Attention calling method selection device, attention calling method selection method, and program

Abstract

To enable selecting an appropriate caution method on the basis of an attribute of a pedestrian, and running-out possibility.SOLUTION: A pedestrian who is present in front of or behind a bus is detected on the basis of a picture photographed by a camera arranged in front of or behind the bus while the bus stops. An attribute of the pedestrian, and running-out possibility of the pedestrian based on a determination reference corresponding to the attribute are determined using a body state of the pedestrian detected from the picture. An instruction according to the caution method is output by selecting the caution method on the basis of the attribute and the running-out possibility.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present disclosure relates to an attention - calling method selection device, an attention - calling method selection method, and a program.

Background Art

[0002] A bus stops at a bus stop or the like for passengers to get on and off. At this time, pedestrians trying to cross the road passing in front of and behind the stopped bus will be hidden by the shadow of the bus body. Therefore, it is difficult for the driver of a following vehicle trying to overtake the bus or the driver of an oncoming vehicle traveling in the opposite lane to see the pedestrians trying to cross. As a result, contact accidents occur between pedestrians jumping out from behind the bus and following vehicles or oncoming vehicles.

[0003] Therefore, in order to prevent contact accidents, devices for warning pedestrians and drivers are known. The danger notification device described in Patent Document 1 detects pedestrians crossing in front of a large vehicle and vehicles trying to overtake the large vehicle. Then, it notifies the pedestrians and the overtaking vehicles of the danger.

[0004] Also, the in - vehicle warning device described in Patent Document 2 identifies the attributes of pedestrians behind a vehicle trying to reverse from the facial features of the facial images. Then, it switches the warning content according to the attributes of the pedestrians.

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0006] In the danger notification device described in Patent Document 1, there may be cases where it is not possible to give a warning corresponding to the attributes of pedestrians and the possibility of running out.

[0007] And in the in-vehicle warning device described in Patent Document 2, there may be cases where it is not possible to give a warning corresponding to the possibility of pedestrians running out.

[0008] An object of the present disclosure is to provide a warning method selection device, method, and program that enable selection of an appropriate warning method.

Means for Solving the Problems

[0009] In order to solve the above problems, in one aspect of the present disclosure, a warning method selection device includes a detection unit that detects pedestrians in front of and behind a bus based on an image captured by cameras installed in front of and behind the bus while the bus is stopped, a determination unit that, when the detection unit detects a pedestrian, determines the attributes of the pedestrian and the possibility of the pedestrian running out based on the determination criteria corresponding to the attributes using the physical state of the pedestrian detected from the image, a selection unit that selects a warning method based on the attributes and the possibility of running out determined by the determination unit, and an output unit that outputs an instruction corresponding to the warning method selected by the selection unit.

[0010] In order to solve the above problems, in one aspect of the present disclosure, a warning method selection method detects pedestrians in front of and behind a bus based on an image captured by cameras installed in front of and behind the bus while the bus is stopped, and when a pedestrian is detected, determines the attributes of the pedestrian and the possibility of the pedestrian running out based on the determination criteria corresponding to the attributes using the physical state of the pedestrian detected from the image, selects a warning method based on the determined attributes and the possibility of the pedestrian running out, and outputs an instruction corresponding to the selected warning method.

[0011] To solve the above problems, in one aspect of the present disclosure, the program causes a computer to detect pedestrians in front of and behind a bus based on an image captured by cameras installed in front of and behind the bus while the bus is stopped, and when a pedestrian is detected, uses the physical state of the pedestrian detected from the image to determine the attributes of the pedestrian and the possibility that the pedestrian will jump out based on the judgment criteria corresponding to the attributes, selects a warning method based on the determined attributes and the possibility that the pedestrian will jump out, and executes a process of outputting an instruction corresponding to the selected warning method.

Advantages of the Invention

[0012] According to the present disclosure, it becomes possible to select an appropriate warning method.

Brief Description of the Drawings

[0013]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

[0014] [First Embodiment] Next, the embodiments will be described in detail with reference to the drawings.

[0015] FIG. 1 is a block diagram showing the configuration of a warning method selection device 10 according to the first embodiment. The warning method selection device 10 includes a detection unit 101 that detects pedestrians in front of and behind the bus based on an image captured by cameras installed in front of and behind the bus while the bus is stopped, and a determination unit 102 that, when the detection unit 101 detects a pedestrian, uses the physical state of the pedestrian detected from the image to determine the possibility that the pedestrian will jump out based on the attributes of the pedestrian and the determination criteria corresponding to the attributes. The selection unit 103 selects a warning method based on the attributes and the possibility of jumping out determined by the determination unit 102, and an output unit 104 that outputs an instruction according to the warning method selected by the selection unit 103.

[0016] Here, the usage situation of the warning method selection device assumed in the present embodiment will be described with reference to FIGS. 2 and 3. FIGS. 2 and 3 are examples, and the position of the vehicle, the attributes of the pedestrians, and the warning method selection method are not limited thereto.

[0017] Figure 2 is a diagram showing the positional relationship between pedestrian A1 who is passing behind the parked bus B1 and attempting to cross the road, and oncoming vehicle C1 traveling in the oncoming lane of the bus. In the situation shown in Figure 2, pedestrian A1 is in a state where it is difficult to be seen by the driver of oncoming vehicle C1 due to the vehicle of bus B1. Therefore, there is a risk of a collision accident between pedestrian A1 and oncoming vehicle C1. Thus, the attention - calling method selection device of the present disclosure issues an attention call to pedestrian A1 and the driver of oncoming vehicle C1.

[0018] Figure 3 is a diagram showing the positional relationship between pedestrian A2 who is passing in front of the parked bus B2 and attempting to cross the road, and following vehicle C2 traveling in the same direction as the bus. In the situation shown in Figure 3, pedestrian A2 is in a state where it is difficult to be seen by the driver of following vehicle C2 due to the vehicle of bus B2. Therefore, there is a risk of a collision accident between pedestrian A2 and following vehicle C2. Thus, the attention - calling method selection device of the present disclosure issues an attention call to pedestrian A2 and the driver of following vehicle C2.

[0019] The configuration of the attention - calling method selection device 10 in the first embodiment will be described by taking the above usage situation as an example.

[0020] The detection unit 101 is an aspect of detection means for detecting pedestrians in front of and behind the bus based on an image captured by a camera installed in front of and behind the bus while the bus is parked. Specifically, the detection unit 101 receives an image captured by a camera installed in front of and behind the bus while the bus is parked. Then, the detection unit 101 detects pedestrians in front of and behind the bus based on the received image. General image recognition techniques are used for the detection of pedestrians. Here, the image used in this embodiment is a moving image.

