JP2016192146A - Traffic control system, traffic control method, program and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a traffic control system capable of automatically optimizing traffic signals at an intersection according to characteristics of a target pedestrian.SOLUTION: The traffic control system includes: pedestrian two dimensional image acquisition means that acquires a two dimensional image of a pedestrian at an intersection; target pedestrian sorting means that sorts whether the pedestrian in the two dimensional image is a target pedestrian or not based on a piece of sorting information of the target pedestrian of traffic control from the acquired two dimensional image; and signal adjustment means that, when the pedestrian is sorted as the target pedestrian, adjusts the signal at the intersection based on adjustment conditions of the traffic signal corresponding to the target pedestrian.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a traffic control system, a traffic control method, a program, and a recording medium.

  At the intersection, a device is provided to protect vulnerable people. Examples of the device include a push button for a crosswalk (Non-Patent Document 1). When an elderly person or the like presses the push button, the lighting time of the green light for pedestrians can be extended more than usual.

July 1, 1994, published by Traffic Engineering Research Group, “Traffic Signal Guide”, p. 73

  However, the push button described in Non-Patent Document 1 needs to be operated by the weak traffic person who is the user. For example, depending on the position where the push button is installed, the operation may be difficult for a weak person such as a wheelchair user. For this reason, it is hard to say that it is effectively utilized for all traffic weak people.

  Accordingly, an object of the present invention is to provide a traffic control system capable of automatically optimizing intersection signals in accordance with the characteristics of pedestrians to be protected.

In order to achieve the above object, the traffic control system of the present invention comprises:
Pedestrian 2D image acquisition means for acquiring a 2D image of a pedestrian at an intersection;
From the acquired two-dimensional image, based on classification information of the target pedestrian for traffic control, the target pedestrian classification means for classifying whether the pedestrian of the secondary original image is the target pedestrian,
When the said pedestrian is classified into the said object pedestrian, the signal adjustment means which adjusts the signal of an intersection based on the adjustment condition of the signal corresponding to the said object pedestrian is characterized by the above-mentioned.

The traffic control method of the present invention
A pedestrian 2D image acquisition step of acquiring a 2D image of a pedestrian at an intersection;
A target pedestrian classification step for classifying whether the pedestrian in the two-dimensional image is the target pedestrian or not based on classification information of the target pedestrian for traffic control from the acquired two-dimensional image;
When the pedestrian is classified as the target pedestrian, a signal adjustment step of adjusting an intersection signal based on a signal adjustment condition corresponding to the target pedestrian is included.

  The program of the present invention causes a computer to execute the traffic control method of the present invention.

  The recording medium of the present invention can be read by a computer in which the program of the present invention is recorded.

  According to the present invention, an intersection signal can be automatically optimized in accordance with the characteristics of a pedestrian to be protected.

FIG. 1 is a block diagram illustrating an example of a traffic control system according to the first embodiment. FIG. 2 is a flowchart illustrating an example of the traffic control method according to the first embodiment. 3A to 3B are explanatory diagrams of specific examples of the traffic control method according to the first embodiment. FIG. 4 is a block diagram illustrating an example of a traffic control system according to the second embodiment. FIGS. 5A to 5B are explanatory diagrams of specific examples of the traffic control method according to the second embodiment. FIG. 6 is a block diagram illustrating an example of a traffic control system according to the third embodiment. FIG. 7 is an explanatory diagram of a specific example of the traffic control method according to the third embodiment.

  In the present invention, the target pedestrian is, for example, a pedestrian that needs protection in passing through an intersection. Examples of the target pedestrian include a so-called traffic weak person. The traffic weak person means, for example, a person whose movement is restricted, a person who is easily damaged by a traffic accident, and specific examples include a wheelchair user, a cane user, a stroller user, a walker user, and a pregnant woman. , Children, elderly people, drunks, etc.

  Next, an embodiment of the present invention will be described. Note that the present invention is not limited to the following embodiments. In addition, in the following FIGS. 1-7, the same code | symbol is attached | subjected to the same part. Moreover, the description of each embodiment can use each other's description unless there is particular mention. Further, the configurations of the embodiments can be combined unless otherwise specified.

