Driving assistance system

Abstract

The present application provides a kind of driving assistance system.The driving assistance system has: multiple first processors, which are respectively mounted on multiple vehicles, and can extract feature quantity related to target object located in the surroundings of the vehicle based on target object information obtained from the surroundings sensor of the vehicle, and can generate map containing the position, moving direction and moving speed of the target object relative to the vehicle based on the feature quantity;And second processor is configured as follows: the feature quantity or the map and the position of the vehicle are obtained from the multiple first processors, generate wide-area map containing information related to the position, moving direction and speed of the multiple vehicles and the target object in the surroundings of the multiple vehicles based on the obtained information, estimate the collision risk of vehicle and target object based on the wide-area map, and send collision risk information to the vehicle estimated as the collision risk is high.

Description

Technical Field

[0001] This invention relates to a driver assistance system that assists the driving operation of each vehicle based on information acquired by sensors mounted on multiple vehicles. Background Technology

[0002] A driver assistance system (see, for example, Patent Document 1 below) is proposed that assists driving operations of each vehicle based on information acquired by sensors mounted on multiple vehicles. This system (hereinafter referred to as the "Prior System") includes a processor, a camera, and an image display device mounted on each vehicle. In addition to images acquired by the vehicle's own camera, the processor generates a top-down image based on images acquired by cameras of other vehicles located around the vehicle and displays it on the image display device. Thus, when it is possible to capture areas (blind spots) that cannot be captured by the vehicle's own camera alone using cameras of other vehicles, the top-down image of the blind spot can be supplemented based on images acquired from the cameras of those other vehicles.

[0003] Patent Document 1: Japanese Patent Application Publication No. 2020-65141 Summary of the Invention

[0004] In the aforementioned existing system, the vehicle's processor acquires images captured by cameras on other vehicles via wireless communication lines. Typically, the data size of images acquired by in-vehicle cameras is relatively large. Therefore, it takes a relatively long time for the vehicle's processor to acquire images from other vehicles. In this case, a display delay in the overhead view may occur. That is, it may reduce the usefulness of the system as a driver assistance system.

[0005] One of the objectives of this invention is to provide a driving assistance system that can provide driving assistance to the driver of a specific vehicle among multiple vehicles at appropriate times.

[0006] To achieve the above objectives, the driving assistance system (1) of the present invention comprises:

[0007] Multiple position sensors (24n) are mounted on multiple vehicles (Vn) and acquire position information representing the position of each vehicle.

[0008] Multiple ambient sensors (21n, 22n) are mounted on the multiple vehicles respectively, and acquire target information related to target objects located around each vehicle.

[0009] Multiple first processors (10n), each mounted on one of the multiple vehicles, are capable of extracting feature quantities (FA) related to targets located around the vehicle based on target information obtained from the surrounding sensors of the vehicle, and of generating a mapping map (M1n) containing the position, direction of movement, and speed of the target relative to the vehicle based on the feature quantities; and

[0010] A second processor (SV) is connected to the plurality of first processors via a wireless communication line, and the second processor is configured to: acquire the feature quantity or the mapping map from the plurality of first processors and acquire the position of the vehicle from the plurality of first processors; generate a wide-area mapping map (M2) based on the acquired information, which includes information related to the position, direction of movement, and speed of the plurality of vehicles and surrounding objects; estimate the collision risk between the vehicle and the object based on the wide-area mapping map; and send collision risk information indicating the high collision risk to the vehicle estimated to have a high collision risk.

[0011] The first processor is configured to: upon obtaining the collision risk information from the second processor, control the vehicle's notification device (30) to issue a prescribed warning to the driver.