[0021] When the detection unit 101 detects a pedestrian, the determination unit 102 is an aspect of a determination means that determines the attributes of the pedestrian and the possibility that the pedestrian will jump out based on the body state of the pedestrian detected from the video and the determination criteria corresponding to the attributes. Specifically, when the detection unit 101 detects a pedestrian, the determination unit 102 detects the body state of the pedestrian from the video. Examples of the body state of the pedestrian include posture, height, walking style, and moving speed. In this embodiment, the types of attributes to be determined are three types: elderly people, people operating a portable terminal, and children, but are not limited thereto. The determination unit 102 determines the attributes of the pedestrian from the body state of the pedestrian. Then, the determination unit 102 determines the possibility that the pedestrian will jump out based on the determination criteria corresponding to the attributes. In this embodiment, the determination is made in two stages: a high possibility of jumping out and a possible jumping out, but it is not limited to this. The determination criteria for the attributes and the determination criteria for the possibility of jumping out corresponding to the attributes of the pedestrian will be described later.

[0022] The selection unit 103 is an aspect of a selection means that selects a warning method based on the attributes and the possibility of jumping out determined by the determination means. The warning method selected by the selection unit 103 based on the attributes of the pedestrian and the possibility of jumping out will be described later.

[0023] The output unit 104 is an aspect of an output means that outputs an instruction corresponding to the warning method selected by the selection means. The output destination of the instruction is the front sign 121, the rear sign 122, the front speaker 131, the rear speaker 132, and the rear display 142, which will be described later. The output destination may be any one or more of the front sign 121, the rear sign 122, the front speaker 131, the rear speaker 132, or the rear display 142.

[0024] FIG. 4 is a block diagram showing the configuration of the alert system according to the first embodiment. The alert system of this embodiment includes an alert method selection device 10, a front camera 111, a rear camera 112, a front sign 121, a rear sign 122, a front speaker 131, a rear speaker 132, and a rear display 142. The configuration of the alert method selection device 10 is as described above. The alert method selection device 10, the front camera 111, the rear camera 112, the front sign 121, the rear sign 122, the front speaker 131, the rear speaker 132, and the rear display 142 are mounted on a bus.

[0025] The front camera 111 is a camera installed in front of the bus. The rear camera 112 is a camera installed behind the bus. When the bus is stopped, the front camera 111 and the rear camera 112 capture images of the front and rear of the bus. Here, the image refers to a moving image. Then, the captured images are transmitted to the detection unit 101 of the alert method selection device 10. The front camera 111 and the rear camera 112 may use, for example, a bus drive recorder.

[0026] The front sign 121 is a sign installed in front of the bus, and the rear sign 122 is a sign installed behind the bus. The front sign 121 and the rear sign 122 may use, for example, a display device for displaying the destination of the bus. The device used is not limited to this. The front sign 121 and the rear sign 122 receive an instruction from the output unit 104. An example of the display on the front sign 121 and the rear sign 122 is shown in FIG. 5.

[0027] The instruction output by the output unit 104 also includes information about the content to be displayed. The information about the content to be displayed includes, for example, the content of the message to be displayed, the border, and the color of the background of the message. The front sign 121 and the rear sign 122 that have received the instruction from the output unit 104 perform a display based on the information included in the instruction.

[0028] In addition, the instructions output from the output unit 104 only include instructions to perform a display corresponding to the attributes of the pedestrian and the likelihood of jumping out, and may not include information about the content to be displayed. In this case, the information about the content to be displayed is associated with the attributes of the pedestrian and the likelihood of jumping out, and is stored in the databases of the front sign 121 and the rear sign 122. The front sign 121 and the rear sign 122 that receive the instructions from the output unit 104 read out the information about the content to be displayed corresponding to the attributes of the pedestrian and the likelihood of jumping out from the databases. Then, the display is performed based on the read information.

[0029] The front speaker 131 is a speaker installed in the front of the bus, and the rear speaker 132 is a speaker installed in the rear of the bus. When the front speaker 131 and the rear speaker 132 receive an instruction from the output unit 104, they output the voice corresponding to that instruction.

[0030] The rear display 142 is a display installed in the rear of the bus. When it receives an instruction from the output unit 104, it performs the display corresponding to that instruction. An example of the display on the rear display is shown in FIG. 6.

[0031] The operation of the warning method selection device 10 in the warning system configured as described above will be described with reference to the flowchart of FIG. 7.

[0032] FIG. 7 is a flowchart showing an overview of the operation of the warning method selection device 10 in the first embodiment.

[0033] First, the detection unit 101 analyzes the video captured by the front camera 111 or the rear camera 112 while the bus is stopped (step S1). Then, based on the analysis result, the detection unit 101 detects whether there is a pedestrian in front of or behind the bus (step S2). If the detection unit 101 does not detect a pedestrian in front of or behind the bus (NO in step S2), the attention - calling method selection device 10 ends its operation. When the detection unit 101 detects a pedestrian in front of or behind the bus (YES in step S2), the determination unit 102 determines the attribute of the pedestrian and the possibility of running out.

[0034] First, the determination unit 102 determines whether the pedestrian is an elderly person (step S3). If it is determined that the pedestrian is an elderly person (YES in step S3), the process proceeds to step S4. In step S4, the determination unit 102 determines the possibility of the elderly person running out, the selection unit 103 selects an attention - calling method, and the output unit 104 outputs an instruction according to the attention - calling method. Steps S3 and S4 will be described in detail later.

[0035] If it is determined that the pedestrian is not an elderly person (NO in step S3), the determination unit 102 determines whether the pedestrian is a person operating a mobile terminal (step S5). If it is determined that the pedestrian is a person operating a mobile terminal (YES in step S5), the process proceeds to step S6. In step S6, the determination unit 102 determines the possibility of the person operating the mobile terminal running out, the selection unit 103 selects an attention - calling method, and the output unit 104 outputs an instruction according to the attention - calling method. Steps S5 and S6 will also be described in detail later.

[0036] When it is determined that the pedestrian is not a person operating a mobile terminal (NO in step S5), the determination unit 102 determines whether the pedestrian is a child (step S7). If it is determined that the pedestrian is a child (YES in step S7), the process proceeds to step S8. In step S8, the determination unit 102 determines the possibility of the child running out, the selection unit 103 selects a warning method, and the output unit 104 outputs an instruction according to the warning method. Steps S7 and S8 will also be described in detail later.

[0037] When the determination unit 102 determines that the pedestrian is not a child (NO in step S7), the selection unit 103 selects to display the video of the front camera 111 on the display (step S9). Then, the output unit 104 outputs an instruction to display the video of the front camera 111 on the display (step S9). Also, when the output unit 104 outputs an instruction according to the warning method in step S4, step S6, or step S8, the process also proceeds to step S9. Then, the warning method selection device 10 ends its operation.