[Embodiment 1]
FIG. 1 is a block diagram showing an example of the configuration of the traffic control system 100 of the present embodiment. The traffic control system 100 includes a pedestrian two-dimensional image acquisition unit 1, a target pedestrian classification unit 2, and a signal adjustment unit 3 as essential components. The means 1, 2 and 3 can be connected to each other via a communication network 10, for example. The communication line network 10 is not particularly limited, and a known communication line network can be used. Specifically, for example, an Internet line, a telephone line, a LAN (Local Area Network), and the like can be used. Good.

  The pedestrian two-dimensional image acquisition means 1 acquires a two-dimensional image of a pedestrian at the intersection. The pedestrian two-dimensional image acquisition unit 1 is not particularly limited, and examples thereof include an RGB camera (also referred to as an RGB color camera). Examples of the area of the 2D image at the intersection acquired by the pedestrian 2D image acquisition unit 1 include an area where the pedestrian waits for a signal, an area of a pedestrian crossing, and the like. For example, the two-dimensional image of the former area can classify pedestrians waiting to cross the signal, and the two-dimensional image of the latter area can, for example, identify pedestrians crossing a pedestrian crossing. Classification is possible.

  The target pedestrian classification means 2 classifies whether the pedestrian in the two-dimensional image is the target pedestrian or not from the acquired two-dimensional image based on the classification information of the target pedestrian for traffic control.

  The classification information of the target pedestrian is not particularly limited, and refers to, for example, person attribute information of the target pedestrian. Examples of the person attribute include whether or not a specific tool is used, a specific target age, a specific physique, or a posture. Examples of the specific tool include a wheelchair, a cane, a stroller, and a walker. Examples of the specific target age include a child and an elderly person. Examples of the specific physique or posture include a pregnant woman, a back bend, and a waist bend.

  Below, an example of the method of classifying whether the pedestrian of the said two-dimensional image is the said target pedestrian by the target pedestrian classification means 2 is demonstrated. The target pedestrian classification means 2 extracts the target pedestrian classification information from the two-dimensional image using, for example, a person attribute determination model created in advance by machine learning (pattern recognition). The extracted classification information of the target pedestrian may be stored by storage means such as a database, for example. And the target pedestrian classification means 2 classifies whether it is the said target pedestrian based on the extracted classification | category information of the target pedestrian, for example. Specifically, when the classification information of the target pedestrian is information that a wheelchair is provided, the target pedestrian is classified as a wheelchair user.

  When the pedestrian is classified as the target pedestrian, the signal adjustment unit 3 adjusts the signal at the intersection based on the adjustment condition of the signal corresponding to the target pedestrian.

  The adjustment conditions are not particularly limited, and include, for example, control of right / left turn signals for vehicles such as extension of a pedestrian green signal, operation of a pedestrian scramble intersection, operation of an audio signal, extension of a red signal for a vehicle, etc. It is done. Examples of the control of the left / right turn signal for a vehicle include extension of display of the right / left turn signal for the vehicle, separation of pedestrians, and the like. For example, a red signal is used. Each adjustment condition corresponds to the target pedestrian. As a specific example, when the target pedestrian is a wheelchair user or an elderly person, the corresponding adjustment condition is, for example, an extension of a pedestrian green signal, and the target pedestrian is a white cane user, The corresponding adjustment condition is, for example, an audio signal operation, and when the target pedestrian is a child, the corresponding adjustment condition is, for example, an operation of a pedestrian scramble intersection.

  The traffic control system 100 according to the present embodiment classifies whether or not the pedestrian at the intersection is the target pedestrian by the target pedestrian classification means 2, and the signal adjustment means 3 determines the signal corresponding to the target pedestrian. Adjust the intersection signal based on the adjustment conditions. Thereby, for example, the person who is the target pedestrian does not need to perform a specific operation such as pressing a push button, and the signal at the intersection can be automatically optimized according to the characteristics of the target pedestrian.

