Artificial structure shape templates: vehicle shape features, utility pole shape features, road / street light features, wall features, etc.}, etc. are stored in a ROM (Read Only Memory) or the like.

  The output signals of the infrared cameras 16R and 16L, the vehicle speed sensor 18, the brake sensor 20, and the yaw rate sensor 22 are configured to be supplied to the CPU 14c as digital signals.

  The CPU 14c captures these digital signals and executes programs while referring to tables, maps, templates, and the like, thereby functioning as various functional means (also referred to as functional units), and to the speaker 24, the MID 28, and the general-purpose display 26. Drive signals (audio signals and display signals) are sent out. These functions can also be realized by hardware.

  In this embodiment, the image processing unit 14 is also referred to as an object detection processing unit 38, a contact possibility attention level determination unit (a contact possibility determination unit or a attention level determination unit), which will be described in detail later, as the various functional units. 42), alarm target setting unit 44, alarm target changing unit 46, and the like. The object detection processing unit 38 includes a pedestrian detection processing unit 39, an animal detection processing unit 40, and an artificial structure detection processing unit 41. The alarm target changing unit 46 also measures a variance value calculating unit 47 that calculates a variance value of pixel values constituting the image, an animal warning sign detecting unit 48, and a determination establishment duration time and an interruption time of a determination condition described later. The determination establishment continuation time counting unit 49 is provided.

  As shown in FIG. 2, the infrared cameras 16R and 16L, which are imaging devices and function as stereo cameras, are arranged on the front bumper portion of the vehicle 12 at a position that is substantially symmetrical with respect to the center portion in the vehicle width direction of the vehicle 12. ing. The two infrared cameras 16R and 16L are fixed so that the optical axes thereof are parallel to each other and the heights from both road surfaces are equal. Note that the infrared cameras 16R and 16L have a characteristic that the output signal level becomes higher (the luminance increases) as the temperature of the monitored object is higher.

  The imaging device is not limited to the configuration example shown in FIG. 2 and can adopt various configurations. For example, the imaging means may be a compound eye camera (stereo camera) or a monocular camera (one camera). In the case of a monocular camera, it is preferable that another ranging means (radar device) is also provided. A digital color camera (hereinafter referred to as a color camera) 16C that mainly uses light having a wavelength in the visible light region as shown by a two-dot chain line block in FIG. It may be used.

  As shown in FIGS. 2 and 3, the dashboard 30 is provided with a handle 32 that protrudes from a lower portion thereof toward a driver seat (not shown). The second panel portion 28p of the MID 28 is disposed at a position aligned with the center position of the handle 32 and in the dashboard 30 (or in the instrument panel 34) near the upper side of the instrument panel 34. . As a result, the driver can view the second panel portion 28p of the MID 28 from the front with the face facing forward of the vehicle 12. The first panel portion 26p of the general-purpose display 26 is disposed at a predetermined portion of the dashboard 30 (specifically, a portion on the right side with respect to the position where the first panel portion 26p is disposed).

  4A is a front view of the first panel portion 26p, and FIG. 4B is a front view of the second panel portion 28p.

  A substantially rectangular first display region 70 that is long in the horizontal direction is provided on substantially the entire surface of the first panel portion 26p shown in FIG. 4A. 4A shows a state in which the first image 72 based on the imaging signal output from the infrared camera 16L is displayed in the first display area 70. FIG. The first panel unit 26p is a module capable of displaying a color image or a monochrome image, and may be configured of, for example, a liquid crystal panel, an organic EL (Electro-Luminescence), an inorganic EL panel, or the like. In this case, the general-purpose display 26 is not limited to an arbitrary image, for example, the first image 72 around the vehicle 12 via the first panel unit 26p, but various images for car navigation (road maps, service information, etc.) Video content and the like can be displayed.

  A substantially rectangular second display region 74 that is long in the horizontal direction is provided on substantially the entire surface of the second panel portion 28p shown in FIG. 4B. This figure shows a state in which a second image 76 in which only a predetermined feature portion in the first image 72 is extracted and appropriately deformed is displayed in the second display area 74. As a specific example, a road icon 78 is displayed below the second display area 74. The road icon 78 includes three lines and lines 77L, 77C, and 77R in order from the left. The width between the lines 77L and 77R is narrower toward the upper side of the second display area 74 that recalls a distant position. That is, the display of the road icon 78 displays the left road side (line 77L), the center line (line 77C), and the right road side (line 77R) according to the road shape when the straight road ahead is viewed from the driver's seat. Can be thought of.