[0038] In step S9, regardless of the attribute of the pedestrian, the video of the front camera 111 is displayed on the rear display 142. By displaying on the rear display, the driver of the following vehicle is prompted to pay attention. When the pedestrian is behind the bus, it may not be necessary to display on the rear display 142.

[0039] In an example of the display on the rear display 142 shown in FIG. 6, the border 142a of the display is displayed in, for example, red to alert the driver.

[0040] In the operation flow of this embodiment, regarding the attributes of pedestrians, judgments are made in the order of elderly people, people operating mobile terminals, and children. However, the order of judgment of these attributes may be swapped. By swapping the order of judgment of the attributes, the attention - calling method selection device of this embodiment is configured to first perform the judgment of the attribute for which attention - calling is desired preferentially. With this configuration, it is possible to shorten the time until the attention - calling method selection device selects and outputs the attention - calling method corresponding to the attribute for which attention - calling is desired preferentially. As a result, it is effective in preventing accidents between pedestrians with attributes for which attention - calling is desired preferentially and vehicles.

[0041] Furthermore, it may be a flow in which judgments are made only for one or two of the attributes among the elderly, people operating mobile terminals, and children. Depending on the usage environment of the attention - calling method selection device, for example, according to the regional characteristics such as a bus route passing through a school route, it is possible to change to a flow in which only the attributes with high necessity are judged. By reducing the types of attributes for which judgments are made and enabling attention - calling with a short flow, it is more effective in preventing accidents.

[0042] Next, with reference to FIGS. 8, 9, and 10, the operations regarding the elderly, people operating mobile terminals, and children will be described respectively.

[0043] First, the operation regarding the elderly will be described with reference to FIG. 8, and the judgment criteria and attention - calling methods regarding the elderly will also be described. FIG. 8 is a flowchart of the operation regarding the elderly.

[0044] First, from step S31 to step S34, it is judged whether the attribute of the pedestrian is an elderly person. This corresponds to step S3 in FIG. 7.

[0045] Here, the criteria for determining whether the pedestrian's attribute is an elderly person are the pedestrian's posture and moving speed. When the pedestrian's posture is a forward stoop and the feet are not lifted, and the moving speed is 1.0 m / s or less, the determination unit 102 determines that the pedestrian is an elderly person. Whether the posture is a forward stoop is determined, for example, based on the kyphosis index. Also, whether the feet are lifted is determined, for example, based on the range of motion of the hip joint. Regarding the moving speed of the elderly, it is set based on the decrease in walking speed due to aging, but it is not limited to this speed.

[0046] In step S31, the determination unit 102 detects the pedestrian's posture from the video. Then, it determines whether the posture is a forward stoop and the feet are not lifted (step S32). If the pedestrian's posture is not a forward stoop or the feet are lifted (NO in step S32), the determination unit determines that the pedestrian is not an elderly person, and the operations related to the elderly person end. Then, it proceeds to step S5.

[0047] If the pedestrian's posture is a forward stoop and the feet are not lifted (YES in step S32), the determination unit 102 detects the pedestrian's moving speed (step S33). Then, it determines whether the moving speed is 1.0 m / s or less (step S34). If the moving speed is faster than 1.0 m / s (YES in step S34), the determination unit determines that the pedestrian is not an elderly person, and the operations related to the elderly person end. Then, it proceeds to step S5.

[0048] If the moving speed is 1.0 m / s or less (NO in step S34), the determination unit determines that the pedestrian is an elderly person, and proceeds to step S41.

[0049] Here, in the present embodiment, the determination unit 102 detects the posture of the pedestrian (step S31), and after making a determination in step S32, detects the moving speed of the pedestrian (step S33). However, in step S31, the determination unit 102 may detect the posture of the pedestrian and the moving speed of the pedestrian simultaneously. Alternatively, a flow of detecting the posture of the pedestrian after detecting the moving speed of the pedestrian first may also be used.

[0050] Next, in step S41, a determination is made regarding the possibility of the elderly person running out. FIG. 11 is a table showing the criteria for determining the possibility of running out corresponding to each attribute. The criterion for determining the possibility of the elderly person running out is the posture of the elderly person. Specifically, it is determined whether the angle of the elderly person's face is downward from 0°. Here, the ground and the horizontal direction are set as 0°. When the face is facing downward from 0°, it is considered that the elderly person is walking while looking down and has insufficient attention to the vehicle. Therefore, it is determined that the possibility of running out is high. On the other hand, when the face is not facing downward from 0°, it is considered that the elderly person is paying attention to the vehicle. Therefore, it is determined that there is a possibility of running out. Here, the angle of the face, which is the criterion for determining the possibility of running out, is not limited to 0°, and an angle at which the elderly person can sufficiently pay attention to the vehicle may be set as appropriate.

[0051] Therefore, in step S41, the determination unit 102 uses the posture of the pedestrian detected in step S31 to determine the possibility of the elderly person running out. When the angle of the elderly person's face is downward from 0° (YES in step S41), the determination unit 102 determines that the possibility of the elderly person running out is high.

[0052] When the attribute of the pedestrian is an elderly person and it is determined that the possibility of running out is high, the methods of alerting are: to display on the front sign 121 or the rear sign 122 that the possibility of the elderly person running out is high in order to prompt the driver to pay attention, and to output a sound that is easy for the elderly person to hear from the front speaker 131 or the rear speaker 132 in order to convey to the elderly person that it is dangerous. These are the two points.

[0053] Therefore, in step S42, the selection unit 103 selects to display on the sign that the elderly person is likely to jump out, and the output unit 104 outputs an instruction to display on the sign. Using an example of the display on the front sign 121 or the rear sign 122 shown in FIG. 5, the specific content of the display will be described. When the possibility of the elderly person jumping out is high, the border 12a of the sign is displayed in red. And the character 12b changes the color of the character alternately between red and black to display a warning message. Also, the background 12c of the character 12b is displayed in yellow. The warning message and colors are not limited to this.

[0054] Furthermore, in order to convey to the elderly person that it is dangerous, the selection unit 103 selects to output a sound that is easy for the elderly person to hear from the front speaker 131 or the rear speaker 132, and the output unit 104 outputs an instruction (step S43). When the detection unit 101 detects a pedestrian in front of the bus, a warning is given using the front speaker 131, and when the detection unit 101 detects a pedestrian behind the bus, a warning is given using the rear speaker 132. Also, as a sound that is easy for the elderly person to hear, for example, a sweep sound that decreases from 2000 Hz to 200 Hz can be considered. The warning sound is not limited to this, and the output unit 104 outputs a warning sound with a frequency that is easy for the elderly person with hearing loss due to aging to hear.

