JP4935795B2 - Pedestrian pop-out prediction device and program - Google Patents

Description

  The present invention relates to a pedestrian jumping prediction apparatus and program, and more particularly to a pedestrian jumping prediction apparatus and program for predicting whether or not a pedestrian jumps on a roadway.

2. Description of the Related Art Conventionally, there is known a technique for performing a warning by determining whether a pedestrian and a vehicle collide when the current speed is constant from the presence, direction, and moving speed of the pedestrian and the vehicle (Patent Document). 1, 2). Also known is a technology that warns of the potential risk of pedestrian jumping by making judgments such as `` There are many popping out when it is close to school '' from the map and time zone, `` There are many popping out when going to school time '' (Patent Document 3). In addition to the presence, orientation, and movement speed of pedestrians and vehicles, a technology is known that determines the possibility of a collision between a pedestrian and a vehicle when the current speed is constant, based on a map or time zone. (Patent Document 4).
Japanese Patent No. 2582119 JP 2005-327177 A JP 2006-92258 A JP 2007-257338 A

  However, the techniques described in Patent Documents 1 to 4 do not take into consideration the intention of the pedestrian to jump out, and thus there is a problem that the pedestrian's jump-out cannot be accurately predicted.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to provide a pedestrian jumping prediction apparatus and program capable of accurately predicting pedestrian jumping.

In order to achieve the above object, the pedestrian jump prediction device according to the first aspect of the present invention is an acquisition means for acquiring a time-series change in the position of a pedestrian existing in front of the target vehicle and a time-series change in moving speed; A road structure acquisition unit that acquires a road structure of a road on which the target vehicle is traveling, and a time series of the position acquired by the acquisition unit when the road structure acquired by the road structure acquisition unit includes a sidewalk. A pedestrian jumps out onto the roadway when the amount of change in the time series change of at least one of the position and the movement speed is discontinuously changed before the boundary between the sidewalk and the roadway. by comparing the pattern of the time series change of the time series change and the moving speed of the previously obtained position when, to predict whether the pedestrian in the roadway where the subject vehicle is traveling pops When the road structure acquired by the road structure acquisition means does not have a sidewalk, the time series change of the position and the time series change of the moving speed acquired by the acquisition means, the pedestrian traffic area and the vehicle traffic on the road Compared with the pattern obtained in advance, in which the amount of change in the time series change of at least one of the position and the moving speed is discontinuously changed before the boundary with the region, the pedestrian is on the roadway. And pop- out prediction means for predicting whether or not to pop out.

According to a second aspect of the present invention, there is provided a program for acquiring a time series change in a position of a pedestrian existing in front of a target vehicle and a time series change in a moving speed, a road on which the target vehicle is traveling. Road structure acquisition means for acquiring a road structure, and when the road structure acquired by the road structure acquisition means has a sidewalk, the time series change of the position and the time series change of the moving speed acquired by the acquisition means; A time series of positions determined in advance when a pedestrian jumps out onto a roadway, the amount of change in the time series change of at least one of the position and the moving speed is discontinuously changed before the boundary between the sidewalk and the roadway change and by comparing the pattern of the time series change in the moving speed, the target vehicle predicts whether the pedestrian in roadway running pops the road structure acquisition unit Therefore, when the acquired road structure does not have a sidewalk, the time series change of the position and the time series change of the moving speed acquired by the acquisition means, and the boundary between the pedestrian traffic area and the vehicle traffic area on the road Whether or not the pedestrian jumps out on the roadway is compared with the previously obtained pattern in which the amount of change in the time series change of at least one of the position and the moving speed is discontinuously changed before this. is a program for functioning as a prediction to jump out predicting unit.

  According to the first invention and the second invention, the acquisition unit acquires the time series change of the position of the pedestrian existing in front of the target vehicle and the time series change of the moving speed. Then, the time series change of the position and the time series change of the movement speed acquired by the acquisition means by the pop-up prediction means, and the time series change of the position and the time series of the movement speed obtained in advance when the pedestrian jumps on the roadway It is possible to predict whether or not a pedestrian jumps out on the roadway on which the target vehicle is traveling by comparing with the change pattern.

  In this way, the time series change of the position of the pedestrian and the time series change of the moving speed are obtained, and the time series change of the position obtained in advance when the pedestrian jumps out on the roadway and the pattern of the time series change of the moving speed and By comparing, it is possible to accurately predict the pedestrian jumping.

  The pedestrian jumping prediction apparatus according to the first invention further includes road structure acquisition means for acquiring the road structure of the road on which the target vehicle is traveling, and the jumping prediction means is obtained by the road structure acquisition means. When the vehicle has a sidewalk, the amount of change in the time series change of the position acquired by the acquisition means and the time series change of the movement speed, and the time series change of at least one of the position and the movement speed before the boundary between the sidewalk and the roadway are calculated. If the road structure acquired by the road structure acquisition means does not have a sidewalk by comparing with the discontinuously changed pattern to predict whether or not a pedestrian will jump on the roadway, the position acquired by the acquisition means Time series change of the vehicle and the time series change of the moving speed, and the time series change of at least one of the position and moving speed before the boundary between the pedestrian traffic area and the vehicle traffic area on the road By comparing the definitive change amount was discontinuously varied patterns, it is possible to predict whether the roadway pedestrian jumps out. This makes it possible to perform pop-up prediction considering the pedestrian's pop-out intention using a pattern in which the amount of change in time series of at least one of the position and moving speed is discontinuously changed before the boundary between the sidewalk and the roadway. It can be carried out.

  The pedestrian jump prediction measure according to the first invention further includes a surrounding vehicle acquisition means for acquiring the presence or absence of a parked vehicle existing in front of the target vehicle, and the presence of the parked vehicle is acquired by the surrounding vehicle acquisition means, And when the parked vehicle exists around the pedestrian, the time-series change of the position and the time-series change of the moving speed acquired by the acquisition means, and at least one of the position and the moving speed before or after the parked vehicle It is possible to predict whether or not a pedestrian will jump out of the roadway by comparing with a pattern in which the amount of change in the series change is discontinuously changed. Accordingly, it is possible to perform pop-up prediction in consideration of the pedestrian's pop-out intention using a pattern in which the amount of change in time series change of at least one of the position and the moving speed before and after the parked vehicle is discontinuously changed.