  The second panel unit 28p is a display module that has a simpler configuration and is less expensive than the first panel unit 26p. In addition to the second image 76 (see FIGS. 9 and 12) described later, the MID 28 can display fuel consumption, current time, information on the instrument panel 34, and the like.

  The vehicle periphery monitoring device 10 according to this embodiment is basically configured as described above. Subsequently, the operation of the vehicle periphery monitoring device 10 will be described along the flowchart of FIG. 5 with reference to other drawings as appropriate.

  First, in step S1, the image processing unit 14 determines the traveling state (running or stopped) of the vehicle 12 from the vehicle speed Vs detected by the vehicle speed sensor 18, and the vehicle is stopped (step S1: NO). The process is stopped.

  When the vehicle is running (step S1: YES), in step S2, the image processing unit 14 is an infrared signal that is an output signal for each frame in a predetermined angle of view range in front of the vehicle captured by the infrared cameras 16R and 16L for each frame. Images (grayscale right image and grayscale left image) are acquired, A / D converted, each grayscale image is stored in the image memory (storage unit 14m), and binarization processing of the stored grayscale image; That is, a conversion process to a binary image in which an area brighter than the luminance threshold is set to “1” (white) and a dark area is set to “0” (black) is performed, and the converted binary image also corresponds to the frame. Each time it is stored in the storage unit 14m. In this binarization process, in the human body, a lump (human body object candidate) including the head, shoulders, torso, and two legs is detected as a lump (aggregate) of “1”. Similarly, in animals such as dogs (small animals) and deer (large animals) (four-legged animals in this embodiment), a lump (animal object candidate) such as the head, torso, tail, and four legs. ) Is detected as a lump (aggregate) of “1”. In the gray scale right image and the gray scale left image, the horizontal position of the same object on the display screen is displayed with a shift, so this shift (parallax) or the like is used to determine the object up to the object by the principle of triangulation. The distance can be calculated.

  Next, in step S3, the image processing unit 14 performs a labeling process. In this labeling process, the “1” chunks constituting the human body object candidate and the “1” chunks constituting the animal target candidate are converted into run-length data for each scan line in the x direction (horizontal direction). , A line with a portion overlapping in the y direction (vertical direction) is regarded as one object, and a so-called labeling process is performed in which each circumscribed rectangle of the object is labeled. In addition, after the process of step S2 and step S3, identification (detection) whether it is a human body or an animal is not yet made.

  Next, in step S4, the pedestrian detection processing unit 39 configuring the target object detection processing unit 38 executes a pedestrian detection process as the target object.

  The pedestrian detection processing unit 39 uses a mask region for a mask region 53 (a region surrounded by an alternate long and short dash line in FIG. 6) that is slightly larger than the labeled object candidate region 52 shown in FIG. If the pixel value in 53 is searched while scanning the pixel value from the upper side to the lower side and from the left side to the right side, and “0” pixels in the dark region continue continuously, the scanned portion is displayed in the image. It is determined as a boundary between the pedestrian candidate 62 (consisting of the head portion 54, the body portion 60, and the leg portion 59) and the road surface 51, and is defined as the lower end OBbm of the object.

  In step S4, the pedestrian detection processing unit 39 searches the image in the mask area 53 on the upper side from the lower end OBbm of the target object while scanning pixel by pixel from left to right. When a change section of the horizontal edge 54 of luminance in the direction (a section in which a pair of “1” and “0” is substantially continuous in the binary image) is detected, the scanning part of the change section is the pedestrian candidate 62 and the background. The upper end OBtm of the object that is the edge of the boundary. In this case, whether or not the candidate is a pedestrian 62 depends on the presence of the head, the torso under the head, the two legs under the torso, and the height Hm being within a predetermined range. It is determined in a known manner.

  Here, for convenience of understanding, the animal detection process executed by the animal detection processing unit 40 in step S7 will be described.

  The animal detection processing unit 40 masks the mask region 153 with respect to the mask region 153 (the region surrounded by the one-dot chain line in FIG. 7) that is slightly larger than the object candidate region 152 labeled in FIG. If the search is performed while scanning pixel values from the upper side to the lower side and the left side to the right side of the pixels in the pixel, and “0” pixels in the dark region continue, the scanned part is the animal in the image. It is determined as a boundary between the candidate 162 (consisting of the head portion 154, the body portion 160, and the leg portion 159) and the road surface 51, and is defined as the lower end OBba of the object.