[0055] On the other hand, when the angle of the elderly person's face is not downward from 0° (NO in step S41), the determination unit 102 determines that the elderly person may jump out. The method of giving a warning when the attribute of the pedestrian is an elderly person and it is determined that there is a possibility of jumping out is to display on the front sign 121 or the rear sign 122 that the elderly person may jump out. By displaying on the front sign 121 or the rear sign 122, the drivers of following vehicles and oncoming vehicles are alerted.

[0056] Therefore, the selection unit 103 selects to display on the signage that the elderly are likely to run out, and the output unit 104 outputs an instruction to display on the signage (step S44). When the elderly are likely to run out, in FIG. 5, the border 12a of the signage is black, and the character 12b is displayed in black. The background 12c of the character is yellow. The content of the warning message is not limited to this, and the content of the message may be changed when the possibility of running out is high.

[0057] When the above step S43 or step S44 is performed, the operation regarding the elderly is completed, and the process proceeds to step S9. The above steps S41 to S44 correspond to step S4 in FIG. 7.

[0058] Next, with reference to FIG. 9, the operation regarding the person operating the mobile terminal will be described, and the judgment criteria and the method of raising awareness regarding the person operating the mobile terminal will also be described. FIG. 9 is a flowchart of the operation regarding the person operating the mobile terminal.

[0059] First, in steps S51 to S54, it is determined whether the attribute of the pedestrian is a person operating the mobile terminal. This corresponds to step S5 in FIG. 7.

[0060] Here, the judgment criteria for whether the attribute of the pedestrian is a person operating the mobile terminal are the posture and the walking style of the pedestrian. Specifically, it will be described. First, the criterion for the posture is whether both hands or one hand is placed in front of the face and the face is directed downward by more than 0°. Further, the criterion for the walking style is whether the person is walking in a zigzag manner. When the width of the pedestrian's zigzag is, for example, 0.39 m or more, it is determined that the walking style is zigzag. The width of the zigzag for determining that the walking style is zigzag is not limited to this. When both hands or one hand is placed in front of the face, the face is directed downward by more than 0°, and the walking style is zigzag, the determination unit 102 determines that the pedestrian is a person operating the mobile terminal.

[0061] In step S51, the determination unit 102 detects the posture of the pedestrian from the video. Then, it determines whether both hands or one hand is placed in front of the face and the face is directed downward by 0° (step S52). If both hands or one hand of the pedestrian is not in front of the face, or the face is not directed downward by 0° (NO in step S52), it is determined that the pedestrian is not a person operating the mobile terminal, and the operation related to the person operating the mobile terminal ends. Then, it proceeds to step S5.

[0062] If both hands or one hand of the pedestrian is placed in front of the face and the face is directed downward by 0° (YES in step S52), the determination unit 102 detects the walking style of the pedestrian (step S53). Then, it determines whether the walking style of the pedestrian is meandering (step S54). If the walking style of the pedestrian is not meandering (NO in step S54), the operation related to the person operating the mobile terminal ends, and it proceeds to step S7.

[0063] If the walking style of the pedestrian is meandering (YES in step S54), it is determined that the pedestrian is a person operating the mobile terminal, and it proceeds to step S61.

[0064] Here, in this embodiment, after the determination unit 102 detects the posture of the pedestrian (step S51) and makes a determination in step S52, it detects the walking style of the pedestrian (step S53). However, in step S51, a flow in which the determination unit 102 simultaneously detects the posture and the walking style of the pedestrian may also be used. Also, the posture of the pedestrian used in step S51 may be detected in step S31 of the operation related to the elderly. Additionally, a flow in which the walking style of the pedestrian is detected first and then the posture of the pedestrian is detected may also be used.

[0065] Next, in step S61, a determination is made regarding the possibility that a person operating the mobile terminal will jump out. From the table shown in FIG. 11, the criterion for determining the possibility that a person operating the mobile terminal will jump out is the posture of the pedestrian. Specifically, the determination unit 102 detects whether the face is raised one or more times within a predetermined time, and determines the possibility that the person operating the mobile terminal will jump out. If the pedestrian is the person operating the mobile terminal and has not raised the face one or more times within the predetermined time, it is determined that the person is not paying attention to the front and the possibility of jumping out is high. On the other hand, even if the pedestrian is the person operating the mobile terminal, if the face is raised one or more times within the predetermined time, it is determined that the person is paying attention to the front, and the determination unit 102 determines that there is a possibility of jumping out.

[0066] Therefore, in step S61, using the posture of the pedestrian detected in step S51, the determination unit 102 determines the possibility that the person operating the mobile terminal will jump out. If the person operating the mobile terminal has not raised the face one or more times within the predetermined time (YES in step S61), the determination unit 102 determines that the possibility that the person operating the mobile terminal will jump out is high.

[0067] When the attribute of the pedestrian is the person operating the mobile terminal and it is determined that the possibility of jumping out is high, the method of attracting attention is, in order to prompt the driver to pay attention, to display on the front sign 121 or the rear sign 122 that the possibility that the person operating the mobile terminal will jump out is high, and, in order to convey to the person operating the mobile terminal that it is dangerous, to output a warning sound from the front speaker 131 or the rear speaker 132. These are the two points.

[0068] Therefore, the selection unit 103 selects to display on the front sign 121 or the rear sign 122 that the possibility that the person operating the mobile terminal will jump out is high, and the output unit 104 outputs an instruction to display on the front sign 121 or the rear sign 122 (step S62). The specific display by the front sign 121 or the rear sign 122 may be the same as in the case of the elderly, or may be changed.

[0069] Furthermore, in order to inform the person operating the mobile terminal that it is dangerous, the selection unit 103 selects to output a warning sound through the speaker, and the output unit 104 outputs an instruction (step S63). When the detection unit 101 detects a pedestrian in front of the bus, a warning is given using the front speaker 131, and when the detection unit 101 detects a pedestrian behind the bus, a warning is given using the rear speaker 132. Here, the warning sound output through the front speaker may output a message prompting attention to the vehicle, or may output only a sound such as "beep beep".

[0070] On the other hand, when the person operating the mobile terminal raises their face one or more times within a predetermined time (NO in step S61), the determination unit 102 determines that the person operating the mobile terminal may jump out. The method of raising a warning when the attribute of the pedestrian is a person operating the mobile terminal and it is determined that there is a possibility of jumping out is to display on the front sign 121 or the rear sign 122 that the person operating the mobile terminal may jump out. By displaying on the front sign 121 or the rear sign 122, the drivers of following vehicles and oncoming vehicles are prompted to pay attention.

[0071] Therefore, the selection unit 103 selects to display on the sign that the person operating the mobile terminal is likely to jump out, and the output unit 104 outputs an instruction to display on the sign (step S64). The specific display by the front sign 121 or the rear sign 122 may be the same as in the case of the elderly, or may be changed.