[0012] In the driver assistance system of this invention, target information acquired by the surrounding sensors of each vehicle is processed by the vehicle's first processor to extract various feature quantities. These feature quantities are then processed by the first processor to generate a mapping map. This feature quantity or mapping map, along with the vehicle's position (current position), is then sent to a second processor. Typically, the data size of this information is smaller than that of information output from surrounding sensors (e.g., the data size of radio wave reflection point data from millimeter-wave radar, or image data from a camera). Therefore, the time for transmitting this information from the first processor of each vehicle to the second processor is short. The second processor generates a wide-area mapping map and calculates the collision risk of each vehicle with targets (moving objects (other vehicles and pedestrians), stationary objects, etc.) based on this map. The second processor then provides collision risk information (trigger information for issuing an alert to the driver in the vehicle) to the driver of the specific vehicle with a high collision risk with the target object. The data size of the collision risk information is relatively small. Therefore, the time for transmitting the collision risk information from the second processor to the first processor is short. Therefore, the data size transmitted and received between the first and second processors is relatively small, thus enabling collision risk information to be provided at the appropriate time.

[0013] In a driving assistance system according to one aspect of the present invention,

[0014] The wireless communication line is a line set up for transmitting and receiving radio waves from mobile phones.

[0015] The second processor is located at a base station that relays the wireless communication line and the wired communication line.

[0016] Typically, mobile phones connect to base stations via wireless communication lines. Base stations are facilities that relay these wireless communication lines to wired communication lines (the Internet). Data sent from a mobile phone is sequentially transmitted to the target destination computer via the wireless communication line, the base station, and multiple server computers connected to the wired communication line. Conversely, data sent from a source computer is transmitted to a specific base station via multiple server computers connected to the wired communication line, and from that base station is transmitted to a specific landline via the wireless communication line. Thus, when sending and receiving data between a mobile phone and a target computer, the transmission and reception process takes a relatively long time due to the multiple server computers involved. In the driver assistance system described in this method, the second processor is located at the mobile phone's base station, allowing the first and second processors to communicate without traversing multiple server computers. Therefore, the time (latency) required for data transmission and reception is relatively small.

[0017] In another aspect of the present invention, a driving assistance system is involved.

[0018] The second processor is only configured in specific base stations among the multiple base stations that cover a pre-selected area as a high-traffic area.

[0019] This reduces the risk of vehicle-object collisions in areas with relatively high traffic volume. Furthermore, the cost of constructing this system is reduced compared to deploying a second processor at all base stations.

[0020] In another aspect of the present invention, a driving assistance system is involved.

[0021] If the second processor determines that the collision risk is high when it detects that the time required for a specific vehicle among the plurality of vehicles to collide with the target object, as estimated based on the wide-area mapping, i.e., the collision margin time, is below a threshold, the second processor determines that the collision risk is high.

[0022] Therefore, it is possible to relatively easily estimate the collision risk of each vehicle with other objects. Attached Figure Description

[0023] Figure 1 This is a block diagram of a driving assistance system according to one embodiment of the present invention.

[0024] Figure 2 It is an example of mapping graph M1n and mapping graph M2.

[0025] Figure 3 This is a flowchart of the first procedure used to implement the alarm function.

[0026] Figure 4 This is a flowchart of the second procedure used to implement the alarm function.

[0027] Figure 5 This is a flowchart of the third program used to implement the alarm function. Detailed Implementation

[0028] (roughly)

[0029] like Figure 1 As shown, an embodiment of the present invention relates to a driver assistance system (DAS) comprising an on-board unit VDn mounted on multiple vehicles Vn (n=1, 2, 3, ...) equipped with autonomous driving functions, and a driver assistance server SV connected to each on-board unit VDn via a wireless communication line. The driver assistance server SV of the DAS has the following function (alarm function): when the autonomous driving function of vehicle Vn is disabled and the driver is performing manual driving operations, it provides information indicating a high collision risk to the driver of vehicle Vn with a high risk of collision with a target object OBm (m=1, 2, 3, ...).

[0030] (Specific structure)

[0031] Next, the structure of the Driver Assistance System (DAS) will be explained in detail. For example... Figure 2 As shown, the vehicle-mounted device VDn includes an ECU 10n, a vehicle-mounted sensor 20n, and a notification device 30n.

[0032] ECU10n (first processor) includes a microcomputer (SoC) with a CPU, ROM, RAM, timers, etc. Furthermore, ECU10n includes a communication device. The communication device is connected to the driver assistance server SV (described later) via a wireless communication line (mobile phone line).

[0033] The vehicle-mounted sensors 20n include millimeter-wave radar 21n, camera 22n, speed sensor 23n, and navigation system 24n.

