Risk estimation device, risk estimation method, and risk estimation program

Abstract

An object detection unit (111) acquires observation data from a sensor device (201) and uses the observation data to detect each of objects present in an observation range of the sensor device (201) as a sensor-detected object, and acquires communication partner data from a wireless device (202) and uses the communication partner data to detect each of objects present in a communication range of the wireless device (202) as a wireless-detected object. A risk estimation unit (112) estimates, on the basis of the information on the sensor-detected objects and the information on the wireless-detected objects, an assumed risk assumed by the presence of the surrounding objects.

Description

Risk estimation device, risk estimation method, and risk estimation program

[0001] The present disclosure relates to a technique for estimating a risk assumed due to the presence of a surrounding object.

[0002] 2. Description of the Related Art Detection systems using sensors such as cameras are used for safety purposes in autonomous vehicles and surveillance systems.

[0003] A known prior art method is to estimate hazard risk by detecting and analyzing objects using images. However, this method has difficulty detecting objects that exist outside the camera's observation range. Another known method is to use low-cost beacons to detect the number and behavior of objects. However, this method does not provide more detailed information than that obtained from beacons.

[0004] Patent Document 1 discloses a technique for setting a risk area based on image information.

[0005] Japanese Patent Application Laid-Open No. 2022-187896

[0006] The present disclosure aims to make it possible to estimate the risk assumed due to the presence of an object outside the observation range of a sensor device.

[0007] The risk estimation device of the present disclosure comprises an object detection unit that acquires observation data from a sensor device, uses the observation data to detect each object present in the observation range of the sensor device as a sensor-detected object, acquires communication partner data from a wireless terminal, and uses the communication partner data to detect each object present in the communication range of the wireless terminal as a wireless-detected object, and a risk estimation unit that estimates an expected risk expected due to the presence of surrounding objects based on information about the sensor-detected objects and information about the wireless-detected objects.

[0008] According to the present disclosure, it is possible to estimate the risk that may be expected due to the presence of an object outside the observation range of a sensor device.

[0009] 1 is a diagram illustrating a configuration of a risk detection system 200 according to a first embodiment. 2 is a diagram illustrating a configuration of a risk estimation device 100 according to the first embodiment. 3 is a flowchart illustrating a risk estimation method according to the first embodiment. 4 is a flowchart illustrating step S140 according to the first embodiment. 5 is a diagram illustrating an example of a situation of Case 1 according to the first embodiment. 6 is a table illustrating risk estimation (S141, S142) according to the first embodiment. 7 is a diagram illustrating an example of a situation of Case 2 according to the first embodiment. 8 is a table illustrating risk estimation (S143, S144) according to the first embodiment. 9 is a diagram illustrating an example of a situation of Case 3 according to the first embodiment. 10 is a table illustrating risk estimation (S145, S146) according to the first embodiment. 11 is a diagram illustrating an overview of a risk estimation method according to the first embodiment. 12 is a flowchart illustrating a risk estimation method according to the second embodiment. 13 is a flowchart illustrating step S240 according to the second embodiment. 14 is a diagram illustrating an overview of a risk estimation method according to the second embodiment. 15 is a flowchart illustrating a risk estimation method according to the third embodiment. 16 is a flowchart illustrating step S340 according to the third embodiment. 17 is a diagram illustrating an overview of a risk estimation method according to the third embodiment. 18 is a diagram illustrating a hardware configuration of a risk estimation device 100 according to an embodiment.

[0010] In the embodiments and drawings, the same or corresponding elements are denoted by the same reference numerals. The description of elements denoted by the same reference numerals as those already described will be omitted or simplified as appropriate. Arrows in the drawings primarily indicate the flow of data or the flow of processing.

[0011] First Embodiment A risk estimation device 100 will be described with reference to FIGS.

[0012] ***Description of Configuration*** The configuration of the risk detection system 200 will be described with reference to Fig. 1. The risk detection system 200 includes a sensor device 201, a wireless device 202, and a risk estimation apparatus 100.

[0013] For example, the risk detection system 200 is mounted on a mobile object such as an automobile or a robot. Examples of the robot include a PMV or an AMR. PMV is an abbreviation for personal mobility vehicle. AMR is an abbreviation for autonomous transport robot.

[0014] For example, the risk detection system 200 is mounted on infrastructure equipment. An example of infrastructure equipment is an RSU. RSU is an abbreviation for Roadside Unit.