  Next, the traffic control method of the present embodiment will be described using the flowchart of FIG.

  The traffic control method of this embodiment is implemented as follows, for example using the traffic control system 100 of this embodiment shown in FIG. In addition, the traffic control method of this embodiment is not limited to use of the traffic control system 100 of FIG.

  First, a two-dimensional image of a pedestrian at an intersection is acquired by the pedestrian two-dimensional image acquisition means 1 (pedestrian two-dimensional image acquisition step, S1). Next, whether or not the pedestrian in the two-dimensional image is the target pedestrian based on the classification information of the target pedestrian for traffic control from the acquired two-dimensional image by the target pedestrian classification means 2. Classify (target pedestrian classification step, S2). Next, when the pedestrian is classified as the target pedestrian, the signal at the intersection is adjusted based on the adjustment condition of the signal corresponding to the target pedestrian (signal adjustment step, S3). When the said pedestrian is not contained in the said object pedestrian, the process of the traffic control method of this embodiment is complete | finished, for example.

  The traffic control method of this embodiment classifies whether the pedestrian at the intersection is the target pedestrian by the target pedestrian classification step (S2), and corresponds to the target pedestrian by the signal adjustment step (S3). The signal at the intersection is adjusted based on the adjustment condition of the signal to be transmitted. Thereby, the target pedestrian himself / herself can automatically optimize the signal of the intersection corresponding to the characteristics of the target pedestrian without requiring the specific operation as described above.

  Next, a specific example of the traffic control method of the present embodiment will be described with reference to FIGS.

  FIGS. 3A and 3B are explanatory diagrams showing an outline of a specific example of the traffic control method of the present embodiment. FIG. 3A shows a situation where an elderly person A who uses a walking stick crosses a pedestrian crossing in an intersection. As shown in FIG. 3A, an RGB camera (pedestrian two-dimensional image acquisition means) 1 can image a pedestrian crossing a pedestrian crossing 12 on a support bar 11 (for example, a utility pole) on which a traffic light is installed. It is fixed to. It is preferable that the photographing angle of the RGB camera can be freely changed, for example. Then, the RGB camera 1 acquires a two-dimensional image of the elderly person A who is crossing the pedestrian crossing 12 in the intersection. FIG. 3B shows a two-dimensional image acquired by the RGB camera 1. The target pedestrian classification means 2 (not shown) uses a person attribute determination model created in advance by machine learning from the acquired two-dimensional image as a classification information of the target pedestrian from the two-dimensional image. Usage information and pedestrian posture information are detected. The detected information is stored in a database. Then, the target pedestrian classification means 2 classifies the target pedestrian as an elderly person based on the use information and posture information of the cane. Next, the signal adjustment means 3 (not shown) adjusts the signal of an intersection based on the corresponding signal adjustment conditions when the target pedestrian is an elderly person. In this case, since the target pedestrian is an elderly person, for example, the pedestrian green light is extended. In this example, a pedestrian crossing a pedestrian crossing is taken as an example. However, for example, a two-dimensional image of a pedestrian waiting for a signal may be acquired.

  In addition, although the specific example of the traffic control method of this embodiment was demonstrated using Fig.3 (a)-(b), the traffic control method of this embodiment is not limited to this.

[Embodiment 2]
FIG. 4 is a block diagram illustrating a configuration example of the traffic control system 200 of the present embodiment. The traffic control system 200 according to the present embodiment further includes a pedestrian three-dimensional image acquisition unit 4 and a target motion classification unit 5 as components in addition to the units of the traffic control system 100 according to the first embodiment. Each means can be connected to each other via the communication network 10, for example.

  The pedestrian 3D image acquisition means 4 acquires the 3D image of the pedestrian. The pedestrian three-dimensional image acquisition unit 4 is not particularly limited, and examples thereof include a camera that can obtain three-dimensional information, and examples thereof include a three-dimensional camera. The region of the three-dimensional image at the intersection acquired by the pedestrian three-dimensional image acquisition unit 4 is, for example, a pedestrian crossing region. The three-dimensional image of the area of the pedestrian crossing can classify, for example, the actions of pedestrians crossing the pedestrian crossing.