[0072] When the above step S63 or step S64 is performed, the operation regarding the person operating the mobile terminal ends, and the process proceeds to step S9. The above steps S61 to S64 correspond to step S6 in FIG. 7.

[0073] Next, with reference to FIG. 10, the operation regarding children will be described, and the judgment criteria and warning methods regarding children will also be described. FIG. 10 is a flowchart of the operation regarding children.

[0074] First, from step S71 to step S72, it is determined whether the attribute of the pedestrian is a child. This corresponds to step S7 in FIG. 7.

[0075] Here, the criterion for determining whether the attribute of the pedestrian is a child is the height of the pedestrian. When the height of the pedestrian is 152.8 cm or less, the determination unit 102 determines that the pedestrian is a child. The height determination criterion of 152.8 cm is set based on the average height of 12-year-olds, but it is not limited to this.

[0076] In step S71, the determination unit 102 detects the height of the pedestrian from the video. Then, it is determined whether the height is 152.8 cm or less (step S72). If the height of the pedestrian is greater than 152.8 cm (NO in step S72), the determination unit 102 determines that the pedestrian is not a child, and the operations related to children end. Then, it proceeds to step S9.

[0077] If the height of the pedestrian is 152.8 cm or less (YES in step S72), the determination unit 102 determines that the pedestrian is a child, and proceeds to step S81.

[0078] Next, in step S81, it is determined whether there is a possibility that the child will run out. From the table shown in FIG. 11, the criterion for determining the possibility that the child will run out is the moving speed of the child. When the moving speed of the child is 2.0 km / h or more, since it is considered that the child is running, it is determined that the possibility of running out is high. On the other hand, when the moving speed is less than 2.0 km / h, since it is considered that the child is walking, it is determined that there is a possibility of running out. The moving speed, which is the criterion for determining the possibility that the child will run out, is not limited to this speed.

[0079] Therefore, in step S81, the determination unit 102 detects the moving speed of the child. Then, it makes a determination regarding the possibility that the child will run out (step S82). When the moving speed of the child is 2.0 km / h or more (YES in step S82), the determination unit 102 determines that the possibility that the child will run out is high.

[0080] When the attribute of the pedestrian is a child and it is determined that there is a high possibility of running out, the method of raising awareness is to display on the rear sign 122 that there is a high possibility of the child running out in order to prompt the driver to pay attention, and to output a sound that is easy for the child to understand from the front speaker 131 in order to convey the danger to the child. These are the two points.

[0081] Therefore, in step S83, the selection unit 103 selects to display on the sign that there is a high possibility of the child running out, and the output unit 104 outputs an instruction to display on the sign. FIG. 5 shows an example of the display on the front sign 121 or the rear sign 122. The specific display by the front sign 121 or the rear sign 122 may be the same as in the case of the elderly, or may be changed.

[0082] Furthermore, the selection unit 103 selects to output from the speaker a sound that is easy for the child to understand in order to convey the danger to the child, and the output unit 104 outputs an instruction (step S84). Here, the sound that is easy for the child to understand may be a message that prompts attention forward using plain words. As a message in plain words, for example, messages such as "Look ahead" and "A car is coming" can be considered. Also, by outputting the sound of a single word instead of a sentence, it may prompt the surroundings forward. As the sound of a single word, for example, words such as "Dangerous" and "Stop" can be considered. Furthermore, the voices of these messages may be output in the voices of characters well-known to children. In addition, music from anime well-known to children may be output.

[0083] On the other hand, when the moving speed of the child is less than 2.0 km / h (NO in step S82), the determination unit 102 determines that there is a possibility that the child will run out. When the attribute of the pedestrian is a child and it is determined that there is a possibility of running out, the method of raising awareness is to display on the front sign 121 or the rear sign 122 that there is a possibility that the child will run out. By displaying on the front sign 121 or the rear sign 122, it prompts the driver of the following vehicle or the oncoming vehicle to pay attention.

[0084] Therefore, the selection unit 103 selects to display on the sign that there is a high possibility that a child will jump out, and the output unit 104 outputs an instruction to display on the sign (step S85). The specific display by the front sign 121 or the rear sign 122 may be the same as that for the elderly, or may be changed.

[0085] When step S84 or step S85 is performed, the operation regarding the child ends, and the process proceeds to step S9. The above steps S81 to S85 correspond to step S8 in FIG. 7.

[0086] The attention - calling method selection device 10 in the above - described embodiment includes a detection unit 101 that detects pedestrians in front of and behind the bus based on an image captured by cameras installed in front of and behind the bus while the bus is stopped, a determination unit 102 that, when the detection unit 101 detects a pedestrian, determines the possibility that the pedestrian will jump out based on the physical state of the pedestrian detected from the image, using the attributes of the pedestrian and the judgment criteria corresponding to the attributes, a selection unit 103 that selects an attention - calling method based on the attributes and the possibility of jumping out determined by the determination unit 102, and an output unit 104 that outputs an instruction corresponding to the attention - calling method selected by the selection unit 103. Since the determination unit 102 determines the possibility that a pedestrian will jump out based on the attributes of the pedestrian and the judgment criteria corresponding to the attributes, an appropriate attention - calling method can be selected.

[0087] And by selecting and outputting an appropriate attention - calling method, it is possible to appropriately alert the driver of the vehicle and the pedestrians in front of and behind the bus. As a result, it is possible to prevent contact accidents between pedestrians and the vehicle.

[0088] In this embodiment, the determination unit 102 uses the physical state of the pedestrian to determine whether the attribute of the pedestrian is any one of an elderly person, a person operating a mobile terminal, or a child. Pedestrians with these attributes are considered to be at high risk of being involved in an accident. This is because pedestrians with these attributes are considered to be insufficiently attentive to vehicles. Furthermore, when the attribute of the pedestrian is a child, the risk of an accident is considered to be high because it is difficult for the driver of the vehicle to recognize the child. Therefore, in this embodiment, it is possible to alert pedestrians with attributes that are at high risk of being involved in an accident. Furthermore, by selecting an alert method according to the attribute of the pedestrian, it is possible to effectively alert pedestrians in dangerous situations.

[0089] Also, in this embodiment, the physical state of the pedestrian used by the determination unit 102 for determining the attribute and the possibility of jumping out includes at least any one of posture, walking style, height, or moving speed. For detecting these physical states, the determination unit 102 of this embodiment uses the video captured by the camera, particularly a moving image. By using the above physical states, it is possible to make a determination not only about the attribute of the pedestrian but also about the possibility of jumping out.