  The pedestrian pop-up prediction device according to the first invention further includes pedestrian information acquisition means for acquiring pedestrian information indicating at least one of a pedestrian state and pedestrian characteristics, and the pop-up prediction means includes the acquisition means. The time series change of the position and the time series change of the movement speed acquired by the above and the time series change of the position obtained in advance and the pattern of the time series change of the movement speed were compared and acquired by the pedestrian information acquisition means. By comparing the pedestrian information with the pedestrian information obtained in advance when the pedestrian jumps out on the roadway, it can be predicted whether or not the pedestrian jumps out on the roadway. Thereby, the pedestrian's jumping out can be predicted with higher accuracy by further comparing the pedestrian's state obtained in advance when the pedestrian jumps out on the roadway or the characteristics of the pedestrian.

  The pedestrian jumping prediction apparatus according to the first aspect of the invention acquires surrounding environment information indicating at least one of a facility existing around the target vehicle, a signal state of a traffic light for a pedestrian, a time, and a situation of a surrounding vehicle. The surrounding environment acquisition means further includes a pop-up prediction means for the time series change of the position and the time series change of the movement speed acquired by the acquisition means, the time series change of the position obtained in advance and the time series change of the movement speed. In addition to comparing the pattern with the surrounding environment information acquired by the surrounding environment acquisition means and the surrounding environment information obtained in advance when the pedestrian jumps out on the roadway, whether or not the pedestrian jumps out on the roadway Can be predicted. This further increases the accuracy of pedestrian pop-up by further comparing the pedestrian's pre-determined surrounding facilities when the pedestrian jumps out onto the roadway, the traffic light signal status for the pedestrian, the time, or the surrounding vehicle status. Can be predicted well.

According to a third aspect of the present invention, there is provided a pedestrian pop-up prediction apparatus, wherein an acquisition unit that acquires a time-series change of a position of a pedestrian existing in front of a target vehicle and a time-series change of a moving speed, and the target vehicle is traveling. Road structure acquisition means for acquiring the road structure of the road, and when the road structure acquired by the road structure acquisition means has a sidewalk, the time series change of the position acquired by the acquisition means and the time series of the moving speed Predicting whether or not the pedestrian jumps out on the roadway on which the target vehicle is running, depending on whether or not the amount of change in at least one of the changes has discontinuously changed before the boundary between the sidewalk and the roadway , When the road structure acquired by the road structure acquisition means does not have a sidewalk, at least the time-series change of the position and the time-series change of the moving speed acquired by the acquisition means Whether or not the pedestrian jumps out on the roadway on which the target vehicle is traveling, depending on whether or not the amount of change in the direction has discontinuously changed before the boundary between the pedestrian traffic area and the vehicle traffic area on the road And pop-out prediction means for prediction.

According to a fourth aspect of the present invention, there is provided a program for acquiring a time series change of a position of a pedestrian existing in front of a target vehicle and a time series change of a moving speed, and a road on which the target vehicle is traveling. When the road structure acquisition means for acquiring the road structure and the road structure acquired by the road structure acquisition means have a sidewalk, the time series change of the position and the time series change of the moving speed acquired by the acquisition means Predicting whether or not the pedestrian jumps out on the roadway on which the target vehicle is running, depending on whether or not the amount of change in at least one has changed discontinuously before the boundary between the sidewalk and the roadway, and the road structure When the road structure acquired by the acquisition unit does not have a sidewalk, the time series change of the position and the time series change of the moving speed acquired by the acquisition unit are small. Whether or not the pedestrian jumps out on the roadway on which the target vehicle is traveling, depending on whether or not the amount of change in one of the roads has changed discontinuously before the boundary between the pedestrian traffic area and the vehicle traffic area on the road is a program for functioning as a pop-out predicting unit predicts a.

  According to the third and fourth aspects of the invention, the acquisition unit acquires the time series change of the position of the pedestrian existing in front of the target vehicle and the time series change of the moving speed. Then, the roadway on which the target vehicle is traveling depends on whether or not the amount of change in at least one of the time-series change of the position and the time-series change of the moving speed obtained by the pop-up prediction means changes discontinuously. Predict whether or not pedestrians will jump out.

  In this way, the time series change of the position of the pedestrian and the time series change of the moving speed are acquired, and whether or not the amount of change in at least one of the time series change of the position and the time series change of the moving speed has changed discontinuously. Therefore, it is possible to accurately predict the jumping out of the pedestrian.

  The pedestrian jumping prediction device includes a vehicle information acquisition unit that acquires the position and speed of the target vehicle, and a target vehicle acquired by the vehicle information acquisition unit when the pedestrian is predicted to jump out of the roadway by the jump prediction unit. Based on the position and speed of the vehicle and the position of the pedestrian, the risk determination means for determining the risk of collision between the target vehicle and the pedestrian, and the driving based on the risk determined by the risk determination means It may further include driving support means for providing support. Thus, an accident can be prevented by performing driving assistance when it is predicted that a pedestrian will jump out of the roadway and the risk of collision between the target vehicle and the pedestrian is high.

  As described above, according to the pedestrian pop-up prediction apparatus and program of the present invention, the time series change of the position of the pedestrian and the time series change of the moving speed are obtained, and are obtained in advance when the pedestrian jumps out on the roadway. The amount of change in at least one of the time-series change of the pedestrian position and the time-series change of the moving speed has changed discontinuously Whether or not the pedestrian jumps out can be predicted with high accuracy can be obtained.

  Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings. In the present embodiment, an example in which the present invention is applied to a driving support system that performs pop-up prediction using an in-vehicle device and outputs an alarm using information transmitted from a roadside information providing device will be described.

  As shown in FIG. 1, the driving support system 10 according to the first embodiment includes an information providing device 12 that is installed on the roadside and transmits pedestrian information indicating the position and moving speed of a pedestrian, and information The pedestrian information from the providing device 12 is used to predict that the pedestrian will jump out, and when it is determined that the risk of collision is high, the vehicle-mounted device 14 outputs an alarm to the driver.