  The target motion classification means 5 classifies whether or not the pedestrian's motion in the 3D image is the target motion based on the acquired motion control target motion classification information from the acquired 3D image.

  The target operation of the traffic control is not particularly limited, and examples thereof include a movement speed and a movement locus.

  The method for calculating the moving speed is not particularly limited, and can be calculated, for example, based on the difference between the moving distance of the target pedestrian and the acquisition time of each three-dimensional image in a three-dimensional image acquired over time. The movement trajectory acquisition method is not particularly limited, and can be acquired by tracing the position of the target pedestrian in each three-dimensional image acquired over time, for example.

  The classification information of the target movement is not particularly limited, but when the target movement is a movement speed, for example, a movement speed within a specific range or a movement speed outside a specific range, a movement speed or a reference value below a reference value The above moving speed can be raised. When the target motion is a movement trajectory, for example, a specific walking pattern or the like can be cited, and examples of the specific walking pattern include staggered walking such as a staggered foot.

  Hereinafter, an example of a method for classifying whether or not the motion of the pedestrian in the three-dimensional image is the target motion by the target motion classification unit 5 will be described. The target action classification means 5 is, for example, when the moving speed of the pedestrian in the three-dimensional image is within or outside a predetermined range, or below a predetermined reference value or above a reference value, or When the movement locus corresponds to a specific walking pattern, the movement of the pedestrian is classified as the target movement.

  In this embodiment, when the pedestrian's movement is classified into the target movement, the signal adjustment section 23 outputs the signal at the intersection based on the signal adjustment condition (target movement adjustment condition) corresponding to the target movement. Adjust. According to the present embodiment, for example, even when the movement of a pedestrian in crossing is classified as a target movement even when the target pedestrian is not classified based on the two-dimensional image, Similarly, the signal of the intersection according to the pedestrian (according to the target movement) can be adjusted.

  In the present embodiment, for example, in the case where the pedestrian is a pedestrian classified as the target pedestrian from the two-dimensional image as described above, the signal adjustment condition (target walk) corresponding to the corresponding pedestrian Person's adjustment condition) and the signal adjustment condition corresponding to the target action, the signal at the intersection is adjusted.

  Examples of the signal adjustment conditions corresponding to the target operation include those exemplified in the first embodiment. Each adjustment condition corresponds to the target action. For example, when the target action is a moving speed within or outside a specific range, the adjustment condition corresponds to an extension of a pedestrian green signal, and the target action When the movement is a staggered walk, the adjustment condition corresponds to the extension of the pedestrian green light.

  As described above, the signal adjusting unit 23 adjusts the signal at the intersection based on the target movement adjustment condition or based on the target pedestrian adjustment condition and the target movement adjustment condition. In the latter case, the signal adjustment means 23 may adjust the signal at the intersection based on the target pedestrian adjustment condition and the target movement adjustment condition, and give priority to either the target pedestrian adjustment condition or the target movement adjustment condition. You may adjust the signal at the intersection.

  Next, the traffic control method of this embodiment will be described.

  The traffic control method of this embodiment is implemented as follows, for example using the traffic control system 200 of this embodiment shown in FIG. In addition, the traffic control method of this embodiment is not limited to use of the traffic control system 200 of FIG.

  First, similarly to the traffic control method of the first embodiment, the pedestrian two-dimensional image acquisition step S1 and the target pedestrian classification step S2 are performed. Next, the pedestrian 3D image acquisition means 4 acquires a 3D image of the pedestrian classified as the target pedestrian (pedestrian 3D image acquisition step, S4). Next, the target motion classification means 5 classifies whether the pedestrian's motion in the 3D image is the target motion based on the classification information of the traffic control target motion from the acquired 3D image. (Target action classification step, S5). Next, based on the signal adjustment condition corresponding to the target pedestrian and the signal adjustment condition corresponding to the target movement when the movement of the pedestrian is classified as the target movement by the signal adjustment unit 23. The signal at the intersection is adjusted (signal adjustment step, S6). When the said pedestrian is not contained in the said object pedestrian, the process of the traffic control method of this embodiment is complete | finished, for example.