[0034] The millimeter-wave radar 21n includes a transceiver unit and a signal processing unit (illustrations omitted). The transceiver unit radiates millimeter-wave radio waves (hereinafter referred to as "millimeter waves") to the surrounding area of ​​the vehicle (vehicle Vn) and receives millimeter waves (reflected waves) reflected by three-dimensional objects (vehicles, pedestrians, guardrails, poles, etc.) within the radiation range. The signal processing unit obtains the position information (distance and direction from the transceiver unit) of each reflection point based on the time from the transceiver unit's transmission of the millimeter wave to the reception of the reflected wave, the phase difference between the transmitted millimeter wave and the received reflected wave, and the attenuation level of the reflected wave. The signal processing unit provides the reflection point information to the vehicle's ECU 10n.

[0035] Camera 22n has multiple camera units. Each camera unit incorporates imaging elements such as lenses, charge-coupled devices (CCDs) or CMOS image sensors (CIS). The camera units are positioned at the front and rear of the vehicle, facing forward and backward respectively. Each camera unit captures image data of the front and rear areas of the vehicle at a specified frame rate. Each camera unit provides the image data to the vehicle's ECU 10n.

[0036] Speed ​​sensor 23n detects the rotational speed (wheel speed) of each wheel of the vehicle and calculates the vehicle's speed spn (measured value) based on the wheel speed. Speed ​​sensor 23n provides this calculation result to the vehicle's ECU 10n.

[0037] The navigation system 24n receives GPS signals from multiple GPS satellites and calculates the vehicle's current position Pn (latitude and longitude) based on these signals. The navigation system 24n provides this calculation result to the vehicle's ECU 10n. The ECU 10n can calculate the vehicle's direction of movement (azimuth) based on the history of the current position Pn obtained from the navigation system 24n. Furthermore, the ECU 10n can calculate the vehicle's direction of movement (turning radius) based on the outputs of sensors that acquire the vehicle's steering angle and sensors that acquire the vehicle's yaw rate. The navigation system 24n also includes an image display device for displaying a map and an audio device for playing voice guidance to the destination.

[0038] The notification device 30n includes an image display device and an audio device. The image display device receives an image display command from the vehicle's ECU 10n and displays an image accordingly (e.g., an image indicating a high risk of collision between the vehicle and another object OBm). The audio device receives a voice playback command from the vehicle's ECU 10n and plays a voice message accordingly (e.g., a voice message indicating a high risk of collision between the vehicle and another object OBm). Additionally, the image display device and audio device of the navigation system 24n can also be used as the notification device 30n (for dual purposes).

[0039] The driver assistance server SV (second processor) includes a microcomputer equipped with a CPU, ROM, RAM, timers, etc. Furthermore, the driver assistance server SV includes a communication device. For example, the driver assistance server SV is installed at a mobile phone base station. The driver assistance server SV connects to the ECUs 10n of multiple vehicles Vn (each vehicle Vn located within the coverage area (communication area) of the base station) via the mobile phone's wireless communication line. That is, the driver assistance server SV can send and receive various types of information with each vehicle Vn. Additionally, the coverage areas of adjacent base stations often partially overlap. Therefore, sometimes the driver assistance server SVs within multiple adjacent mobile phone base stations receive radio waves transmitted from the ECUs 10n of the vehicles Vn. In this case, information acquired from the vehicles Vn is shared among the multiple driver assistance server SVs.

[0040] (Alarm function)