  Similarly, in step S7, the animal detection processing unit 40 searches while scanning pixel by pixel from the left side to the right side with respect to the image in the mask area 153 upward from the lower end OBba of the object. When a change interval of the horizontal edge 154 of luminance in the vertical direction (in the binary image, a interval in which a pair of “1” and “0” is substantially continuous) is detected, the scanning portion of the change interval is the animal candidate 162 and the background. The upper end OBta of the object that is the edge of the boundary.

  It should be noted that there are various methods for determining whether the pedestrian candidate 62 (FIG. 6) or the animal candidate 162 (FIG. 7) is used. For example, the height of the pedestrian Hm (Hm = OBtm−OBbm). : The unit is, for example, the pedestrian candidate 62 when the number of pixels is higher than the height Ha (Ha = OBta-OBba: the unit is the number of pixels) of an animal (especially a dog). . Moreover, it can determine by comparing with each threshold value by ratio of trunk | body height bhm and bha and head height hhm and hha. Usually, the ratio hm / bhm (= head height / body height) of a pedestrian (human body) is smaller than the ratio hha / bha (= head height / body height) of an animal such as a dog. Further, when four legs (legs 159) are detected, it can be an animal candidate 162, and when two legs (legs 59) are detected, it can be a pedestrian candidate 62. If the horizontal orientation can be determined, it can be determined by the ratio of the head widths hwm and hwa to the total widths Wm and Wa (hwm / Wm = head width / body width or hwa / Wa = head width / total width).

  In step S5 following the pedestrian detection process in step S4, it is determined whether or not the pedestrian candidate 62 has been detected in step S4. If the pedestrian candidate 62 has not been detected, the animal detection process in step S7 described above is performed. If it has been detected, however, the contact possibility attention level determination unit 42 performs pedestrian priority determination processing in step S6. In this pedestrian priority determination process, the pedestrian candidate 62 with the highest level of attention among the detected pedestrian candidates 62 is specified.

  Specifically, for example, the distance to each pedestrian candidate 62 and the movement vector of each pedestrian candidate 62 are calculated, and further, brake operation amounts Br that are outputs of the brake sensor 20, the vehicle speed sensor 18, and the yaw rate sensor 22. Based on the vehicle speed Vs and the yaw rate Yr, a contact margin time TTC (Time To Contact or Time To Collision) is calculated, and the vehicle 12 is the pedestrian candidate 62 ( The pedestrian candidate 62) with the shortest contact allowance time TTC is determined as the pedestrian candidate 62 with the highest level of attention (highest priority). The contact allowance time TTC between the host vehicle 12 and the pedestrian candidate 62 is referred to as contact allowance time TTCp.

  Next, in step S7, the above-described animal detection process is executed. In step S8, it is determined whether or not the animal candidate 162 has been detected in step S7. In step S9, the contact possibility level determination unit 42 performs an animal priority determination process. In this animal priority determination process, as in step S6, the animal candidate 162 with the highest attention level among the detected animal candidates 162 is specified.

  Specifically, for example, the distance to each animal candidate 162 and the movement vector of each animal candidate 162 are calculated, and further, the brake operation amount Br, the vehicle speed, which are the outputs of the brake sensor 20, the vehicle speed sensor 18, and the yaw rate sensor 22, are calculated. Based on Vs and the yaw rate Yr, a contact allowance time TTC (the contact allowance time TTC with the animal candidate 162 is referred to as a contact allowance time TTCa) is calculated, and the vehicle 12 has the highest possibility of contact among the animal candidates 162. The animal candidate 162 with the highest level of attention (the highest priority) is determined as the animal candidate 162 with the highest attention level (the animal candidate 162 with the shortest contact allowance time TTCa).

  Next, in step S10, the alarm target setting unit 44 has the highest possibility of contact detected in step S9 and the pedestrian candidate 62 having the highest possibility of contact (high attention level) detected in step S6 ( An animal candidate 162 having a high attention level is selected, and it is determined whether or not the attention (priority) level is the same (equivalent). Specifically, when the difference between the contact allowance time TTCp of the pedestrian candidate 62 with the highest attention level and the contact allowance time TTCa of the animal candidate 162 with the highest attention level is within the threshold time ΔTth (| TTCp−TTCa | ≦ ΔTth), it is determined that they are at the same attention (priority) level.