  The information providing apparatus 12 performs a predetermined image processing on the camera 16 for photographing the pedestrian and the image photographed by the camera 16, thereby determining the position and moving speed of the pedestrian represented by the photographed image. An image processing unit 18 that detects at any time and an information transmission unit 20 that transmits pedestrian information indicating the detected position and moving speed of the pedestrian to the in-vehicle device 14 at any time by wireless communication are provided.

  The in-vehicle device 14 is transmitted from the GPS 22 that detects the position of the host vehicle, the vehicle speed sensor 24 that detects the vehicle speed of the host vehicle, the camera 26 for photographing the periphery of the host vehicle, and the roadside information providing device 12. Based on the information receiving unit 28 that receives pedestrian information by wireless communication, the output of the GPS 22, the vehicle speed sensor 24, and the camera 26, and the pedestrian information received by the information receiving unit 28, it is predicted that the pedestrian will jump out, And if it determines with the risk of colliding with a pedestrian being high, the computer 32 which outputs a warning to a driver via the warning device 30 is provided.

  The alarm device 30 includes a speaker and a display, and outputs an alarm sound as driving assistance and displays an alarm message.

  The computer 32 includes a CPU, a ROM that stores a program of a driving support processing routine that will be described later, a RAM that stores data, an HDD, and a bus that connects these. If the computer 32 is described with function blocks divided for each function realizing means determined based on hardware and software, as shown in FIG. 1, the computer 32 converts the pedestrian information received by the information receiving unit 28. A pedestrian vehicle information collection unit 34 that collects the position of the own vehicle from the GPS 22 and the vehicle speed from the vehicle speed sensor 24, a road map database 36 that stores map information including facility information and road structure information, and a camera 26 recognizes the situation of the surrounding vehicle and the signal condition of the traffic signal by performing predetermined image processing from the image taken by the camera 26, collects the surrounding environment information indicating the situation of the surrounding vehicle and the signal condition of the traffic signal, and from the road map database 36 Peripheral information collection that collects and collects information on the road structure of the road on which the vehicle is traveling and facilities around the vehicle Unit 38 and the time-series change pattern of the pedestrian position and moving speed based on the pedestrian jump-out case, the situation of the surrounding vehicle based on the pedestrian jump-out case, the signal status of the traffic light, and the surrounding facility information are stored. Whether or not a pedestrian existing in front of the own vehicle will jump out is predicted based on the pop-up example database 40, the pedestrian information, the surrounding environment information, and the time-series change pattern of the pedestrian position and moving speed. In addition, a jump-out risk prediction unit 42 that determines a risk of collision with a pedestrian existing in front of the host vehicle and a driving support control unit 44 that controls an alarm by the alarm device 30 are provided.

  The pedestrian vehicle information collection unit 34 corresponds to the acquisition unit and the vehicle information acquisition unit of the present invention, and the surrounding information collection unit 38 includes the road structure acquisition unit, the surrounding vehicle acquisition unit, and the surrounding environment acquisition of the present invention. Corresponds to the means. The pop-out risk prediction unit 42 corresponds to the pop-out prediction unit and the risk level determination unit of the present invention.

  The pedestrian vehicle information collection unit 34 collects the pedestrian information received by the information reception unit 28, and as illustrated in FIG. 2, the time series change and the moving speed of the position of the pedestrian existing in front of the host vehicle. Get time series change of. In FIG. 2, the X-axis direction is a direction parallel to the traveling direction of the roadway and the sidewalk, and the Y-axis direction is a direction parallel to the width direction of the roadway and the sidewalk. Further, the pedestrian vehicle information collection unit 34 acquires the position of the host vehicle obtained from the GPS 22 and the vehicle speed of the host vehicle obtained from the vehicle speed sensor 24.

  The surrounding information collection unit 38 performs predetermined image processing (for example, pattern recognition processing) on the image photographed by the camera 26, so that the situation of the surrounding vehicle existing around the own vehicle is, for example, the own vehicle Presence and location of parked vehicles, whether or not there is traffic, presence or location of vehicles in the adjacent lane, presence and location of vehicles in the opposite lane, and the road on which the vehicle is traveling Recognize and collect traffic volume. In addition, the peripheral information collection unit 38 recognizes and collects the signal status of the pedestrian traffic light by performing predetermined image processing (for example, pattern recognition processing) on the image captured by the camera 26. In addition, the peripheral information collection unit 38 acquires and collects information about the road structure of the road on which the host vehicle is traveling and the facilities existing around the host vehicle from the road map database 36.

  In the pop-up case database 40, for each type of road structure, the time-series change of the position of the pedestrian and the time-series change of the moving speed when the pedestrian jumps on the roadway are obtained from the analysis result of the pop-up case. It is remembered. For example, as shown in FIG. 3, with respect to a road structure having a sidewalk, before the boundary between the sidewalk and the roadway, the position where the amount of change in the movement speed changes discontinuously and the pattern of the time series change in the movement speed, As shown in FIG. 4, before the boundary between the sidewalk and the roadway, the amount of change in the position of the pedestrian changes discontinuously and the amount of change in the movement speed changes discontinuously. As shown in FIG. 5, the pattern of the time series change and the pattern of the time series change of the moving speed and the position where the amount of change in the position of the pedestrian discontinuously changes before the boundary between the sidewalk and the roadway, It is stored in the pop-up case database 40. In addition, for a road structure where there is no sidewalk, at a position where the amount of change in the movement speed discontinuously changes before the boundary between the pedestrian traffic area and the vehicle traffic area on the predetermined road. The position where the amount of change in the pedestrian's position changes discontinuously and the amount of change in the movement speed changes discontinuously before the boundary of the pedestrian traffic area and the vehicle traffic area. The pattern of the time-series change of the moving speed and the pattern of the time-series change of the moving speed and the position where the amount of change in the position of the pedestrian discontinuously changes before the boundary between the pedestrian-passing area and the vehicle-passing area. It is stored in the pop-up case database 40.