  In the traffic control system and the traffic control method of the present embodiment, the target motion classification means or step classifies whether or not a pedestrian's motion at an intersection is the target motion, and the signal adjustment means or step determines the target. The signal at the intersection is adjusted based on not only the signal adjustment condition corresponding to the pedestrian but also the signal adjustment condition corresponding to the target motion. Thereby, for example, the signal of the intersection can be automatically optimized according to the movement of the pedestrian crossing the pedestrian crossing. For this reason, for example, for a pedestrian who is difficult to determine whether it is a target pedestrian from a two-dimensional image, such as a drunken person, the signal of an intersection can be automatically optimized. Even when the target pedestrian is classified from the two-dimensional image, more suitable signal control is possible according to the walking motion of the pedestrian.

  Next, a specific example of the traffic control method of the present embodiment will be described with reference to FIG.

  FIGS. 5A and 5B are explanatory views showing an outline of a specific example of the traffic control method of the present embodiment. FIG. 5A shows a situation where a drunk person A crosses a pedestrian crossing in an intersection. As shown in FIG. 5A, an RGB camera (pedestrian two-dimensional image acquisition means) 1 and a three-dimensional camera (pedestrian three-dimensional image acquisition means) 4 are connected to a support bar 11 on which a traffic light is installed. The pedestrian crossing 12 is fixed so as to be able to take an image. Then, the RGB camera 1 and the three-dimensional camera 4 acquire a two-dimensional image and a three-dimensional image of the drunk person A who is crossing the pedestrian crossing 12 in the intersection. FIG. 5B is a diagram in which the trajectory of movement of the drunk person A is traced from the three-dimensional image acquired with time by the three-dimensional camera 4. The target pedestrian classification means 2 (not shown) classifies the drunken person A of the two-dimensional image as a target pedestrian from the acquired two-dimensional image based on the classification information of the target pedestrian. To do. Here, when the classification information of the target pedestrian is not detected from the two-dimensional image for the drunken person A, the drunken person A is classified as not the target pedestrian. On the other hand, the target motion classification means 5 classifies the motion of the drunk person A at the intersection as a target motion. Specifically, the target motion classification unit 5 detects staggered walking as the target motion classification information from the three-dimensional image. Information on the detected wobbling walking is stored in a database. Then, the target motion classification means 4 classifies the target motion as a staggered walk based on the information on the staggered walk. Next, the signal adjustment means 23 adjusts the signal at the intersection based on the adjustment condition of the signal corresponding to the staggered walking. In this case, since the target motion is staggered walking, the pedestrian green signal is extended.

  FIG. 5 shows an example in which the target pedestrian is not classified by the two-dimensional image but is classified by the three-dimensional image as the target motion. However, the present invention is not limited to this. Similarly, signal adjustment is possible even when the walking motion of a pedestrian classified as a person is further classified as a target motion by a three-dimensional image.

  In addition, although the specific example of the traffic control method of this embodiment was demonstrated using Fig.5 (a)-(b), the traffic control method of this embodiment is not limited to this.

[Embodiment 3]
FIG. 6 is a block diagram illustrating an exemplary configuration of the traffic control system 300 according to the present embodiment. The traffic control system 300 of the present embodiment further includes a vehicle three-dimensional image acquisition unit 6 and a target vehicle situation classification unit 7 as components in addition to the units of the traffic control system 100 of the first embodiment. Each means can be connected to each other via the communication network 10, for example. Note that the traffic control system of the present embodiment may include, for example, a vehicle three-dimensional image acquisition unit 6 and a target vehicle situation classification unit 7 as components in addition to the units of the traffic control system of the second embodiment.

  The vehicle 3D image acquisition means 6 acquires a vehicle 3D image. The three-dimensional image acquired by the vehicle three-dimensional image acquisition means 6 is, for example, an image of a car that is lined up in front of the signal in the traveling direction (a car that is lined up at an intersection). , Right turn car, left turn car, etc. The vehicle three-dimensional image acquisition unit 6 is not particularly limited, and examples thereof include those exemplified in the pedestrian three-dimensional image acquisition unit of the second embodiment.