[0015] The sensor device 201 performs various observations of the surrounding environment. The observations include photography, measurement, and the like. Examples of the sensor device 201 include a camera and a LiDAR. LiDAR is an abbreviation for Light Detection And Ranging.

[0016] Data obtained by observation by the sensor device 201 is called observation data.

[0017] The wireless device 202 is a device that communicates wirelessly. The wireless device 202 is also called a transmitting terminal. An example of the wireless device 202 is a BLE beacon or a Wi-Fi device. BLE is an abbreviation for Bluetooth Low Energy. Bluetooth is a registered trademark. Wi-Fi is an abbreviation for Wireless Fidelity. Wi-Fi is a registered trademark.

[0018] Data indicating the communication partner of the wireless device 202 is called communication partner data.

[0019] The configuration of the risk estimation device 100 will be described with reference to Fig. 2. The risk estimation device 100 is a computer equipped with hardware such as a processor 101, a memory 102, an auxiliary storage device 103, and an interface 104. These pieces of hardware are connected to each other via signal lines.

[0020] The processor 101 is an IC that performs arithmetic processing and controls other hardware. For example, the processor 101 is a CPU, a GPU, or a combination of these. IC is an abbreviation for Integrated Circuit. CPU is an abbreviation for Central Processing Unit. GPU is an abbreviation for Graphics Processing Unit.

[0021] The memory 102 is a volatile or non-volatile storage device. The memory 102 is also called a primary storage device or a main memory. For example, the memory 102 is a RAM. Data stored in the memory 102 is saved in the secondary storage device 103 as needed. RAM is an abbreviation for Random Access Memory.

[0022] The auxiliary storage device 103 is a non-volatile storage device. For example, the auxiliary storage device 103 is a ROM, a HDD, a flash memory, or a combination of these. Data stored in the auxiliary storage device 103 is loaded into the memory 102 as needed. ROM is an abbreviation for Read Only Memory. HDD is an abbreviation for Hard Disk Drive.

[0023] The interface 104 is a port to which various devices are connected. For example, a sensor device 201 and a wireless device 202 are connected to the interface 104.

[0024] The risk estimation device 100 includes elements such as an object detection unit 111, a risk estimation unit 112, and a risk notification unit 113. These elements are realized by software.

[0025] The auxiliary storage device 103 stores a risk estimation program for causing the computer to function as an object detection unit 111, a risk estimation unit 112, and a risk notification unit 113. The risk estimation program is loaded into the memory 102 and executed by the processor 101. The auxiliary storage device 103 also stores an OS. At least a portion of the OS is loaded into the memory 102 and executed by the processor 101. The processor 101 executes the risk estimation program while running the OS. OS is an abbreviation for Operating System.

[0026] Data (input data, output data, etc.) of the risk estimation program is stored in the storage unit 120. The memory 102 functions as the storage unit 120. However, a storage unit such as the auxiliary storage unit 103, a register in the processor 101, or a cache memory in the processor 101 may function as the storage unit 120 instead of or together with the memory 102.

[0027] The risk estimation program can be recorded (stored) in a computer-readable manner on a non-volatile recording medium such as an optical disk or flash memory.

[0028] ***Explanation of Operation*** The operational procedure of the risk estimation device 100 corresponds to a risk estimation method. Also, the operational procedure of the risk estimation device 100 corresponds to a processing procedure by a risk estimation program.

[0029] The risk estimation method will be described with reference to Fig. 3. In step S110, the object detection unit 111 acquires observation data from the sensor device 201 and detects a sensor-detected object using the observation data.

[0030] The sensor-detected objects are objects that exist within the observation range (sensor observation range) of the sensor device 201. The sensor observation range is the range of an area that can be observed by the sensor device 201. Examples of sensor-detected objects include automobiles, pedestrians, and bicycles.

[0031] For example, the observation data is image data created by a camera capturing images of the surroundings. The object detection unit 111 processes the image data to detect objects detected by each sensor. For example, the observation data is three-dimensional point cloud data created by a LiDAR measuring the surroundings. The object detection unit 111 processes the three-dimensional point cloud data to detect objects detected by each sensor.

[0032] In step S120, the object detection unit 111 acquires communication partner data from the wireless device 202 and detects a wirelessly detected object using the communication partner data.