  Further, in the pop-out case database 40, a pedestrian obtained from the analysis result of the pop-up case corresponding to the case where a parked vehicle exists in front of the host vehicle and the parked vehicle is present around the pedestrian. A time-series change of the position of the pedestrian and a time-series change of the moving speed when jumping out on the roadway are stored. For example, when there is a parked vehicle in front of the host vehicle and the parked vehicle is in the vicinity of a pedestrian, the position where the amount of change in the movement speed changes discontinuously before or after the parked vehicle and The pattern of the time-series change of the moving speed and the position and moving speed of the pedestrian's position change discontinuously before and after the parked vehicle, and the moving speed change amount discontinuously changes. The time series change pattern and the position where the amount of change in the position of the pedestrian discontinuously changes before and after the parked vehicle and the time series change pattern of the moving speed are stored in the pop-up case database 40.

  In addition, the situation of surrounding vehicles when a pedestrian is likely to jump out on the roadway is, for example, that there is a traffic jam, that there is a stopped vehicle in the adjacent lane, that there is a stopped vehicle in the opposite lane, traffic It is stored in the pop-up example database 40 that the road has a small amount. In addition, the signal status of the traffic light for pedestrians when there is a high possibility that the pedestrian will jump out of the roadway, for example, the change from red to blue, the change from blue to flashing blue, and the flashing blue It is stored in the pop-up case database 40 that it is a change from red to red. In addition, as information on the surrounding facilities when there is a high possibility that pedestrians will jump out onto the roadway, for example, there are shops on the sidewalk across the roadway, there are parks in the vicinity, event venues in the vicinity, sightseeing The fact that there is a school and a kindergarten in the vicinity is stored in the pop-up example database 40.

  The pop-out risk prediction unit 42 predicts whether or not a pedestrian will jump out on the roadway, as will be described below.

  First, a pattern of time-series changes in the position and movement speed of a pedestrian corresponding to the type of road structure collected by the surrounding information collection unit 38 and the presence / absence of a parked vehicle in front of the host vehicle is extracted from the pop-up example database 40. Compare the time-series change of the position and movement speed of the pedestrian ahead of the own vehicle acquired by reading and the pedestrian vehicle information collection unit 34 with the time-series change pattern of the position and movement speed of the read pedestrian, The degree of coincidence is calculated. When there are a plurality of read patterns, the highest matching degree among the matching degrees calculated from each of the plurality of patterns may be used.

  In addition, the situation of surrounding vehicles based on the pedestrian jumping case, the surrounding facility information, and the signal status of the traffic light are read from the jumping case database 40 and collected by the surrounding information collecting unit 38, and the surrounding facilities The degree of coincidence is calculated by comparing the information and the situation of the traffic light for pedestrians with the situation of the read surrounding vehicles, the surrounding facility information, and the signal situation of the traffic lights.

  If the value obtained by combining the two degrees of coincidence calculated as described above is equal to or greater than the threshold value, it is predicted that a pedestrian will jump out on the roadway on which the vehicle is traveling, and the value obtained by combining the two degrees of coincidence is less than the threshold value. If so, it is predicted that the pedestrian will not jump out of the road.

  Further, when it is predicted that the pedestrian will jump out of the roadway, the jump-out risk prediction unit 42 determines the risk of collision between the pedestrian and the host vehicle, as will be described below.

  Based on the current position and average moving speed of the pedestrian ahead of the host vehicle and the current position and current speed of the host vehicle obtained from the position and moving speed of the pedestrian collected by the pedestrian vehicle information collecting unit 34, TTC (Time To Collision) is calculated, and the risk of collision between the pedestrian in front of the host vehicle and the host vehicle is determined based on the TTC. In addition, you may make it calculate TTC using the general moving speed of a predetermined pedestrian, without using the average moving speed of a pedestrian.

  The driving support control unit 44 causes the alarm device 30 to output an alarm to the driver when the degree of risk determined by the pop-out risk prediction unit 42 is high.

  Next, operation | movement of the driving assistance system 10 which concerns on 1st Embodiment is demonstrated. First, in the information providing apparatus 12 installed on the roadside, a pedestrian is detected from an image taken by the camera 16, and the detected position and movement speed of the pedestrian are detected, so that the position and movement of the pedestrian are detected. Pedestrian information indicating the speed is transmitted to the vehicle-mounted device 14 existing in the vicinity as needed.

  Further, the driving support processing routine shown in FIG. 6 is executed in the computer 32 of the in-vehicle device 14.

  First, in step 100, the pedestrian information received by the information receiving unit 28 is collected for a predetermined period, and the time series change of the position of the pedestrian existing in front of the host vehicle and the time series change of the moving speed are acquired. .

  In step 102, the position of the host vehicle obtained from the GPS 22 and the vehicle speed of the host vehicle obtained from the vehicle speed sensor 24 are acquired. In the next step 104, from the image taken by the camera 26, the situation of surrounding vehicles including the presence and position of the parked vehicle and the signal situation of the traffic light for pedestrians are recognized and acquired from the road map database 36. Then, the road structure of the road on which the host vehicle is traveling and the facility information existing in the vicinity are acquired.

  In step 106, the pattern of the time-series change of the position and moving speed when the pedestrian jumps out on the roadway corresponding to the road structure and the presence / absence and position of the parked vehicle acquired in step 104 from the pop-up case database 40 is obtained. In addition to reading, the situation of surrounding vehicles, the signal situation of traffic lights for pedestrians, and information on surrounding facilities when pedestrians jump out on the roadway are read.

  In step 108, the time-series change of the position and moving speed of the pedestrian ahead of the host vehicle acquired in step 100 is compared with the time-series change pattern of the position and moving speed read in step 106, The degree of coincidence is calculated. In addition, the situation of the surrounding vehicle acquired in step 104, the signal situation of the traffic signal for pedestrians, and the surrounding facility information, the situation of the surrounding vehicle read in step 106, the signal situation of the traffic signal for pedestrians, and The degree of coincidence is calculated by comparing with neighboring facility information. Whether or not the pedestrian jumps out is predicted based on a value obtained by combining the two calculated degrees of coincidence.

  In step 110, it is determined whether or not a pedestrian is predicted to jump out in step 108. If it is predicted that the pedestrian will not jump out, the process returns to step 100, while the pedestrian jumps out. If it is predicted, the routine proceeds to step 112.