  The target vehicle situation classification means 7 classifies the number of the vehicles in the three-dimensional image from the acquired three-dimensional image based on the classification information of the target vehicle situation of traffic control as to whether or not the target vehicle situation.

  The classification information of the target vehicle situation is not particularly limited, and examples thereof include a large or small number of straight vehicles in the target vehicle situation, a large or small number of right-turn cars, and a large or small number of left-turn cars. The criterion for determining whether the number of vehicles is large or small may be based on, for example, a predetermined evaluation criterion.

  Hereinafter, an example of a method for classifying whether or not the vehicle status of the three-dimensional image is the target vehicle status by the target vehicle status classification unit 7 will be described. For example, when the target vehicle status is larger or smaller than a predetermined evaluation criterion, the target vehicle status classification unit 7 classifies the number of vehicles in the three-dimensional image as the target vehicle status.

  When the vehicle condition is classified into the target vehicle situation, the signal adjustment unit 33 adjusts the signal at the intersection based on a signal adjustment condition (target vehicle condition adjustment condition) corresponding to the target vehicle situation.

  In the present embodiment, for example, as described above, when the pedestrian is a pedestrian classified as the target pedestrian from the two-dimensional image, or when the pedestrian is classified as the target motion from the three-dimensional image. In the case of a person, a signal adjustment condition (target pedestrian adjustment condition) corresponding to the corresponding pedestrian, a signal adjustment condition (target movement adjustment condition) corresponding to the target movement, and the target vehicle condition adjustment condition Based on the signal of the intersection.

  The signal adjustment conditions corresponding to the target vehicle situation are not particularly limited.For example, when there are many straight ahead vehicles, the vehicle green signal is extended, and when there are many right turn vehicles, the right turn permission signal is extended, If there are many cars, the pedestrian signal may be red.

  As described above, the signal adjustment means 33 is based on the target vehicle situation adjustment condition, on the basis of the target pedestrian adjustment condition and the target vehicle situation adjustment condition, or on the target pedestrian adjustment condition and the target motion adjustment. An intersection signal is adjusted based on the condition and the target vehicle condition adjustment condition. The signal adjustment means 33 may adjust the signal of the intersection based on, for example, the target pedestrian adjustment condition, the target motion adjustment condition, and the target vehicle situation adjustment condition. The signal may be adjusted.

  Next, the traffic control method of this embodiment will be described.

  The traffic control method of this embodiment is implemented as follows using the traffic control system 300 of this embodiment shown in FIG. 6, for example. In addition, the traffic control method of this embodiment is not limited to use of the traffic control system 300 of FIG.

  First, similarly to the traffic control method of the first embodiment, the pedestrian two-dimensional image acquisition step S1 and the target pedestrian classification step S2 are performed. Next, the vehicle three-dimensional image acquisition means 6 acquires a three-dimensional image of the vehicle (vehicle three-dimensional image acquisition step, S7). Next, whether or not the number of the vehicles in the three-dimensional image is the target vehicle status based on classification information of the target vehicle status of traffic control from the acquired three-dimensional image by the target vehicle status classification means 7. (Target vehicle situation classification step, S8). Next, based on the adjustment condition of the signal corresponding to the target pedestrian by the adjusting means 33 and the adjustment condition of the signal corresponding to the target vehicle situation when the number of vehicles is classified into the target vehicle situation. The signal at the intersection is adjusted (signal adjustment step S3).

  In addition, the traffic control method of this embodiment may perform S7-S8 after performing S1-2, S4-5 in the traffic control method of Embodiment 2, for example.

  In the traffic control system and the traffic control method of the present embodiment, the target vehicle situation classification means or step classifies whether the vehicle situation is the target vehicle situation, and the signal adjustment means or step determines the target pedestrian. The signal at the intersection is adjusted based on not only the signal adjustment condition corresponding to the vehicle condition but also the signal adjustment condition corresponding to the target vehicle situation. Thereby, the signal of the intersection can be automatically optimized in accordance with the situation of the cars lined up at the intersection.