[0090] By being able to judge not only the attributes of pedestrians but also the possibility of them darting out, an appropriate alert method can be selected. The alert method selection device of this embodiment selects an alert method based on the attributes of pedestrians and the possibility of pedestrians darting out. The effect of selecting an alert method based on the attributes of pedestrians is as described above. And by selecting an alert method based on the possibility of darting out, it is possible to prompt appropriate risk avoidance actions for the vehicle driver and pedestrians. For example, even if an alert is given, if the driver does not recognize the high level of danger, sufficient attention may not be paid to pedestrians and an accident may occur. Similarly, pedestrians may also be at risk of accidents due to misrecognition of danger. On the other hand, despite an alert being given in a situation where the danger level is not high, if the driver perceives it as a very dangerous situation, excessive risk avoidance actions such as sudden braking may be taken, potentially triggering other accidents. However, the alert method selection device of this embodiment can perform appropriate alerting to the driver by selecting an alert method based on the possibility of pedestrians darting out. As a result, the risk of accidents is appropriately recognized by pedestrians and the driver. As a result, pedestrians and the driver can take appropriate risk avoidance actions. Therefore, accidents caused by insufficient attention of the driver and pedestrians can be prevented. Furthermore, the possibility of accidents caused by the driver's excessive risk avoidance actions can also be reduced.

[0091] Furthermore, the attention - calling method selected by the selection unit 103 is changed according to the attributes and the possibility of popping out. For example, when the possibility of an elderly person popping out is high, the selection unit 103 selects, as the attention - calling method, to output a sound that is easy for the elderly person to hear. By outputting a sound that is easy for the elderly person to hear, appropriate attention can be called to the elderly person. In addition, the selection unit 103 selects, as the attention - calling method, to output to the signage installed in front of and behind the bus. And the display on the signage varies according to the possibility of a pedestrian popping out. By changing the display on the signage according to the possibility of popping out, appropriate attention can be called to pedestrians and drivers. As a result, pedestrians and drivers can take appropriate danger - avoidance actions, and contact accidents can be prevented.

[0092] [Modification Example of the First Embodiment] A modification example of the first embodiment of the present disclosure will be described. Hereinafter, within the scope where the description of this modification example is not unclear, the description of the content overlapping with the above - mentioned description will be omitted.

[0093] In this modification example, the detection unit 101 also detects pedestrians and vehicles. And when the detection unit 101 detects a vehicle, the determination unit 102 uses the physical state of the pedestrian to determine the attribute of the pedestrian and the possibility of the pedestrian popping out based on the determination criteria corresponding to the attribute. Specifically, even when a pedestrian is detected, in a situation where no vehicle is detected, that is, in a situation where there is no danger of a contact accident, the configuration is such that the attribute of the pedestrian and the possibility of popping out are not determined.

[0094] In the configuration of this modification example, unnecessary attention - calling can be reduced. That is, even when a pedestrian is detected, in a situation where no vehicle is detected, since there is no danger of a contact accident, attention - calling is unnecessary. Such unnecessary attention - calling can be reduced. By reducing attention - calling in unnecessary situations, attention - calling in situations where the risk of contact accidents is high can be carried out more effectively.

[0095] As another modification example, the detection unit 101 may detect a plurality of pedestrians. First, the detection unit 101 detects a plurality of pedestrians in front of and behind the bus. Then, when the detection unit 101 detects a plurality of pedestrians, the determination unit 102 detects the physical states of the plurality of pedestrians. Then, based on the physical states of the plurality of pedestrians, one pedestrian to be determined as the target for attribute and possibility of running out determination is determined. Then, for the pedestrian who is the target for determination, similar to the first embodiment, determination of the attribute and the possibility of running out is performed.

[0096] As a method for determining the pedestrian who is the target for determination, a method using the moving speeds of a plurality of pedestrians can be considered. The determination unit 102 detects the moving speeds of the plurality of pedestrians. Then, the pedestrian with the fastest moving speed among the plurality of pedestrians is determined as the pedestrian who is the target for determination, and the attribute and the possibility that the pedestrian based on the determination criteria corresponding to the attribute runs out are determined.

[0097] By the determination unit 102 determining the possibility that the pedestrian based on the determination criteria corresponding to the attribute of the pedestrian with the fastest moving speed among the plurality of pedestrians runs out, a contact accident with the vehicle can be prevented. This is because it is considered that the pedestrian with the fastest moving speed has the highest possibility of first invading the vehicle's path among the plurality of pedestrians. That is, it is considered that the pedestrian is the one with the highest risk of being involved in a contact accident with the vehicle. Therefore, it is effective in preventing accidents by determining the attribute and the possibility of running out for the pedestrian with a high accident risk and giving a warning.

[0098] As another method for determining the pedestrian who is the target for determination, a method using the heights of a plurality of pedestrians can also be considered. The determination unit 102 detects the heights of the plurality of pedestrians. Then, the person with the smallest height among the plurality of pedestrians is determined as the pedestrian who is the target for determination.

[0099] The pedestrian with the smallest height is considered to be hidden behind the bus and the most difficult to be seen by the driver. Therefore, by determining the attribute and the possibility of running out for the pedestrian with the smallest height and giving a warning, the possibility of preventing a contact accident between the vehicle and the pedestrian can be increased.

[0100] [Second Embodiment] Next, a second embodiment of the present disclosure will be described in detail with reference to the drawings. Hereinafter, the description of the overlapping content with the above description will be omitted as long as the description of this embodiment is not unclear.

[0101] FIG. 12 is a block diagram showing the configuration of a warning system including a warning method selection device 20 according to the second embodiment of the present disclosure.

[0102] Referring to FIG. 12, the warning method selection device 20 in this embodiment further includes a recording unit 205 as compared with the warning method selection device of the first embodiment.

[0103] The recording unit 205 in the warning method selection device 20 is an aspect of a recording means for recording the number of times the output unit 104 outputs an instruction according to the warning method and the information of the output location. Specifically, the number of times the output unit 104 outputs an instruction according to the warning method is linked and recorded with the information of the location where the instruction according to the warning method is output. FIG. 13 is an example of the data stored in the recording unit 205. As the information of the location where the instruction according to the warning method is output, the name of the bus stop is recorded. Also, the number of times the warning method is output is recorded in association with each bus stop name. The location information may be the name of the bus route, the destination, or the location information of the bus stop. Further, the number of times the instruction according to the warning method is output may be recorded for each attribute or for each possibility of popping up.

[0104] And when there is a location where the number of times the instruction is output exceeds a predetermined number of times, the recording unit 205 transmits the information of the location where the output number exceeds the predetermined number of times to the output unit 104. Here, the predetermined number of times is a number set as needed, and different numbers of times may be set for each bus stop.

[0105] The output unit 104 receives information on locations where the number of outputs has exceeded a predetermined number from the recording unit. Then, it transmits the location information to a database 251 such as an external map service. This map service includes maps used by pedestrians and drivers for route search, as well as maps managed by bus companies, prefectures, cities, and towns.