  When the pedestrian candidate 62 and the animal candidate 162 having the same attention level exist (step S10: YES), the alarm target changing unit 46 prioritizes the pedestrian candidate 62 and the animal candidate 162 having the same attention level in step S11. Degree determination processing (pedestrian / animal priority determination processing) is performed. In this priority determination process, any candidate among the pedestrian candidate 62 and the animal candidate 162 of the same attention level is subjected to an alarm enhancement process or the like {in this specification, the alarm is performed by the speaker 24. Image output by alarm output and / or image highlighting is included. } Is determined.

  In the priority determination process executed by the alarm target changing unit 46, the situation of the host vehicle 12 and the surrounding environment of the host vehicle 12 are compared to the pedestrian candidate 62 and the animal candidate 162 having the same level of caution as the alarm target. In consideration of at least one of the situations, a process for determining whether or not to change the setting of the alarm target from the default (initial setting) pedestrian candidate 62 to the animal candidate 162 is performed.

Specifically, if any one of the first to seventh determination conditions described below is met, the alarm target setting at the same attention level is set as a default pedestrian candidate. 62 to change to animal candidate 162 from.

  The first determination condition is that when the vehicle speed Vs detected by the vehicle speed sensor 18 is equal to or higher than the threshold vehicle speed Vsth (Vs ≧ Vsth) as the situation of the host vehicle 12, the road on which the host vehicle 12 is traveling is an expressway, countryside It is assumed that the road is a road or a mountain road, and it is estimated that there is a higher possibility that an animal will jump out than the possibility that a pedestrian will jump out, and an alarm target for the pedestrian candidate 62 and the animal candidate 162 having the same attention level Is changed from the pedestrian candidate 62 to the animal candidate 162.

  The second determination condition is calculated by the variance value calculation unit 47 that calculates the variance value Var of the pixel values constituting the image (grayscale image) captured by the infrared camera 16L as the surrounding environment state of the host vehicle 12. When the variance value Var of the image is equal to or less than the threshold variance value Varth (Var ≦ Varth), it is considered that the road on which the vehicle 12 is traveling is not a city road but a country road, a mountain road, or the like. Estimating that there is a higher possibility of animals popping out than popping out, and changing the alarm target setting for pedestrian candidates 62 and animal candidates 162 having the same level of caution from pedestrian candidates 62 to animal candidates 162 To do.

  The third determination condition is that the surrounding environment of the host vehicle 12 is a specific object other than the pedestrian candidate 62 and the animal candidate 162 from the image captured by the infrared camera 16L, for example, a vehicle, a wall, a road light / a streetlight. When the number of detections within a predetermined time for an artificial structure such as a utility pole is less than or equal to the predetermined number of times, it is considered that the road on which the vehicle 12 is traveling is not a city road but a country road or a mountain road. Estimating that there is a higher possibility of an animal jumping out than the possibility of the pedestrian jumping out, and setting the alarm target for the pedestrian candidate 62 and the animal candidate 162 having the same level of caution from the pedestrian candidate 62 to the animal. Change to candidate 162. In addition, what is necessary is just to implement the artificial structure detection process by the artificial structure detection process part 41 just before step S10 process, for example.

  The fourth determination condition refers to the total number σ of the pedestrian candidates 62 and the animal candidates 162 in the image detected by the pedestrian detection processing unit 39 and the animal detection processing unit 40 as the surrounding environment state of the host vehicle 12. However, when the total number σ is less than the threshold σth (σ ≦ σth), it is considered that the road on which the vehicle 12 is traveling is not a city road, but a country road or a mountain road, and a pedestrian jumps out. Estimating that the possibility of an animal jumping out is higher than the possibility of coming in, and changing the setting of the alarm target for the pedestrian candidate 62 and the animal candidate 162 having the same attention level from the pedestrian candidate 62 to the animal candidate 162 .

  The fifth determination condition is that information from the navigation device 25 that provides the surrounding environment information of the position or traveling direction of the host vehicle 12, for example, road type information such as an expressway, a main road, a mountain road, or animal jumping out caution information. When detected, the surrounding environment that is currently running is considered to be an environment with few pedestrians, and it is estimated that animals are more likely to jump out than pedestrians pop out. The alarm target setting for the pedestrian candidate 62 and the animal candidate 162 is changed from the pedestrian candidate 62 to the animal candidate 162.