  Next, a specific example of the traffic control method of the present embodiment will be described with reference to FIG.

  FIG. 7 is an explanatory diagram of a specific example of the traffic control method of the present embodiment. FIG. 7 shows a state in which a plurality of target vehicles lined up at the intersection turn right on the pedestrian crossing in the intersection. As shown in FIG. 7, an RGB camera (pedestrian two-dimensional image acquisition means) 1 and a three-dimensional camera 4 are fixed to a support bar 11 on which a traffic light is installed. The RGB camera 1 is fixed so that a pedestrian crossing the pedestrian crossing 12 can be imaged, and the three-dimensional camera 4 is fixed so as to be able to image cars arranged at the intersection. Then, the RGB camera 1 and the three-dimensional camera 4 acquire a two-dimensional image of the elderly person A and the right-turning car at the pedestrian crossing 12 in the intersection. Next, the target pedestrian classification means 2 (not shown), from the acquired two-dimensional image, the elderly person A of the two-dimensional image based on the classification information of the target pedestrian as in the first embodiment, Classify as a target pedestrian. On the other hand, the target vehicle status classification means 7 classifies the vehicle status at the intersection as to whether it is the target vehicle status. Specifically, the target vehicle situation classification unit 7 detects information on the number of right-turn cars as the classification information of the target vehicle situation from the three-dimensional image. The detected information is stored in a database. Then, the target vehicle status classification means 7 classifies that there are many right-turn cars as the target vehicle status based on the information. Next, the signal adjustment means 33 adjusts the signal at the intersection based on the adjustment condition of the signal corresponding to the elderly person A and the adjustment condition that there are many right-turn cars. In this case, the pedestrian green signal is extended and the right turn allowable signal is extended.

  In FIG. 7, for example, the three-dimensional camera 4 may be fixed so as to be able to capture an image of a movement of a pedestrian crossing a pedestrian crossing. In this case, as described above, it is possible to classify whether or not the pedestrian's movement is the target movement and adjust the signal at the intersection.

  In addition, although the specific example of the traffic control method of this embodiment was demonstrated using FIG. 7, the traffic control method of this embodiment is not limited to this.

[Embodiment 4]
The program of this embodiment is a program that can execute the traffic control method of Embodiments 1 to 3 on a computer. Or the program of this embodiment may be recorded on a computer-readable recording medium, for example. The recording medium is not particularly limited, and examples thereof include a read only memory (ROM), a hard disk (HD), an optical disk, a floppy (registered trademark) disk (FD), and the like.

  While the present invention has been described with reference to the embodiments, the present invention is not limited to the above embodiments. Various changes that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.

  Although the traffic control system and the traffic control method of this embodiment are not particularly limited, the traffic control system and the traffic control method can be suitably applied to a system that adjusts an intersection signal according to a pedestrian or a car.

DESCRIPTION OF SYMBOLS 1 Pedestrian two-dimensional image acquisition means 2 Target pedestrian classification means 3 Signal adjustment means 10 Communication network 100 Traffic control system

Claims (16)