[0033] A wirelessly detected object is an object that carries a wireless terminal and exists within the communication range (wireless communication range) of the wireless device 202. A wireless terminal is a terminal that has a function of communicating wirelessly. Examples of wireless terminals are an in-vehicle device and a smartphone. Examples of wirelessly detected objects are a car equipped with an in-vehicle device, a pedestrian carrying a smartphone, a bicycle driven by a person carrying a smartphone, etc.

[0034] For example, the communication partner data indicates the identifier (ID) of each wireless terminal that is a communication partner. The object detection unit 111 refers to the communication partner data and identifies each wireless terminal as a wireless detection object.

[0035] In steps S130 and S140, the risk estimation unit 112 estimates the assumed risk based on the information on the sensor-detected object and the information on the wireless-detected object.

[0036] The assumed risk is a risk assumed due to the presence of a surrounding object. For example, the assumed risk is a risk assumed for a mobile body equipped with the risk detection system 200. For example, the risk estimation unit 112 estimates an assumed risk due to the presence of an object outside the line of sight of the sensor device 201.

[0037] In step S130, the risk estimation unit 112 determines the magnitude relationship between the number of sensor detections and the number of wireless detections.

[0038] The sensor detection number is the number of objects detected by the sensor. The risk estimation unit 112 counts the number of objects detected by the sensor to obtain the sensor detection number. The wireless detection number is the number of objects detected by wireless. The risk estimation unit 112 counts the number of objects detected by wireless to obtain the wireless detection number.

[0039] In step S140, the risk estimation unit 112 estimates an assumed risk depending on the magnitude relationship between the number of sensor detections and the number of wireless detections.

[0040] Details of step S140 will be described with reference to FIG. 4. If the number of sensor detections and the number of wireless detections match, the process proceeds to step S141. In step S141, the risk estimation unit 112 estimates the estimated risk due to the presence of an object within the sensor observation range. In step S142, the risk estimation unit 112 estimates the estimated risk due to the presence of an object outside the sensor observation range.

[0041] FIG. 5 shows an example of a situation in which the number of sensor detections and the number of wireless detections match. The entire range shown in FIG. 5 is included in the wireless communication range. An automobile 210 is a mobile body equipped with a risk detection system 200. The fan-shaped shading represents the sensor observation range. Black people are pedestrians carrying wireless devices and present within the sensor observation range. Shaded people are pedestrians not carrying wireless devices. White people are pedestrians carrying wireless devices and present outside the sensor observation range. Apart from the pedestrians shown in FIG. 5, there are no other objects carrying wireless devices within the wireless communication range.

[0042] There are five pedestrians within the sensor observation range. However, one pedestrian is hidden behind the other pedestrians and is not detected by the sensor device 201. Therefore, the number of sensor detections is "4." There are four pedestrians carrying wireless terminals within the wireless communication range. Therefore, the number of wireless detections is "4." Therefore, the number of sensor detections "4" and the number of wireless detections "4" match.

[0043] If the number of sensor detections and the number of wireless detections match, the following possibilities are possible: 1. In the sensor observation range, an object without a wireless terminal is hidden behind another object and is not detected by the sensor device 201. 2. In the sensor observation range, there is an object without a wireless terminal. This object is detected by the sensor device 201. 3. Outside the sensor detection range, there is an object without a wireless terminal. 4. Outside the sensor detection range, there is an object with a wireless terminal. Because of these possibilities, even if the number of sensor detections and the number of wireless detections match, it is not possible for the sensor device 201 to identify all objects, and it is necessary to continue to be vigilant against risks that exist in the surrounding area.

[0044] The assumed risks estimated in steps S141 and S142 will be described with reference to FIG. 6. Within the sensor observation range (step S141), there is a possibility that an object hidden behind another object will not be detected. Therefore, the risk estimation unit 112 estimates a "contact accident" as the assumed risk. In other words, the risk estimation unit 112 estimates that there is a risk of a contact accident. Outside the sensor observation range (step S142), there is a possibility that an object without a wireless terminal will not be detected. Therefore, the risk estimation unit 112 estimates a "surrounding risk" as the assumed risk. In other words, the risk estimation unit 112 estimates that there is a risk in the vicinity of the risk detection system 200.