  In step 112, based on the position and moving speed of the pedestrian in front of the host vehicle acquired in step 100 and the position and vehicle speed of the host vehicle acquired in step 102, the pedestrian and host vehicle in front of the host vehicle are Determine the risk of collision.

  In step 114, it is determined whether or not the degree of risk determined in step 114 is greater than or equal to a threshold. If the degree of risk is less than the threshold, the process returns to step 100, but the degree of risk is greater than or equal to the threshold. If there is, the alarm device 30 is controlled to output an alarm to the driver in step 116, and the process returns to step 100.

  By executing the above driving support processing routine, for example, as shown in FIGS. 7A and 7B, when the amount of change in the movement speed changes discontinuously before the boundary between the sidewalk and the roadway, Since it can be predicted that the pedestrian will jump out of the road, it can be predicted that the pedestrian will jump out of the road approximately 4 seconds before the time when the pedestrian actually jumps out of the road.

  As described above, according to the driving support system according to the first embodiment, the time series change of the position of the pedestrian in front of the host vehicle and the time series change of the moving speed are acquired, and the pedestrian jumps out on the roadway. By comparing with the time-series change of the position obtained in advance and the time-series change pattern of the moving speed, it is possible to accurately predict the pedestrian jumping in consideration of the pedestrian jumping intention.

  Furthermore, by further comparing the facilities that exist in the vicinity around the pedestrian when there is a high possibility that the pedestrian will jump out on the roadway, the traffic light signal status for the pedestrian, and the surrounding vehicle status, Predict with high accuracy.

  In addition, to prevent accidents with pedestrians by outputting an alarm to the driver when a pedestrian is expected to jump out on the roadway and the risk of collision between the host vehicle and the pedestrian is high Can do.

  In addition, it is possible to predict pedestrians jumping out using conventional static information earlier and more accurately, and the possibility of avoiding pedestrian accidents is increased. In addition, false information can be reduced.

  Next, a second embodiment will be described. In addition, since the structure of the driving assistance system which concerns on 2nd Embodiment is the structure similar to 1st Embodiment, it attaches | subjects the same code | symbol and abbreviate | omits description.

  The second embodiment differs from the first embodiment in that the pedestrian's state, pedestrian characteristics, time, etc. are further used to predict whether or not the pedestrian will jump out. .

  In the second embodiment, the image processing unit 18 of the information providing device 12 performs predetermined image processing on the image captured by the camera 16, so that the position and movement of the pedestrian represented by the captured image are displayed. It detects speed and recognizes pedestrian conditions and characteristics.

  For example, as the pedestrian state, the pedestrian's body direction, the pedestrian's face direction, the presence / absence of the accompanying person, the accompanying person's behavior, the pedestrian's action, etc. are recognized by the image recognition processing. Further, as the characteristics of the pedestrian, the age of the pedestrian, the clothes of the pedestrian, the luggage of the pedestrian, etc. are recognized by the image recognition process.

  Moreover, the information transmission part 20 transmits the pedestrian information which shows the position, moving speed, pedestrian state, and the characteristic of a pedestrian detected to the vehicle equipment 14 by radio | wireless communication.

  The pedestrian vehicle information collecting unit 34 of the in-vehicle device 14 collects the pedestrian information received by the information receiving unit 28, and the time series change of the position of the pedestrian existing in front of the own vehicle and the time series of the moving speed. While acquiring a change, the state of the pedestrian existing in front of the host vehicle and the characteristics of the pedestrian are acquired. Further, the pedestrian vehicle information collection unit 34 acquires the position of the host vehicle obtained from the GPS 22 and the vehicle speed of the host vehicle obtained from the vehicle speed sensor 24. In addition, the pedestrian vehicle information collection part 34 respond | corresponds to the acquisition means, pedestrian information acquisition means, and vehicle information acquisition means of this invention.

  In addition, the surrounding information collection unit 38 collects the situation of surrounding vehicles, the signal situation of traffic lights for pedestrians, the road structure of the road on which the vehicle is traveling and facility information existing in the vicinity, and the current time get.

  In the pop-out example database 40, the time-series change of the position of the pedestrian and the time-series change of the moving speed when the pedestrian jumps to the roadway, the situation of the surrounding vehicles when the pedestrian is likely to jump to the roadway, The signal status of traffic lights for pedestrians and surrounding facility information are stored. Further, the pop-out example database 40 stores the state of the pedestrian, the characteristics of the pedestrian, and the time zone when the pedestrian is likely to jump out onto the roadway.

  The pedestrian state when the pedestrian is likely to jump out onto the roadway, for example, that the body is facing the roadway, that the face is facing the roadway, that the companion has crossed the road first It is remembered that they are crazy about conversations with their companions, that they are talking on mobile phones, that they are looking for buildings, and that they are moving fast (running).

  The characteristics of pedestrians when there is a high possibility of pedestrians jumping onto the roadway include, for example, being older, being an infant, being dressed as a tourist such as a yukata, and having large luggage Etc. are stored.

  As a time zone when a pedestrian is likely to jump out onto the roadway, a commuting time zone in the morning and evening and a time zone with a low traffic volume (midnight and early morning) are stored.

  The pop-out risk prediction unit 42 includes a time-series pattern of pedestrian position and moving speed corresponding to the type of road structure collected by the surrounding information collection unit 38 and the presence / absence of a parked vehicle, and the pedestrian. The degree of coincidence is calculated by comparing the position of the pedestrian ahead of the host vehicle acquired by the vehicle information collection unit 34 and the time-series change in the moving speed. Further, the pop-out risk prediction unit 42 outputs the pop-up case database 40 based on the situation of the surrounding vehicle based on the pedestrian pop-up case, the surrounding facility information, the traffic light signal state, the pedestrian state, the characteristics of the pedestrian, and the time zone. , The pedestrian state and pedestrian characteristics collected by the pedestrian vehicle information collection unit 34, the situation of surrounding vehicles collected by the surrounding information collection unit 38, surrounding facility information, and traffic lights for pedestrians The degree of coincidence is calculated by comparing the current situation and the current time with the situation of the read surrounding vehicle, the surrounding facility information, the traffic light signal situation, the pedestrian state, the pedestrian characteristics, and the time zone.