[0106] Based on the location information received from the output unit 104, the map service displays dangerous locations, such as dangerous bus stops, on the map. An example of a screen showing a dangerous bus stop on the map is shown in FIG. 14. Additionally, it is also conceivable to display a list of dangerous bus stops on the homepage of the bus company or the homepage of the city, town, or village based on the location information transmitted by the output unit 104.

[0107] The warning method selection device 10 in the above-described embodiment includes a recording unit 205 that records the number of times the output unit 104 has output an instruction according to the warning method and the location information of the output. The output unit 104 outputs information on locations where the number of times recorded by the recording unit 205 has exceeded a predetermined number. A location where the number of instructions output according to the warning method has exceeded a predetermined number is considered to be a location with a high risk of contact accidents. Therefore, by outputting information on such dangerous locations, it is possible to widely inform the public about dangerous locations. As a result of wide dissemination, pedestrians and drivers pay attention, making it possible to prevent contact accidents.

[0108] [Modification Example of the Second Embodiment] A modification example of the first embodiment of the present disclosure will be described. Hereinafter, descriptions of overlapping content with the above description will be omitted as long as the description of this modification example is not made unclear.

[0109] In this modification example, based on the information on locations where the number of times the recording unit 205 has recorded an instruction output according to the warning method has exceeded a predetermined number, the detection unit 101 detects pedestrians.

[0110] Specifically, the recording unit 205 transmits information on locations where the number of attention calls has exceeded a predetermined number to the detection unit 101. Then, upon receiving the location information, the detection unit 101 detects pedestrians based on the video captured by the camera when the bus stops at that location.

[0111] When the bus stops at a location where the number of attention calls recorded by the recording unit exceeds a predetermined number, the detection unit 101 detects pedestrians in front of and behind the bus. This configuration enables prevention of accidents at dangerous locations.

[0112] [Hardware Configuration Example] FIG. 15 is a diagram showing an example of a hardware configuration in which the attention call method selection device 10 in the present disclosure is realized by a computer 30 including a processor. As shown in FIG. 15, the attention call method selection device 10 includes a processor 301, a ROM (Read Only Memory) 302, a RAM (Random Access Memory) 303, a storage device 304 such as a hard disk for storing programs, a communication interface for network connection, and an input / output interface for inputting and outputting data. Each component is connected via a bus 307.

[0113] The processor 301 operates an operating system to control the entire computer. The processor 301 includes, for example, a CPU (Central Processing Unit), a DSP (Digital Signal Processor), a GPU (Graphics Processing Unit), etc. The processor loads programs stored in the storage device 304, ROM 302, etc. Then, the processor 301 executes each process coded in the program. Also, the processor 301 functions as part or all of the computer 30. And the processor 301 may execute the processes or instructions in the illustrated flowchart based on the program.

[0114] The ROM 302 stores application programs, programs according to each embodiment, and the like. Also, the RAM 303 is used as a work area for the processor 801. And the storage device 304 includes, for example, a semiconductor memory such as a flash memory, an HDD (Hard Disk Driver), and the like. The storage device 304 stores, for example, programs of an OS (Operating System), application programs, programs according to each embodiment, and the like.

[0115] The input / output interface 305 is connected to peripheral devices. The connection method may be wired or wireless. Examples of peripheral devices include a camera, a speaker, a display, and signage.

[0116] The communication interface 306 is connected to a communication network such as a LAN (Local Network) or a WAN (Wide Area Network) through a wireless or wired communication line. Note that the communication network may be composed of a plurality of communication networks. Thereby, the computer is connected to an external device via the communication network.

[0117] The above is an example of the hardware configuration of a computer. The computer may have some of the configurations shown in FIG. 15, or the computer may have configurations other than those shown in FIG. 15.

[0118] Although the present disclosure has been described with reference to the above embodiments, the present disclosure is not limited to the above embodiments. The configurations and details of each present disclosure may include embodiments to which various changes that can be grasped by those skilled in the art within the scope of the present disclosure are applied. The present disclosure may include embodiments in which the matters described in this specification are appropriately combined or replaced as necessary. For example, the matters described using a specific embodiment can also be applied to other embodiments within a range where no contradiction occurs. For example, although a plurality of operations are described in order in the form of a flowchart, the described order does not limit the order of executing the plurality of operations. Therefore, when implementing each embodiment, the order of the plurality of operations can be changed within a range that does not substantially interfere with the content thereof.

[0119] Some or all of the above embodiments can also be described as follows. However, some or all of the above embodiments are not limited to the following.