[0045] Returning to FIG. 4 , the description of step S140 continues. If the number of sensor detections is greater than the number of wireless detections, the process proceeds to step S143. In step S143, the risk estimation unit 112 estimates the estimated risk due to the presence of an object within the sensor observation range. In step S144, the risk estimation unit 112 estimates the estimated risk due to the presence of an object outside the sensor observation range.

[0046] Figure 7 shows an example of a situation where the number of sensor detections is greater than the number of wireless detections. There are five pedestrians within the sensor observation range. Therefore, the number of sensor detections is "5." There are four pedestrians carrying wireless devices within the wireless communication range. Therefore, the number of wireless detections is "4." Therefore, the number of sensor detections "5" is greater than the number of wireless detections "4."

[0047] If the number of sensor detections is greater than the number of wireless detections, the following possibilities are possible: 1. There is an object without a wireless terminal within the sensor observation range. A certain number of objects will be detected by the sensor device 201. However, it is not possible to deny the existence of an object hidden behind other objects in a densely populated area, for example. 2. There is an object with a wireless terminal outside the sensor observation range. 3. There is an object without a wireless terminal outside the sensor observation range. The possible scenarios are almost the same as when the number of sensor detections and the number of wireless detections match. However, since there is definitely an object without a wireless terminal, it is necessary to be even more vigilant in the surrounding area.

[0048] The expected risks estimated in steps S143 and S144 will be described with reference to FIG. 8 . Within the sensor observation range (step S143), there is a possibility that an object without a wireless terminal will not be detected. Therefore, the risk estimation unit 112 estimates "contact accident" and "crowding risk" as expected risks. In other words, the risk estimation unit 112 estimates that there is a risk of a contact accident and congestion. Outside the sensor observation range (step S144), there is a possibility that an object that cannot be detected by the sensor device 201 (a non-line-of-sight object) will not be detected. Therefore, the risk estimation unit 112 estimates "surrounding risk" as the expected risk. In other words, the risk estimation unit 112 estimates that there is a risk in the vicinity of the risk detection system 200.

[0049] Returning to FIG. 4 , the description of step S140 continues. If the number of sensor detections is less than the number of wireless detections, the process proceeds to step S145. In step S145, the risk estimation unit 112 estimates the estimated risk due to the presence of an object within the sensor observation range. In step S146, the risk estimation unit 112 estimates the estimated risk due to the presence of an object outside the sensor observation range.

[0050] 9 shows an example of a situation in which the number of sensor detections is less than the number of wireless detections. Four pedestrians are present within the sensor observation range. However, one pedestrian is hidden behind the other pedestrians and is not detected by the sensor device 201. Therefore, the number of sensor detections is "3." Five pedestrians carrying wireless terminals are present within the wireless communication range. Therefore, the number of wireless detections is "5." Therefore, the number of sensor detections "3" is less than the number of wireless detections "5."

[0051] If the number of sensor detections is less than the number of wireless detections, the following possibilities are possible: 1. An object carrying a wireless terminal exists outside the sensor observation range. As this object approaches, there is a risk that the object may jump out of sight. 2. An object carrying a wireless terminal exists in a densely populated area within the sensor observation range, hidden behind other objects. In this case, there is a risk of congestion in the densely populated area and a risk that the object may jump out of the densely populated area. 3. An object without a wireless terminal exists outside the sensor observation range. The major difference from when the number of sensor detections and the number of wireless detections match is that there are always objects that are not detected by the sensor device 201.

[0052] The expected risks estimated in steps S145 and S146 will be described with reference to FIG. 10 . Within the sensor observation range (step S145), there is a possibility that an object hidden behind another object will not be detected. Therefore, the risk estimation unit 112 estimates "contact accident" and "crowding risk" as expected risks. In other words, the risk estimation unit 112 estimates that there is a risk of a contact accident and congestion. Outside the sensor observation range (step S146), there is a possibility that an object outside the line of sight will not be detected. Therefore, the risk estimation unit 112 estimates "surrounding risk" as the expected risk. In other words, the risk estimation unit 112 estimates that there is a risk in the vicinity of the risk detection system 200.

[0053] Returning to FIG. 3, step S150 will be described.

[0054] In step S150, the risk notification unit 113 notifies the estimated assumed risk.