  The pop-out risk prediction unit 42 predicts that the pedestrian will jump out of the roadway if the value obtained by combining the two coincidence degrees calculated as described above is equal to or greater than the threshold value, and the value obtained by combining the two coincidence degrees is If it is less than the threshold, it is predicted that the pedestrian will not jump out of the road.

  In addition, since the other structure and effect | action of the driving assistance system of 2nd Embodiment are the same as that of 1st Embodiment, description is abbreviate | omitted.

  In this way, when pedestrians are likely to jump out onto the roadway, the pedestrian jumps out can be predicted more accurately by further comparing the pedestrian state, pedestrian characteristics, and time zones that have been obtained in advance. can do.

  Next, a third embodiment will be described. In addition, about the part which becomes the structure similar to 1st Embodiment, the same code | symbol is attached | subjected and description is abbreviate | omitted.

  The third embodiment is different from the first embodiment in that the roadside information providing apparatus performs pedestrian pop-up prediction and collision risk determination.

  As shown in FIG. 8, when the driving support system 310 according to the third embodiment is installed on the roadside and predicts that a pedestrian will jump out and determines that the risk of collision is high, it transmits an alarm message. An information providing apparatus 312 and an in-vehicle device 314 that outputs an alarm to the driver when an alarm message is received from the information providing apparatus 312 are provided.

  The information providing apparatus 312 is based on a camera 316 for photographing a pedestrian, a vehicle, and the surrounding environment, an information transmission unit 20 that transmits information to the in-vehicle device 14 by wireless communication, and an image photographed by the camera 316. The computer 318 is configured to transmit an alarm message to the in-vehicle device 314 by the information transmitting unit 20 when it is predicted that the pedestrian will jump out and the risk of collision between the vehicle and the pedestrian is high.

  The computer 318 includes a CPU, a ROM that stores a program for a driving support processing routine, which will be described later, a RAM that stores data, an HDD, and a bus that connects them. When the computer 318 is described in terms of functional blocks divided for each function realizing means determined based on hardware and software, the computer 318 performs predetermined processing on an image photographed by the camera 316 as shown in FIG. By performing image processing, the position and moving speed of the pedestrian represented by the photographed image and the position and moving speed of the target vehicle are detected, and the image processing unit 320 recognizes the situation of the surrounding vehicle and the signal situation of the traffic light. A pedestrian vehicle information collection unit 322 that collects pedestrian information indicating the position and moving speed of the pedestrian detected by the image processing unit 320, collects the position and vehicle speed of the target vehicle, and a road map database 36. In addition to collecting the situation of surrounding vehicles and the signal situation of traffic lights recognized by the image processing unit 320, the road map A peripheral information collection unit 324 for collecting acquires facility information of the road structure and surrounding roads subject vehicle is traveling from the database 36, and a case database 40 jumping out, and a danger prediction unit 42 pops out.

  The information transmission unit 20 transmits an alarm message to the in-vehicle device 314 of the target vehicle when the degree of risk determined by the pop-out risk prediction unit 42 is high.

  The vehicle-mounted device 314 includes an information receiving unit 28, an alarm device 30, and a driving support control unit 344 that outputs an alarm to the driver by the alarm device 30 when an alarm message is received by the information receiving unit 28. ing.

  Next, the operation of the driving support system 310 according to the third embodiment will be described. First, in the information providing apparatus 312 installed on the roadside, pedestrians and vehicles are detected from the images taken by the camera 316, and the detected pedestrian position and moving speed, and the position and vehicle speed of the target vehicle are detected as needed. To detect. Further, in the information providing apparatus 312, the situation of the surrounding vehicle and the signal situation of the traffic light for the pedestrian are recognized and acquired from the image photographed by the camera 316 as needed.

  Further, the driving support processing routine shown in FIG. 9 is executed in the computer 318 of the information providing apparatus 312. In addition, about the process similar to 1st Embodiment, the same code | symbol is attached | subjected and detailed description is abbreviate | omitted. In addition, the driving support processing routine is executed for each of the target vehicles detected from the image captured by the camera 316.

  First, in step 350, pedestrian information detected from an image photographed by the camera 316 is collected for a predetermined period, and the time series change of the position of the pedestrian existing in front of the target vehicle and the time series change of the moving speed. To get.

  In step 352, the position and vehicle speed of the target vehicle detected from the image captured by the camera 316 are acquired. In the next step 354, the situation of surrounding vehicles including the presence and position of the parked vehicle and the signal situation of the traffic light for pedestrians are recognized and acquired from the image taken by the camera 316, and from the road map database 36. The information on the road structure of the road on which the target vehicle is traveling and the facility information existing in the vicinity are acquired.

  In step 106, the pattern of the time-series change of the position and moving speed when the pedestrian jumps out on the roadway corresponding to the road structure and the presence / absence and position of the parked vehicle acquired in step 354 from the pop-out case database 40 is obtained. In addition to reading, the situation of surrounding vehicles, the signal situation of traffic lights for pedestrians, and information on surrounding facilities when pedestrians jump out on the roadway are read.

  In step 108, the time series change of the position and movement speed of the pedestrian ahead acquired in step 350 is compared with the time series change pattern of the position and movement speed read in step 106, and the degree of coincidence is calculated. calculate. In addition, the situation of the surrounding vehicle acquired in step 354, the signal situation of the traffic signal for pedestrians, and the information on the surrounding facilities, the situation of the surrounding vehicle read in step 106, the signal situation of the traffic lights for pedestrians, and The degree of coincidence is calculated by comparing with information on surrounding facilities. Whether or not the pedestrian jumps out is predicted based on a value obtained by combining the two calculated degrees of coincidence.

  In step 110, it is determined whether or not the pedestrian is predicted to jump out in step 108. If it is predicted that the pedestrian does not jump out, the process returns to step 350. If it is predicted to jump out, the process proceeds to step 112.

  In step 112, the risk that the pedestrian and the target vehicle collide based on the position and moving speed of the pedestrian in front of the target vehicle acquired in step 350 and the position and vehicle speed of the target vehicle acquired in step 352. Determine the degree.