  The sixth determination condition is that when the animal warning sign detection unit 48 detects a warning sign that the animal may jump out by template matching or the like from the image captured by the infrared camera 16L, a pedestrian jumps out. Estimating that there is a higher possibility of an animal jumping out than the possibility of coming, setting the alarm target for the pedestrian candidate 62 and the animal candidate 162 having the same attention level, and setting the alarm target from the pedestrian candidate 62 to the animal candidate Change to 162. When a color camera 16C as a visible region imaging device is mounted as the imaging device, the animal warning sign detection unit 48 uses an edge extraction, template matching, or the like from a color visible image by the color camera 16C. If a warning sign indicating that there is a possibility of jumping out is detected, it is estimated that there is a higher possibility that the animal will jump out than the possibility that the pedestrian will jump out, and the pedestrian candidate 62 having the same attention level The alarm target setting for the animal candidate 162 may be changed from the pedestrian candidate 62 to the animal candidate 162.

  The seventh determination condition is that the establishment time (continuation time) of the first to fourth determination conditions or the sixth determination condition, excluding the fifth determination condition, has continued for a threshold duration or longer. When the time is counted by the pedestrian, even if the time less than the threshold interruption time, the first to fourth determination conditions, or the continuous establishment time (continuation time) of the sixth determination condition is interrupted, It is preferable to keep the state changed from the candidate 62 to the animal candidate 162. For example, the vehicle speed Vs according to the first determination condition is normally (basically) on country roads and mountain roads, and is running at a high vehicle speed of Vs ≧ Vsth, but temporarily due to up / down roads, curves, etc. The vehicle speed Vs may be lower than the threshold vehicle speed as Vs <Vsth. However, in such a situation, it is still the case that the vehicle is traveling on a country road or a mountain road. This is because priority is given to the candidate 162.

  After the above-described pedestrian / animal priority determination process in step S11 or when the determination in step S10 is negative, emphasis of the highest priority alarm target (highest priority object) detected in step S12 Process.

  Here, a specific example of the emphasis process of the highest priority alarm target when the pedestrian candidate 62 and the animal candidate 162 having the same attention level exist in the image will be described.

  FIG. 8 shows a display of a grayscale image 200 captured by the infrared camera 16L, which is captured as a digital signal in the storage unit 14m.

  In the gray scale image 200 of this urban area, there are an animal candidate 162 crossing the road Rd and a pedestrian candidate 62 crossing the animal candidate 162 while walking.

  In this case, it is determined that the attention level of the possibility of contact with the pedestrian candidate 62 and the animal candidate 162 that exist in the approximate center of the road Rd is the same, but the alarm target setting is changed from the default pedestrian candidate 62 to the animal candidate 162. Since none of the above-described first to sixth determination conditions to be changed applies, the alarm target is set to the default pedestrian candidate 62. That is, in urban areas, relatively large animals such as dogs are not stray dogs, and many animals are being walked, and such walking animals are accessible to pedestrians because their behavior is restricted by reeds, etc. Since the nature (risk level) can be said to be lower than that of pedestrians, the alarm target is set as the pedestrian candidate 62.

  In this case, the display on the second panel portion 28p of the MID 28 is switched from the normal display such as fuel consumption to the second image 76 of the alarm display shown in FIG. In this case, the image processing unit 14 (which also serves as an emphasis processing unit for the highest priority object) serves as a central region of the road Rd in the gray scale image 200 (a region of the center 1/3 of the gray scale image 200). Corresponding to the position of the existing pedestrian candidate 62 (FIG. 8), the pedestrian icon Hi is displayed in a prominent color (here yellow) at the position (center position) of the line 77C, and an alarm is generated through the speaker 24. And alert the driver. Note that the colors of the lines 77L, 77C, and 77R are white. The background color of the second image 76 is black.

  If the position of the pedestrian candidate 62 is not in the central area of the road Rd but in the right area of the road Rd (the area on the right third of the grayscale image 200), the pedestrian icon Hi is displayed above the line 77R. When the position of the pedestrian candidate 62 is present in the left region of the road Rd (the left one-third region of the gray scale image 200) in the gray scale image 200, the user walks above the line 77L. A person icon Hi is displayed.