[0055] For example, the risk detection system 200 is mounted on an automobile. The risk notification unit 113 uses a car navigation system to notify the driver of the anticipated risk by voice. The driver operates the accelerator, brake, steering wheel, etc. in accordance with the anticipated risk. For example, the risk detection system 200 is mounted on a mobile object that is equipped with an automatic driving device and travels autonomously. The risk notification unit 113 inputs data indicating the anticipated risk into the automatic driving device. The automatic driving device controls the speed, driving direction, etc. in accordance with the anticipated risk.

[0056] The risk notification unit 113 may notify the magnitude of the assumed risk along with the assumed risk. For example, the risk notification unit 113 determines the magnitude of the assumed risk based on the difference between the number of sensor detections and the number of wireless detections (detection difference). The magnitude of the assumed risk increases as the detection difference increases.

[0057] The features of the risk estimation method will be described with reference to FIG. 11 . The risk estimation device 100 performs detection in the same manner as conventional methods, using a sensor device 201. The sensor device 201 alone cannot detect objects outside of the line of sight. The risk estimation device 100 also performs communication using a wireless device 202. The wireless device 202 is not affected by obstacles within its detection range, and is therefore capable of detecting objects outside of the line of sight. The risk estimation device 100 estimates risks outside of the line of sight based on the difference between the detection situation by the sensor device 201 and the detection situation by the wireless device 202. By utilizing the difference in detection situations as well as sensor fusion, it is possible to estimate the risk of objects that cannot be detected by the sensor device 201.

[0058] ***Effects of First Embodiment*** Sensor devices such as cameras or LiDAR alone cannot recognize objects that exist outside the line of sight due to obstacles such as the shadows of objects, behind fences, and behind intersections. This makes it difficult to observe risks that are outside the line of sight. Portable terminal devices such as smartphones are capable of wireless communication such as BLE beacons or Wi-Fi. However, it is difficult to obtain detailed information about a detection target using only the portable terminal device. The first embodiment has the following objectives: The number and positions of objects that exist outside the line of sight of the sensor device 201 are ascertained using the wireless device 202. This expands the detection range, improves detection accuracy, and obtains information for risk avoidance.

[0059] The first embodiment has the following features. The risk estimation device 100 uses the wireless device 202 to measure the number and positions of objects present in the vicinity. As a result, the risk estimation device 100 detects objects that are outside the line of sight of the sensor device 201. The risk estimation device 100 fuses the wireless device 202 with the sensor device 201. The risk estimation device 100 then identifies objects that cannot be detected by the sensor device 201 and estimates the risk of congestion in the vicinity or the driver's own risk of contact. By having these features, the first embodiment has the following effects. It is possible to identify objects that are outside the line of sight and estimate various risks outside the line of sight based on the movement and number of those objects. As a result, it is possible to prevent accidents from occurring.

[0060] Second Embodiment A second embodiment of estimating a potential risk associated with the approach of an out-of-sight object will be described with reference to Figs. 12 to 14, focusing mainly on the differences from the first embodiment.

[0061] ***Description of Configuration*** The configuration of the risk estimation device 100 is the same as that in the first embodiment.

[0062] ***Description of Operation*** The risk estimation method will be described with reference to Fig. 12. In step S210, the object detection unit 111 acquires observation data from the sensor device 201 and detects a sensor-detected object using the observation data. Step S210 corresponds to step S110 in the first embodiment.

[0063] At this time, the object detection unit 111 estimates the position information of each sensor-detected object. For example, the observation data is image data created by a camera capturing images of the surroundings. The object detection unit 111 processes the image data to calculate the relative position of each sensor-detected object with respect to the camera. For example, the observation data is three-dimensional point cloud data created by a LiDAR measuring the surroundings. The object detection unit 111 processes the three-dimensional point cloud data to calculate the relative position of each sensor-detected object with respect to the LiDAR.

[0064] In step S220, the object detection unit 111 acquires communication partner data from the wireless device 202 and detects a wirelessly detected object using the communication partner data. Step S220 corresponds to step S120 in the first embodiment.

[0065] At this time, the object detection unit 111 estimates the position information of each wirelessly detected object. For example, if the communication partner has a positioning function, the communication partner data indicates the position of the communication partner. An example of the positioning function is satellite positioning. An example of a satellite positioning system is the Global Positioning System (GPS). For example, the wireless device 202 calculates the direction and distance of the communication partner by DOA estimation using an array antenna. The communication partner data indicates the direction and distance of each communication partner. DOA is an abbreviation for Direction of Arrival. For example, the communication partner data indicates the reception strength of radio waves from the communication partner for each communication partner. The reception strength varies depending on the distance of the communication partner and the presence or absence of an obstruction. The object detection unit 111 refers to the communication partner data and estimates the distance of the communication partner for each communication partner based on the reception strength.