Pedestrian 2D image acquisition means for acquiring a 2D image of a pedestrian at an intersection;
Target pedestrian classification means for classifying whether the pedestrian in the two-dimensional image is the target pedestrian from the acquired two-dimensional image based on the classification information of the target pedestrian for traffic control,
When the said pedestrian is classified into the said object pedestrian, the traffic control system characterized by including the signal adjustment means which adjusts the signal of an intersection based on the adjustment condition of the signal corresponding to the said object pedestrian. The classification information of the target pedestrian is a classification of at least one target pedestrian selected from the group consisting of a wheelchair user, a cane user, a stroller user, a walker user, a pregnant woman, a child, an elderly person, and a drunk person. The traffic control system according to claim 1 which is information. The adjustment condition is at least one selected from the group consisting of an extension of a pedestrian green signal, an operation of a pedestrian scramble intersection, an operation of an audio signal, an extension of a red signal for a vehicle, and a control of a right / left turn signal for a vehicle. The traffic control system according to claim 1, wherein the traffic control system is an adjustment condition. further,
Pedestrian 3D image acquisition means for acquiring a 3D image of a pedestrian at an intersection;
From the acquired three-dimensional image, based on the classification information of the target operation of traffic control, including a target motion classification means for classifying whether the pedestrian motion of the three-dimensional image is the target motion,
The said signal adjustment means adjusts the signal of an intersection based on the adjustment conditions of the signal corresponding to the said object movement, when the movement of the said pedestrian is classified into the said object movement. The traffic control system according to one item. The signal adjusting means classifies the pedestrian of the two-dimensional image as the target pedestrian, and classifies the motion of the target pedestrian as the target motion, and a signal corresponding to the target motion of the target pedestrian The traffic control system according to claim 4, wherein the signal at the intersection is adjusted based on the adjustment condition. The traffic control system according to claim 4 or 5, wherein the classification information of the target motion is a movement speed or a movement trajectory. further,
Vehicle three-dimensional image acquisition means for acquiring a three-dimensional image of the vehicle before the intersection;
From the acquired three-dimensional image, based on the classification information of the target vehicle situation of traffic control, including the target vehicle situation classification means for classifying whether the vehicle situation of the three-dimensional image is the target vehicle situation,
7. The signal according to claim 1, wherein when the vehicle condition is classified into the target vehicle situation, the signal adjusting unit adjusts an intersection signal based on an adjustment condition of a signal corresponding to the target vehicle situation. The traffic control system according to item. A pedestrian 2D image acquisition step of acquiring a 2D image of a pedestrian at an intersection;
A target pedestrian classification step for classifying whether the pedestrian in the two-dimensional image is the target pedestrian or not based on classification information of the target pedestrian for traffic control from the acquired two-dimensional image;
And a signal adjustment step of adjusting an intersection signal based on an adjustment condition of a signal corresponding to the target pedestrian when the pedestrian is classified as the target pedestrian. The classification information of the target pedestrian is a classification of at least one target pedestrian selected from the group consisting of a wheelchair user, a cane user, a stroller user, a walker user, a pregnant woman, a child, an elderly person, and a drunk person. The traffic control method according to claim 8, which is information. The adjustment condition is at least one selected from the group consisting of an extension of a pedestrian green signal, an operation of a pedestrian scramble intersection, an operation of an audio signal, an extension of a red signal for a vehicle, and a control of a right / left turn signal for a vehicle. The traffic control method according to claim 8 or 9, which is an adjustment condition. further,
A pedestrian 3D image acquisition step of acquiring a 3D image of the pedestrian at the intersection;
From the acquired three-dimensional image, based on classification information of the target operation of traffic control, including a target motion classification step of classifying whether the pedestrian motion of the three-dimensional image is the target motion,
The signal adjustment step adjusts a signal at an intersection based on an adjustment condition of a signal corresponding to the target motion when the pedestrian's motion is classified as the target motion. The traffic control method according to item. The signal adjusting step classifies the pedestrian of the two-dimensional image as the target pedestrian, and the signal corresponding to the target motion of the target pedestrian when the motion of the target pedestrian is classified as the target motion. The traffic control method according to claim 11, wherein the signal at the intersection is adjusted based on the adjustment condition. The traffic control method according to claim 11 or 12, wherein the classification information of the target motion is a movement speed or a movement locus. further,
A vehicle 3D image acquisition step of acquiring a 3D image of the vehicle in front of the intersection;
From the acquired three-dimensional image, based on classification information of the target vehicle situation of traffic control, including the number of the vehicles of the three-dimensional image, a target vehicle situation classification step that classifies whether or not the target vehicle situation,
The signal adjustment step adjusts a signal at an intersection based on a signal adjustment condition corresponding to the target vehicle situation when the number of the vehicles is classified into the target vehicle situation. The traffic control method according to one item. A program for causing a computer to execute the traffic control method according to any one of claims 8 to 14. A computer-readable recording medium on which the program according to claim 15 is recorded.

Images