  Furthermore, simultaneously with the alarm display of the pedestrian icon Hi on the MID 28, the first image 72 of the first panel unit 26p constituting the general-purpose display 26 of the navigation device 25 is displayed from a normal display such as a map as shown in FIG. In addition, the display is switched to the alarm display (highlighted display) of the grayscale image (infrared video) 200A corresponding to the grayscale image 200 shown in FIG. At this time, the pedestrian candidate 62 is displayed by being surrounded by a red frame 80 that is a conspicuous color that alerts the driver that the highest degree of danger is present, and the animal candidate 162 is a yellow frame 82 that is a conspicuous color that alerts the driver. It is displayed surrounded by.

  Next, as shown in FIG. 11, a case will be described in which the grayscale image 202 captured as a digital signal in the storage unit 14m is a grayscale image obtained by capturing the “country road” with the infrared camera 16L.

  In the gray scale image 202 of the country road, there are an animal candidate 162 crossing the road Rd and a pedestrian candidate 62.

  In this case, the attention level of the possibility of contact with the candidate animal 162 that exists in the approximate center of the road Rd and the pedestrian candidate 62 that exists on the right side of the road Rd is determined to be equivalent, but the first to sixth determinations described above are used. Since at least one of the conditions is met, the alarm target changing unit 46 changes the alarm target setting from the default pedestrian candidate 62 to the animal candidate 162. As described above, on country roads, highways, etc., animals are large and small, and there are many wild animals (wild animals). On the other hand, pedestrians are restricted in behavior because the vehicle speed of the traveling vehicle 12 is relatively high, and it can be said that the caution level of contact possibility, in other words, the danger level is low. The setting is an animal candidate 162.

  In this case, the display on the second panel portion 28p of the MID 28 is switched from the normal display such as fuel consumption to the second image 76 of the alarm display shown in FIG. The image processing unit 14 (which also serves as an emphasis processing unit for the highest-priority object) is an animal that exists in the central region of the road Rd (the central one-third region of the grayscale image 200) in the grayscale image 202. In correspondence with the position of the candidate 162 (FIG. 11), the animal icon Ai is displayed in a conspicuous color (yellow here) at the position (center position) of the line 77C, and an alarm is generated through the speaker 24 to Call attention.

  Further, simultaneously with the alarm display of the animal icon Ai on the MID 28, the first image 72 of the first panel unit 26p constituting the general-purpose display 26 of the navigation device 25 is displayed from the normal display such as a map as shown in FIG. Then, the alarm display (highlighted display) of the gray scale image (infrared video) 202A corresponding to the gray scale image 202 shown in FIG. 11 is switched. At this time, the animal candidate 162 is displayed by being surrounded by a red frame 80 that is a conspicuous color that alerts the driver that the degree of danger is high, and the pedestrian candidate 62 is surrounded by a yellow frame 82 that is a conspicuous color that calls attention. To display.

[Outline of Embodiment]
As described above, the vehicle periphery monitoring device 10 according to the above-described embodiment can make contact with pedestrians and animals using images captured by the infrared cameras 16R and 16L that are imaging devices mounted on the vehicle 12. Whether or not to be alarmed is determined according to the detected attention level of the contact possibility.

  In this case, the vehicle periphery monitoring apparatus 10 has an equivalent attention level that the pedestrian candidate 62 and the animal candidate 162 in the grayscale images 200 and 202 stored in the storage unit 14m determine as the alarm target. In this case, by default, the alarm target setting unit 44 that sets the pedestrian candidate 62 as the alarm target, and the pedestrian candidate 62 and the animal candidate 162 that have the same level of attention as the alarm target are determined. An alarm target changing unit 46 that changes the setting of the alarm target from the pedestrian candidate 62 to the animal candidate 162 in consideration of at least one of the situation of the own vehicle 12 and the surrounding environment situation of the own vehicle 12.

  In this way, at least one of the situation of the own vehicle 12 and the surrounding environment situation of the own vehicle 12 is taken into consideration for the pedestrian candidate 62 and the animal candidate 162 having the same level of caution as the alarm target. Since the setting of the alarm target is changed from the pedestrian candidate 62 to the animal candidate 162, the animal is preferentially used in a situation where the appearance frequency of the animal is relatively higher than the appearance frequency of the pedestrian. The alarm target can be set, and the alarm target can be set in accordance with the actual state of the danger level.