[0066] In step S230, the risk estimation unit 112 determines whether or not there is an out-of-line-of-sight object based on the position information of each sensor-detected object and the position information of each wireless-detected object.

[0067] A non-line-of-sight object is a wirelessly detected object that is not a sensor-detected object. The risk estimation unit 112 finds a wirelessly detected object whose position information does not match any of the sensor-detected objects as a non-line-of-sight object.

[0068] If a non-line-of-sight object is present, the process proceeds to step S240. In step S240, the risk estimation unit 112 measures movement information of the non-line-of-sight object based on the position information of the non-line-of-sight object at each time, and estimates an assumed risk due to the presence of the non-line-of-sight object based on the movement information of the non-line-of-sight object.

[0069] Specifically, the risk estimation unit 112 measures the distance of the out-of-sight object as movement information of the out-of-sight object, and estimates the expected risk according to changes in the distance of the out-of-sight object.

[0070] Details of step S240 will be described with reference to Fig. 13. In step S241, the risk estimation unit 112 starts tracking an out-of-line-of-sight object.

[0071] Tracking of out-of-sight objects is performed, for example, as follows: The object detection unit 111 acquires communication partner data at each time and estimates position information of the out-of-sight object. The risk estimation unit 112 records the position information of the out-of-sight object at each time.

[0072] In step S242, the risk estimation unit 112 measures the distance of the non-line-of-sight object from the wireless device 202 based on the position information of the non-line-of-sight object.

[0073] In step S243, the risk estimation unit 112 compares the distance of the out-of-line-of-sight object with a proximity threshold, which is a predetermined value.

[0074] If the distance of the non-line-of-sight object is equal to or less than the proximity threshold, the process proceeds to step S244. If the distance of the non-line-of-sight object is greater than the proximity threshold, the process proceeds to step S245.

[0075] In step S244, the risk estimation unit 112 estimates the expected risk to be a “contact accident.” After step S244, the process of step S240 ends.

[0076] In step S245, the risk estimation unit 112 compares the distance of the out-of-line-of-sight object with a far-away threshold, which is a predetermined value.

[0077] If the distance of the non-line-of-sight object is less than the far threshold, the process proceeds to step S242. If the distance of the non-line-of-sight object is equal to or greater than the far threshold, the process proceeds to step S246.

[0078] In step S246, the risk estimation unit 112 estimates the assumed risk as a "peripheral risk." After step S246, the process of step S240 ends.

[0079] The description will continue with reference back to Figure 12. After step S240, the process proceeds to step S260.

[0080] If it is determined in step S230 that no out-of-sight object exists, the process proceeds to step S250. In step S250, the risk estimation unit 112 estimates a "peripheral risk" as the assumed risk. After step S250, the process proceeds to step S260.

[0081] In step S260, the risk notification unit 113 notifies the assumed risk. Step S260 is the same as step S150 in the first embodiment.

[0082] ***Effects of Embodiment 2*** The risk estimation device 100 uses the wireless device 202 to identify objects that cannot be detected by the sensor device 201, estimates the movement information (distance and direction) of the object, and estimates the risk associated with the approaching object. Embodiment 2 makes it possible to sense the approach of an object and estimate the risk associated with the approaching object.

[0083] 14 shows an example of a situation in which an object (pedestrian, automobile) approaches the risk detection system 200. According to the second embodiment, it is possible to estimate, for example, the risk of a collision accident caused by an object jumping out from behind a fence and the risk of a collision accident involving an object approaching from behind.

[0084] Third Embodiment A third embodiment of estimating a potential risk associated with a change in the traveling direction of an out-of-sight object will be described with reference to Figs. 15 to 17, focusing mainly on the differences from the first and second embodiments.

[0085] ***Description of Configuration*** The configuration of the risk estimation device 100 is the same as that in the first embodiment.

[0086] ***Description of Operation*** The risk estimation method will be described with reference to Figure 15. In step S310, the object detection unit 111 acquires observation data from the sensor device 201 and detects sensor-detected objects using the observation data. At this time, the object detection unit 111 estimates position information of each sensor-detected object. Step S310 is the same as step S210 in the second embodiment.