  In step 114, it is determined whether or not the degree of risk determined in step 114 is greater than or equal to a threshold. If the degree of risk is less than the threshold, the process returns to step 350, but the degree of risk is greater than or equal to the threshold. If there is, in step 356, the information transmission unit 20 transmits an alarm message to the vehicle-mounted device 314 of the target vehicle, and the process returns to step 350.

  In the in-vehicle device 314, when the information receiving unit 28 receives an alarm message for the host vehicle transmitted from the information providing device 312, the alarm device 30 outputs an alarm to the driver.

  As described above, according to the driving support system according to the third embodiment, the time series change of the position of the pedestrian in front of the target vehicle and the time series change of the moving speed are acquired, and the pedestrian jumps out on the roadway. By comparing with the time-series change of the position obtained in advance and the time-series change pattern of the moving speed, it is possible to accurately predict the pedestrian jumping in consideration of the pedestrian jumping intention.

  In addition, when it is predicted that a pedestrian will jump out on the roadway and the risk of collision between the target vehicle and the pedestrian is high, an alarm message is sent to the vehicle-mounted device of the target vehicle, Accidents can be prevented.

  In the above embodiment, the case where the information providing apparatus installed on the roadside performs the prediction of the pedestrian's popping out and the determination of the collision risk is described as an example. However, the present invention is not limited to this. In an information providing apparatus installed on the center side such as an information center, prediction of pedestrian pop-up and determination of collision risk may be performed.

  The second embodiment described above may also be applied to the third embodiment described above.

  Next, a fourth embodiment will be described. In addition, since the structure of the driving assistance system which concerns on 4th Embodiment is a structure similar to 1st Embodiment, it attaches | subjects the same code | symbol and abbreviate | omits description.

  In the fourth embodiment, based on the time-series changes in the position and movement speed of the pedestrian, it is determined whether the change in position or the change in movement speed has changed discontinuously. The point of predicting whether or not to jump out is different from the first embodiment.

  In the fourth embodiment, the peripheral information is based on the time-series changes in the position and moving speed of the pedestrian ahead of the host vehicle acquired by the pedestrian vehicle information collection unit 34 by the pop-out risk prediction unit 42 of the in-vehicle device 14. At least one of the change amount of the pedestrian's position and the change amount of the moving speed changes discontinuously depending on the type of road structure collected by the collection unit 38, the presence / absence of the parked vehicle, and the trouble of the boundary corresponding to the position. If it is determined that at least one of the change amount of the pedestrian's position and the change amount of the moving speed is discontinuously changed before the corresponding boundary, the pedestrian jumps out on the roadway. If it is determined that neither the amount of change in the position of the pedestrian nor the amount of change in the moving speed has discontinuously changed before the corresponding boundary, the pedestrian will not jump out of the roadway. Predict.

  In addition, since the other structure and effect | action of the driving assistance system of 4th Embodiment are the same as that of 1st Embodiment, description is abbreviate | omitted.

  In this way, the time series change of the position of the pedestrian and the time series change of the moving speed are acquired, and whether or not the amount of change in at least one of the time series change of the position and the time series change of the moving speed has changed discontinuously. Therefore, it is possible to accurately predict the jumping out of the pedestrian.

  Note that the contents of the second embodiment or the third embodiment described above may also be applied to the fourth embodiment.

  In the first to fourth embodiments described above, a case where an alarm is output as driving assistance has been described as an example. However, the present invention is not limited to this, and braking intervention control or the like may be used as driving assistance. Vehicle intervention control may be performed.

  Further, it may be predicted whether or not the pedestrian jumps out by comparing only the time-series change pattern of the pedestrian position and the moving speed without using the surrounding environment information or the like.

  Further, the degree of danger of colliding with a pedestrian may be determined in consideration of the surrounding environment such as weather and light and dark. In this case, information indicating weather and light / dark may be acquired from a wiper switch or brightness sensor mounted on the vehicle.

  The program according to the present invention may be provided by being stored in a storage medium.

1 is a block diagram showing a driving support system according to a first embodiment of the present invention. It is a figure which shows the example of the time series change of the position of a walking vehicle, and the time series change of a moving speed. It is the figure which showed the pattern in which the variation | change_quantity of the moving speed changed discontinuously before the boundary of a sidewalk and a roadway. It is the figure which showed the pattern in which the variation | change_quantity of the moving speed changed discontinuously before the boundary of a sidewalk and a roadway, and the variation | change_quantity of the position changed discontinuously. It is the figure which showed the pattern in which the variation | change_quantity of the position changed discontinuously before the boundary of a sidewalk and a roadway. It is a flowchart which shows the content of the driving assistance processing routine in the computer of the vehicle equipment which concerns on the 1st Embodiment of this invention. (A) A graph showing the relationship between the change in the position of the sidewalk in the direction of travel and the change in movement speed when the pedestrian jumps out on the roadway, and (B) the width direction of the sidewalk when the pedestrian jumps out onto the roadway It is the graph which showed the relationship between the change of the position of this, the change of a moving speed, and time. It is a block diagram which shows the driving assistance system which concerns on the 3rd Embodiment of this invention. It is a flowchart which shows the content of the driving assistance processing routine in the computer of the information provision apparatus which concerns on the 3rd Embodiment of this invention.