  In this case, as the situation of the host vehicle 12, when the vehicle speed Vs detected by the vehicle speed sensor 18 is equal to or higher than the threshold vehicle speed Vsth, it is considered that the vehicle is traveling on a highway, a country road, a mountain road, etc. 1), under such circumstances, it is estimated that there is a higher possibility that an animal will jump out than the possibility that a pedestrian will jump out, and the alarm target setting is changed from the pedestrian candidate 62 to the animal candidate 162. It is preferable to change. Note that the threshold vehicle speed Vsth is, for example, an expressway according to the known road type when the type of road (highway or country road) is known from information such as the navigation device 25 or roadside broadcasting. Then, the threshold vehicle speed Vsth may be set higher than the country road.

  In addition, as a surrounding environment situation of the host vehicle 12, a variance value calculation unit 47 that calculates a variance value Var of the grayscale images 200 and 202 captured by the infrared camera 16L is provided, and the alarm target change unit 46 is calculated. When the variance value Var is equal to or less than the threshold variance value Varth (when the second determination condition described above is satisfied), it is preferable to change the setting of the alarm target from the pedestrian candidate 62 to the animal candidate 162. That is, the variance value Var of the grayscale images 200 and 202 that are infrared images is lower when traveling on country roads (see FIG. 11) and when traveling on mountain roads or the like than when traveling on urban roads (see FIG. 8). In country roads, mountain roads, and the like, it is preferable to change the setting of the alarm target from the pedestrian candidate 62 to the animal candidate 162, assuming that the possibility that the animal will jump out is higher than the possibility that the pedestrian jumps out. The threshold variance value Varth may also be changed between, for example, dusk and midnight.

  Further, the surrounding environment of the host vehicle 12 includes a specific object other than the pedestrian candidate 62 and the animal candidate 162 from the grayscale images 200 and 202 captured by the infrared camera 16L, for example, artificial structures such as road lights and street lights. When the number of detections for an object within a predetermined time is less than the predetermined number, it is considered as a country road, etc., and it is estimated that an animal is more likely to jump out than a pedestrian may jump out. It is preferable to change the setting from the pedestrian candidate 62 to the animal candidate 162.

  Furthermore, when the surrounding environment state of the host vehicle 12 takes into account the total number σ of the pedestrian candidates 62 and the animal candidates 162 in the detected grayscale images 200 and 202, and the total number σ is as small as a threshold σth or less. When the above-mentioned third determination condition is satisfied, it is not when driving on a city road, but when driving on a road such as a country road or mountain road, and an animal may jump out due to the possibility of a pedestrian jumping out. It is preferable to change the setting of the alarm target from the pedestrian candidate 62 to the animal candidate 162. The threshold value σth of the total number σ may also be changed, for example, according to road congestion.

  Further, the surrounding environment currently being traveled is determined by information from the navigation device 25 that detects the surrounding environment information of the host vehicle 12, for example, road type information such as highways, main roads, mountain roads, or animal jumping out caution information. If it is determined that the environment is low (when the above-described fourth determination condition is satisfied), it is estimated that there is a higher possibility that the animal will jump out than the possibility that the pedestrian will jump out. It is preferable to change from the pedestrian candidate 62 to the animal candidate 162.

  In addition, when the color camera 16C as a visible region imaging device that images the front of the road is mounted, there is a possibility that an animal may jump out of the color image captured by the color camera 16C as the surrounding environment information of the host vehicle 12. When the warning sign is detected by edge extraction processing, pattern matching processing, or the like (when the above-described fifth determination condition is satisfied), the possibility that the animal will jump out is higher than the possibility that the pedestrian jumps out. It is preferable to change the setting of the alarm target from the pedestrian candidate 62 to the animal candidate 162. Note that, under certain conditions, a warning sign indicating that there is a possibility that the animal may jump out by the infrared camera 16L can be detected.

  Furthermore, a determination establishment duration measuring unit that measures the duration (continuous establishment time) of the time established from the time when the first to fourth determination conditions, the sixth determination condition, or the seventh determination condition is satisfied. 49, when the continuation time has continued for more than the threshold continuation time, a continuous continuation flag is set, and when the continuous continuation flag is lowered, a predetermined hysteresis time (fixed time) is provided, As long as the first to third determination conditions or the fifth determination condition is not satisfied, the alarm target setting is changed from the pedestrian candidate 62 to the animal candidate 162 even when the continuous establishment time is interrupted. It is preferable to keep the changed state. The hysteresis time may also be changed according to road type information.