[0087] In step S320, the object detection unit 111 acquires communication partner data from the wireless device 202 and detects wirelessly detected objects using the communication partner data. At this time, the object detection unit 111 estimates position information of each wirelessly detected object. Step S320 is the same as step S220 in the second embodiment.

[0088] In step S330, the risk estimation unit 112 determines whether or not there is a non-line-of-sight object based on the position information of each sensor-detected object and the position information of each wireless-detected object. Step S330 is the same as step S230 in the second embodiment.

[0089] If a non-line-of-sight object is present, the process proceeds to step S340. In step S340, the risk estimation unit 112 measures movement information of the non-line-of-sight object based on the position information of the non-line-of-sight object at each time, and estimates an assumed risk due to the presence of the non-line-of-sight object based on the movement information of the non-line-of-sight object.

[0090] Specifically, the risk estimation unit 112 measures the movement path of the out-of-sight object as movement information of the out-of-sight object, and estimates the expected risk according to changes in the movement path of the out-of-sight object.

[0091] Step S340 will be described in detail with reference to Fig. 16. In step S341, the risk estimation unit 112 starts tracking of the out-of-line-of-sight object. Step S341 is the same as step S241 in the second embodiment.

[0092] In step S342, the risk estimation unit 112 measures the movement path of the out-of-sight object based on the position information of the out-of-sight object at each time.

[0093] In step S343, the risk estimation unit 112 calculates the amount of change in the travel route. Then, the risk estimation unit 112 compares the amount of change in the travel route with a threshold value. The threshold value is a predetermined value.

[0094] If the amount of change in the movement path is less than the threshold, the process proceeds to step S342. If the amount of change in the movement path is equal to or greater than the threshold, the process proceeds to step S344.

[0095] In step S344, the risk estimation unit 112 determines whether the traveling direction of the out-of-sight object has changed based on the position information of the out-of-sight object at each time.

[0096] Specifically, the risk estimation unit 112 determines whether the traveling direction of the out-of-sight object has changed from the direction toward the object having the risk detection system 200.

[0097] If the traveling direction of the out-of-line-of-sight object has changed, the process proceeds to step S342. If the traveling direction of the out-of-line-of-sight object has not changed, the process proceeds to step S345.

[0098] In step S345, the risk estimation unit 112 estimates a "peripheral risk" as an assumed risk.

[0099] The description will continue with reference back to Figure 15. After step S340, the process proceeds to step S360.

[0100] If it is determined in step S330 that no out-of-sight object exists, the process proceeds to step S350. In step S350, the risk estimation unit 112 estimates a "peripheral risk" as the assumed risk. After step S350, the process proceeds to step S360.

[0101] In step S360, the risk notification unit 113 notifies the assumed risk. Step S360 is the same as step S150 in the first embodiment.

[0102] ***Effects of Embodiment 3*** The risk estimation device 100 tracks out-of-line-of-sight objects detected by the wireless device 202. Each wireless terminal is assigned an ID, allowing for individual identification. Furthermore, since the approximate location can be determined, the object's movement path can be ascertained. The risk estimation device 100 estimates risk when the object's movement path changes suddenly. FIG. 17 shows an example of the movement path of an object (pedestrian). The dashed circle represents an obstacle. The dotted arrow indicates the movement path of the pedestrian avoiding the obstacle and proceeding toward the automobile 210. The solid arrow indicates the movement path of the pedestrian proceeding in a direction different from the automobile 210. An object may deviate significantly from its direction of travel due to the presence of an obstacle or other reasons. If the object subsequently returns to its original path, it is possible that the object avoided the obstacle. In this case, the risk estimation device 100 estimates a "surrounding risk." Furthermore, if the object continues moving forward without returning to its original path, it is considered that the object's direction of travel has changed. In this case, the risk estimation device 100 determines that there is no risk. According to the third embodiment, it is possible to track out-of-sight objects and estimate risks.

[0103] *** Supplementary Information about the Embodiment *** The hardware configuration of the risk estimation device 100 will be described with reference to Figure 18. The risk estimation device 100 includes a processing circuit 109. The processing circuit 109 is hardware that realizes an object detection unit 111, a risk estimation unit 112, and a risk notification unit 113. The processing circuit 109 may be dedicated hardware, or may be a processor 101 that executes a program stored in memory 102.