Explanation of symbols

10, 310 Driving support system 12, 312 Information providing device 14, 314 Vehicle-mounted device 16, 26, 316 Camera 24 Vehicle speed sensor 30 Alarm device 32, 318 Computer 34, 322 Pedestrian vehicle information collection unit 36 Road map database 38, 324 Information collection unit 40 Pop-up case database 42 Risk prediction unit 44, 344 Driving support control unit 320 Image processing unit

Claims (8)

An acquisition means for acquiring a time series change of a position of a pedestrian existing in front of the target vehicle and a time series change of a moving speed;
Road structure acquisition means for acquiring the road structure of the road on which the target vehicle is traveling;
When the road structure acquired by the road structure acquisition unit has a sidewalk, the time series change of the position and the time series change of the moving speed acquired by the acquisition unit, and before the boundary between the sidewalk and the roadway, A pattern of a time-series change of the position and a time-series change of the movement speed obtained in advance when the pedestrian jumps on the roadway, the amount of change in the time-series change of at least one of the position and the movement speed discontinuously changed And predicting whether or not the pedestrian jumps out on the roadway on which the target vehicle is running, and if the road structure acquired by the road structure acquisition means does not have a sidewalk, acquired by the acquisition means The position and the moving speed before the boundary between the pedestrian traffic area and the vehicle traffic area on the road, and the time-series change of the position and the time-series change of the moving speed. Discontinuous changing the amount of change in at least one of the time-series change, by comparing the pattern obtained in advance, and pop-out predicting unit predicts whether the pedestrian on the road pops
Pedestrian pop-up prediction device. It further includes surrounding vehicle acquisition means for acquiring the presence or absence of a parked vehicle that exists in front of the target vehicle,
When it is acquired by the surrounding vehicle acquisition means that the parked vehicle exists, and the parked vehicle exists in the vicinity of a pedestrian, the time-series change of the position and the movement speed acquired by the acquisition means The time series change is compared with the pattern in which the change amount in the time series change of at least one of the position and the moving speed before or after the parked vehicle is discontinuously changed, and the pedestrian jumps out on the roadway. whether predicting claim 1 Symbol placement of the pedestrian jumping out predicting unit. Further includes pedestrian information acquisition means for acquiring pedestrian information indicating at least one of the state of the pedestrian and the characteristics of the pedestrian,
The pop-up prediction means includes the time series change of the position and the time series change of the movement speed acquired by the acquisition means, and the pattern of the time series change of the position and the time series change of the movement speed obtained in advance. In addition to comparing, the pedestrian information acquired by the pedestrian information acquisition means is compared with the pedestrian information obtained in advance when the pedestrian jumps out on the roadway, and whether the pedestrian jumps out on the roadway. The pedestrian jumping prediction apparatus according to claim 1 or 2, which predicts whether or not. A surrounding environment acquisition means for acquiring surrounding environment information indicating at least one of a facility existing around the target vehicle, a signal status of a traffic light for a pedestrian, a time, and a status of the surrounding vehicle;
The pop-up prediction means includes the time series change of the position and the time series change of the movement speed acquired by the acquisition means, and the pattern of the time series change of the position and the time series change of the movement speed obtained in advance. Whether to compare the surrounding environment information acquired by the surrounding environment acquisition means with the surrounding environment information obtained in advance when the pedestrian jumps out on the roadway, and whether or not the pedestrian jumps out on the roadway The pedestrian pop-out prediction apparatus according to any one of claims 1 to 3 , wherein the pedestrian jump-out prediction apparatus predicts the above. An acquisition means for acquiring a time series change of a position of a pedestrian existing in front of the target vehicle and a time series change of a moving speed;
Road structure acquisition means for acquiring the road structure of the road on which the target vehicle is traveling;
When the road structure acquired by the road structure acquisition unit has a sidewalk, the amount of change in at least one of the time-series change of the position and the time-series change of the moving speed acquired by the acquisition unit is a difference between the sidewalk and the roadway. Whether or not the pedestrian jumps out on the roadway on which the target vehicle is traveling is predicted based on whether or not the vehicle has changed discontinuously before the boundary, and the road structure acquired by the road structure acquisition unit If not, the amount of change in at least one of the time-series change of the position and the time-series change of the moving speed acquired by the acquisition means is discontinuous before the boundary between the pedestrian traffic area and the vehicle traffic area on the road Pop-out prediction means for predicting whether or not the pedestrian jumps out on the roadway on which the target vehicle is traveling, depending on whether or not
Pedestrian pop-up prediction device. Vehicle information acquisition means for acquiring the position and speed of the target vehicle;
Based on the position and speed of the target vehicle acquired by the vehicle information acquisition means and the position of the pedestrian when the pedestrian is predicted to jump out on the roadway by the popping prediction means, the target vehicle And a risk determination means for determining a risk of collision between the pedestrian and the pedestrian,
The pedestrian jump prediction device according to any one of claims 1 to 5 , further comprising driving support means for performing driving support based on the risk determined by the risk determination means. Computer
An acquisition means for acquiring a time-series change of a position of a pedestrian existing in front of the target vehicle and a time-series change of a moving speed;
Road structure acquisition means for acquiring a road structure of a road on which the target vehicle is traveling; and
When the road structure acquired by the road structure acquisition unit has a sidewalk, the time series change of the position and the time series change of the moving speed acquired by the acquisition unit, and before the boundary between the sidewalk and the roadway, A pattern of a time-series change of the position and a time-series change of the movement speed obtained in advance when the pedestrian jumps on the roadway, the amount of change in the time-series change of at least one of the position and the movement speed discontinuously changed And predicting whether or not the pedestrian jumps out on the roadway on which the target vehicle is running, and if the road structure acquired by the road structure acquisition means does not have a sidewalk, acquired by the acquisition means The position and the moving speed before the boundary between the pedestrian traffic area and the vehicle traffic area on the road, and the time-series change of the position and the time-series change of the moving speed. Discontinuous changing the amount of change in at least one of the time-series change, by comparing the pattern obtained in advance, in order to function as a pop-out predicting unit predicts whether the pedestrian on the road pops Program. Computer
An acquisition means for acquiring a time-series change of a position of a pedestrian existing in front of the target vehicle and a time-series change of a moving speed;
Road structure acquisition means for acquiring a road structure of a road on which the target vehicle is traveling; and
When the road structure acquired by the road structure acquisition unit has a sidewalk, the amount of change in at least one of the time-series change of the position and the time-series change of the moving speed acquired by the acquisition unit is a difference between the sidewalk and the roadway. Whether or not the pedestrian jumps out on the roadway on which the target vehicle is traveling is predicted based on whether or not the vehicle has changed discontinuously before the boundary, and the road structure acquired by the road structure acquisition unit If not, the amount of change in at least one of the time-series change of the position and the time-series change of the moving speed acquired by the acquisition means is discontinuous before the boundary between the pedestrian traffic area and the vehicle traffic area on the road Depending on whether or not the pedestrian jumps out on the roadway on which the target vehicle is traveling. Program for

Images