  Further, in the above-described embodiment, the MID 28 in the instrument panel 34 on which the driver's attention is paid on average for a long time together with the alarm, the monitoring object having the highest contact possibility attention level (highest priority). The icon (pedestrian icon Hi or animal icon Ai) is displayed on the left side, in the center, or on the right side of the road so that the driver can see at a glance. It is possible to grasp the presence of the monitoring object and whether the monitoring object is a pedestrian or an animal. In addition to displaying icons in the MID 28, the monitoring object is displayed in a gray scale image 200, 202 on the general-purpose display 26 surrounded by an emphasis frame (red frame 80, etc.). By referring to the gray scale images 200 and 202, it is possible to visually recognize an actual pedestrian or an actual animal.

  In addition, this invention is not limited to embodiment mentioned above, Of course, it can change freely in the range which does not deviate from the main point of this invention.

10 ... Vehicle periphery monitoring device 12 ... Vehicle (own vehicle)
DESCRIPTION OF SYMBOLS 14 ... Image processing unit 16L, 16R ... Infrared camera 16C ... Color camera 26 ... General-purpose display 28 ... MID 44 ... Alarm object setting part 46 ... Alarm object change part

Claims (8)

Whether the contact possibility based on the contact allowance time with respect to the pedestrian and the animal is detected using an image picked up by the image pickup device mounted on the vehicle, and whether or not to be a warning target according to the detected attention level of the contact possibility In the vehicle periphery monitoring device for determining whether
When the pedestrian and the animal in the captured image have the same level of caution as determining the candidate for the alarm target, by default, the alarm target setting that sets the pedestrian as the alarm target And
Considering at least one of the situation of the host vehicle and the surrounding environment of the host vehicle with respect to the pedestrian and the animal having the same attention level as the candidate for the alarm target, the pedestrian When it is estimated that the animal is more likely to jump out onto the traveling road of the host vehicle than the possibility of jumping out onto the traveling road of the own vehicle, the default setting is canceled and the alarm target setting is changed. An alarm target changing unit for changing from the pedestrian to the animal;
A vehicle periphery monitoring device comprising: The vehicle periphery monitoring device according to claim 1,
As a situation of the host vehicle, a vehicle speed sensor that detects the vehicle speed of the host vehicle,
When the vehicle speed detected by the vehicle speed sensor is greater than or equal to a threshold vehicle speed, the alarm target changing unit changes the setting of the alarm target from the pedestrian to the animal. . The vehicle periphery monitoring device according to claim 1,
The imaging device is an infrared camera;
A dispersion value calculation unit that calculates a dispersion value of pixel values constituting the image as a surrounding environment situation of the host vehicle,
The alarm target changing unit is
If the calculated pre-SL component dispersion value is equal to or less than the threshold variance value, the vehicle periphery monitoring apparatus characterized by changing the setting of the alarm object to said animal from the pedestrian. The vehicle periphery monitoring device according to claim 1,
The alarm target changing unit is
When the number of detections within a predetermined time of an object other than the pedestrian and the animal detected from an image captured by the imaging device is less than or equal to a predetermined number A vehicle periphery monitoring device, wherein the setting of an object is changed from the pedestrian to the animal. The vehicle periphery monitoring device according to claim 1,
The alarm target changing unit is
Considering the total number of the pedestrians and the animals in the detected image as the surrounding environment situation of the host vehicle, if the total number is less than or equal to a threshold value, the alarm target setting is set to the walking A vehicle periphery monitoring device, wherein the vehicle is changed from the person to the animal. The vehicle periphery monitoring device according to claim 1,
And a navigation device for detecting surrounding environment information of the host vehicle.
The alarm target changing unit is
The vehicle periphery monitoring characterized in that, when it is determined from the information of the navigation device that the surrounding environment currently running is an environment with few pedestrians, the setting of the alarm target is changed from the pedestrian to the animal. apparatus. The vehicle periphery monitoring device according to claim 1,
When the warning target change unit detects a warning sign that an animal may jump out from the image captured by the imaging device as the surrounding environment information of the host vehicle, the warning target change unit sets the warning target A vehicle periphery monitoring device, wherein the animal is changed from a pedestrian to the animal. In the vehicle periphery monitoring device according to any one of claims 1 to 4, 6, and 7,
The alarm target setting is further changed from the pedestrian to the animal, and further includes a duration measuring unit for measuring a duration from the establishment time of the determination, and the duration from the establishment time of the determination is a threshold time or more. Then, even if the establishment of the determination is interrupted for a certain period of time, the setting of the alarm target is retained in the changed animal at the time of the interruption. Monitoring device.

Images