[0104] When the processing circuit 109 is dedicated hardware, the processing circuit 109 may be, for example, a single circuit, a multiple circuit, a programmed processor, a parallel programmed processor, an ASIC, an FPGA, or a combination thereof. ASIC is an abbreviation for Application Specific Integrated Circuit. FPGA is an abbreviation for Field Programmable Gate Array.

[0105] The risk estimation device 100 may include multiple processing circuits replacing the processing circuit 109 .

[0106] In the processing circuit 109, some functions may be realized by dedicated hardware, and the remaining functions may be realized by software or firmware.

[0107] Thus, the functions of the risk estimation device 100 can be realized by hardware, software, firmware, or a combination of these.

[0108] The "part" of each element of the risk estimation device 100 may be read as a "process," a "step," a "circuit," or a "circuitry."

[0109] The first to third embodiments may be implemented in combination with one another. The risk estimation unit 112 estimates an assumed risk using the method in each embodiment. The risk notification unit 113 notifies the assumed risk estimated using the method in each embodiment.

[0110] Each embodiment is an example of a preferred embodiment and is not intended to limit the technical scope of the present disclosure. Each embodiment may be implemented in part or in combination with other embodiments. Procedures described using flowcharts, etc. may be modified as appropriate.

[0111] 100 Risk estimation device, 101 Processor, 102 Memory, 103 Auxiliary storage device, 104 Interface, 109 Processing circuit, 111 Object detection unit, 112 Risk estimation unit, 113 Risk notification unit, 120 Memory unit, 200 Risk detection system, 201 Sensor device, 202 Wireless device, 210 Automobile.

Claims

1. A risk estimation device comprising: an object detection unit that acquires observation data from a sensor device, uses the observation data to detect each object present within the observation range of the sensor device as a sensor-detected object, acquires communication partner data from a wireless terminal, and uses the communication partner data to detect each object present within the communication range of the wireless terminal as a wireless-detected object; and a risk estimation unit that estimates an assumed risk assumed due to the presence of surrounding objects based on information about the sensor-detected objects and information about the wireless-detected objects.

2. The risk estimation device according to claim 1, wherein the risk estimation device estimates the assumed risk due to the presence of an object outside the observation range.

3. A risk estimation device as described in claim 1 or claim 2, wherein the risk estimation unit determines the magnitude relationship between the sensor detection number, which is the number of objects detected by the sensor, and the wireless detection number, which is the number of objects detected by the wireless, and estimates the expected risk based on the magnitude relationship.

4. A risk estimation device as described in any one of claims 1 to 3, wherein the object detection unit estimates position information of each of the sensor-detected objects and position information of each of the wirelessly detected objects, and the risk estimation unit determines that an object among the wirelessly detected objects that is not a sensor-detected object is a non-line-of-sight object based on the position information of each of the sensor-detected objects and the position information of each of the wirelessly detected objects, measures movement information of the non-line-of-sight objects based on the position information of the non-line-of-sight objects at each time, and estimates the expected risk due to the presence of the non-line-of-sight object based on the movement information of the non-line-of-sight objects.

5. A risk estimation device as described in claim 4, wherein the risk estimation unit measures the distance of the out-of-sight object as the movement information of the out-of-sight object, and estimates the expected risk according to changes in the distance of the out-of-sight object.

6. A risk estimation device as described in claim 4 or claim 5, wherein the risk estimation unit measures the movement path of the out-of-sight object as the movement information of the out-of-sight object, and estimates the expected risk according to changes in the movement path.

7. A risk estimation method that acquires observation data from a sensor device, uses the observation data to detect each object present within the observation range of the sensor device as a sensor-detected object, acquires communication partner data from a wireless terminal, uses the communication partner data to detect each object present within the communication range of the wireless terminal as a wireless-detected object, and estimates an expected risk due to the presence of surrounding objects based on the information on the sensor-detected objects and the information on the wireless-detected objects.

8. A risk estimation program for causing a computer to execute the following steps: an object detection process for acquiring observation data from a sensor device, using the observation data to detect each object present within the observation range of the sensor device as a sensor-detected object, acquiring communication partner data from a wireless terminal, and using the communication partner data to detect each object present within the communication range of the wireless terminal as a wireless-detected object; and a risk estimation process for estimating potential risks associated with the presence of surrounding objects based on information about the sensor-detected objects and information about the wireless-detected objects.

Images