[0120] (Appendix 1) Detection means for detecting pedestrians in front of and behind the bus based on an image captured by a camera installed in front of and behind the bus while the bus is stopped; When the detection means detects the pedestrian, determination means for determining the possibility that the pedestrian jumps out based on the body state of the pedestrian detected from the image, based on the attribute of the pedestrian and the judgment criterion corresponding to the attribute; Selection means for selecting a warning method based on the attribute and the possibility of jumping out determined by the determination means; Output means for outputting an instruction corresponding to the warning method selected by the selection means A warning method selection device comprising: (Appendix 2) The body state of the pedestrian includes at least any one of posture, walking style, height, or moving speed The warning method selection device according to Appendix 1. (Appendix 3) The determination means determines, using the physical state of the pedestrian, whether the attribute of the pedestrian is any one of an elderly person, a person operating a portable terminal, or a child. The attention-calling method selection device according to appendix 1 or 2. (Appendix 4) The physical state of the pedestrian includes the moving speed and posture of the pedestrian. The determination means determines, based on the moving speed and the posture, whether the attribute of the pedestrian is an elderly person. The attention-calling method selection device according to appendix 3. (Appendix 5) When the attribute of the pedestrian is determined to be an elderly person, the determination means determines the possibility of running out based on the posture of the pedestrian. The attention-calling method selection device according to appendix 4. (Appendix 6) The determination means determines that the possibility of running out is high. The selection means selects, as the attention-calling method, to output a sound that is easy for the elderly to hear. The attention-calling method selection device according to appendix 3 or 5. (Appendix 7) The physical state of the pedestrian includes the walking manner and posture of the pedestrian. The determination means determines, based on the walking manner and the posture, whether the attribute of the pedestrian is a person operating a portable terminal. The attention-calling method selection device according to appendix 3. (Appendix 8) When the attribute of the pedestrian is a person operating a portable terminal, the determination means determines the possibility of running out based on the walking manner of the pedestrian. The attention-calling method selection device according to appendix 7. (Appendix 9) The physical state of the pedestrian includes the height and moving speed of the pedestrian. The determination means determines that the attribute of the pedestrian is a child based on the height, and determines the possibility of running out based on the moving speed. The attention-calling method selection device according to appendix 3. (Appendix 10) When the determination means determines that the attribute of the pedestrian is a child and the possibility of running out is high, the selection means selects to output a sound that is easy for children to understand as the attention - attracting method. An attention - attracting method selection device according to Supplementary Note 3 or 9. (Supplementary Note 11) The selection means selects to output to a sign installed in front of and behind the bus as the attention - attracting method, The display on the sign varies according to the possibility of the pedestrian running out. An attention - attracting method selection device according to any one of Supplementary Notes 1 to 10. (Supplementary Note 12) The detection means further detects a vehicle, When the detection means detects the vehicle, the determination means uses the physical state of the pedestrian detected from the video to determine the attribute of the pedestrian and the possibility of the pedestrian running out based on the determination criteria corresponding to the attribute. An attention - attracting method selection device according to any one of Supplementary Notes 1 to 11. (Supplementary Note 13) The detection means detects a plurality of pedestrians, When the detection means detects the plurality of pedestrians, the determination means determines the attribute of the pedestrian with the fastest moving speed among the plurality of pedestrians and the possibility of the pedestrian running out based on the determination criteria corresponding to the attribute. An attention - attracting method selection device according to any one of Supplementary Notes 2 to 12. (Supplementary Note 14) It includes a recording means for recording the number of times the output means outputs an instruction according to the attention - attracting method and the information of the output location, The output means outputs the information of the location where the number of times recorded by the recording means exceeds a predetermined number of times. An attention - attracting method selection device according to any one of Supplementary Notes 1 to 13. (Supplementary Note 15) When the bus stops at a location where the number of times recorded by the recording means exceeds a predetermined number of times, The detection means detects pedestrians in front of and behind the bus. The warning method selection device described in Supplementary Note 14. (Supplementary Note 16) Based on the video captured by cameras installed in front of and behind the bus while the bus is stopped, pedestrians in front of and behind the bus are detected. When the pedestrians are detected, using the physical states of the pedestrians detected from the video, the attributes of the pedestrians and the possibility of the pedestrians jumping out based on the judgment criteria corresponding to the attributes are judged. Based on the judged attributes and the possibility of jumping out, a warning method is selected. An instruction corresponding to the selected warning method is output. Warning method selection method. (Supplementary Note 17) On a computer, Based on the video captured by cameras installed in front of and behind the bus while the bus is stopped, pedestrians in front of and behind the bus are detected. When the pedestrians are detected, using the physical states of the pedestrians detected from the video, the attributes of the pedestrians and the possibility of the pedestrians jumping out based on the judgment criteria corresponding to the attributes are judged. Based on the judged attributes and the possibility of jumping out, a warning method is selected. An instruction corresponding to the selected warning method is output. A program for executing the process.

Explanation of Signs

[0121] 10, 20 Warning method selection device 101 Detection unit 102 Judgment unit 103 Selection unit 104 Output unit 111 Front camera 112 Rear camera 121 Front signage 122 Rear signage 131 Front speaker 132 Rear speaker 142 Rear display 205 Recording unit 251 Database 30 Computer 301 Processor 302 ROM 303 RAM 304 Storage Device 305 Input / Output Interface 306 Communication Interface 307 Bus

Claims

1. Detection means for detecting pedestrians in front of and behind the bus based on an image captured by a camera installed in front of and behind the bus while the bus is stopped; When the detection means detects the pedestrian, using the physical state of the pedestrian detected from the image, determining the attribute of the pedestrian, and based on a judgment criterion corresponding to the determined attribute, Judgment means for judging the possibility that the pedestrian will jump out; Selection means for selecting a warning method based on the attribute judged by the judgment means and the possibility that the pedestrian will jump out; Output means for outputting an instruction corresponding to the warning method selected by the selection means In a warning method selection device comprising: The physical state of the pedestrian includes at least one of posture, walking style, height, and moving speed; When the detection means detects a plurality of the pedestrians, the judgment means is based on the attribute of the pedestrian with the fastest moving speed among the plurality of pedestrians and the judgment criterion corresponding to the attribute. , Judge the possibility that the pedestrian will jump out Warning method selection device.

2. The judgment means uses the physical state of the pedestrian to judge whether the attribute of the pedestrian with the fastest moving speed is an elderly person, a person operating a portable terminal, a child, or any other. The warning method selection device according to claim 1.

3. The judgment means judges whether the attribute of the pedestrian is an elderly person based on the moving speed and the posture. The warning method selection device according to claim 2.

4. The judgment means judges whether the attribute of the pedestrian is a person operating a portable terminal based on the walking style and the posture. The warning method selection device according to claim 2.

5. The judgment means judges whether the attribute of the pedestrian is a child based on the height and the moving speed. The warning method selection device according to claim 2.

6. When the selection means determines that the attribute of the pedestrian is an elderly person and the possibility of jumping out is high, as the warning method, it is selected to output a sound that is easy for the elderly to hear. The warning method selection device according to claim 3.

7. The selection means selects to output a warning sound as the warning method when it is determined that the attribute of the pedestrian is a person operating a portable terminal and the possibility of jumping out is high. The warning method selection device according to claim 4.

8. When the selection means determines that the attribute of the pedestrian is a child and the possibility of running out is high, as the attention - calling method, it selects to output a sound that is easy for children to understand. The attention - calling method selection device according to claim 5.

9. The output means further includes a display, The selection means selects to output the video of the camera to the display. The attention - calling method selection device according to claim 1.

10. The output means further includes a sign, The detection means further detects a vehicle, When the detection means detects the vehicle, the selection means selects to output, as the attention - calling method, a message corresponding to the attribute of the pedestrian determined by the determination means to have a high possibility of running out, to the sign. The attention - calling method selection device according to claim 1.

11. It further includes a recording means for recording the number of times an instruction corresponding to the attention - calling method is output and information on the output location, The output means outputs information on the location where the number recorded by the recording means exceeds a predetermined number. The attention - calling method selection device according to any one of claims 1 to 10.

12. Based on the video taken by cameras installed in front of and behind the bus while the bus is stopped, pedestrians in front of and behind the bus are detected, When the pedestrian is detected, the attribute of the pedestrian is determined using the physical state of the pedestrian detected from the video, and based on the judgment criteria corresponding to the determined attribute, the possibility that the pedestrian will run out is judged. Based on the determined attribute and the possibility that the pedestrian will run out, an attention - calling method is selected, An instruction corresponding to the selected attention - calling method is output. In the attention - calling method selection method, the physical state of the pedestrian includes at least one of posture, walking style, height, and moving speed. When a plurality of pedestrians are detected, based on the attribute of the pedestrian with the fastest moving speed among the plurality of pedestrians and the judgment criteria corresponding to the attribute, the possibility that the pedestrian will run out is judged. Attention - calling method selection method.

13. A program for causing a computer to execute the attention - calling method selection method according to claim 12.

Images