[0041] Each vehicle's ECU10n sequentially acquires various information (feature quantities) from its onboard sensors 20n. Each time information is acquired, it is applied to various neural networks (DNN1, backbone network) to extract various feature quantities. For example, ECU10n extracts various feature quantities used to obtain the distance Δdnm and direction between the vehicle and the target object OBm, and the speed (relative speed sprnm (vector)) of the target object OBm relative to the vehicle. Next, ECU10n applies these feature quantities to a predefined neural network (DNN2, head) to generate a top-down view (mapping map M1n) with the vehicle's center of gravity as the origin On, representing the types (moving objects (other vehicles and pedestrians) and stationary objects) and positions (distance and direction from the origin On) of each target object OBm existing around the vehicle. The top of the mapping map M1n is "North," and the bottom is "South." Furthermore, the right side of the mapping map M1n is "East," and the left side is "West." Furthermore, the mapping map M1n contains information related to the direction of movement and speed of each target OBm. ECU 10n displays the image representing this mapping map M1n on the image display device of the notification device 30n. ECU 10n also obtains the vehicle's current position Pn (latitude and longitude of the origin On) from the navigation system 24n. ECU 10n transmits the mapping map M1n and the current position Pn to the driver assistance server SV via a wireless communication line (mobile phone line). Additionally, information related to target OBm located in the blind spots of the millimeter-wave radar 21n and camera 22n is missing from this mapping map M1n. For example, in… Figure 2In the example shown, pedestrian Ped, located diagonally to the right front of vehicle V1 (n=1), cannot be identified by vehicle V1's sensors due to the presence of vehicle V2 (n=2). Therefore, pedestrian Ped is not included in map M11. In contrast, vehicle V2's sensors can identify pedestrian Ped, and therefore include pedestrian Ped in map M12.

[0042] The driver assistance server SV obtains the mapping map M1n and the current position Pn from the ECU10n of each vehicle Vn via a wireless communication line (mobile phone line). Based on the current position Pn obtained from each vehicle Vn (the latitude and longitude of the origin On of the mapping map M1n), the driver assistance server SV synthesizes (coordinate transforms) the mapping map M1n obtained from each vehicle Vn, generating a mapping map M2 (wide-area mapping map) that includes information related to the type, location (latitude and longitude), direction of movement, and speed of target objects OBx (x=1, 2, 3, ...) within the area covered by the base station where the driver assistance server SV is installed. Additionally, for example, sometimes target object OB1 (m=1) in mapping map M11 (n=1) is the same as target object OB1 (m=1) in mapping map M12 (n=2). The driver assistance server SV identifies overlapping target objects in mapping map M2 as a single target object. In this specification, "x" is used as the index for identifying each target object recognized by the server computer SSC. Furthermore, suppose that the communication status of a vehicle Vn located at the outer edge of the area covered by a base station equipped with a driver assistance server SV is not good. In this case, the driver assistance server SV can obtain (share) information (mapping map M1n) related to the vehicle Vn within the outer edge from the driver assistance server SVs of adjacent base stations whose coverage areas overlap with the outer edge. This improves the accuracy of the mapping map M2.

[0043] Next, the driver assistance server SV selects a vehicle Vn sequentially based on the information contained in the mapping map M2 (the position, type (vehicle Vn and other targets), direction of movement, and speed of each target OBx). Whenever a vehicle Vn is selected, the driver assistance server SV calculates the collision risk of that selected vehicle Vn with other target OBx. Specifically, the driver assistance server SV obtains the distance Δdnx and relative speed spnx of the selected vehicle Vn with each other target OBx, and calculates (estimates) the time required for vehicle Vn to collide (make contact) with a target OBx, i.e., the collision margin time TTcnx (=Δdnx / spnx). Then, if the collision margin time TTcnx is below the threshold TTCth, the driver assistance server SV sends collision risk information to vehicle Vn. Specifically, the driver assistance server SV sends information representing the positional relationship between vehicle Vn and target OBx (the position of target OBm in the mapping map M1n) as collision risk information. Furthermore, the driver assistance server SV can estimate the trajectories of each vehicle Vn and the target object OBx after the current time point based on the information contained in the mapping map M2. If these trajectories intersect, the server estimates the difference between the arrival time of vehicle Vn at the intersection and the arrival time of target object OBx. Then, if the difference at that time is below a threshold, the driver assistance server SV can send collision risk information to vehicle Vn.

[0044] When the ECU 10n of vehicle Vn receives collision risk information from the driver assistance server SV, it executes an alarm process. Specifically, the CPUn sends an instruction to the notification device 30n to change the display method of the image of the target object OBm in the mapping map M1n, so that the driver can identify the target object OBm with a high collision risk. Furthermore, the ECU 10n sends an instruction to the notification device 30n to generate a prescribed alarm sound.

[0045] Next, refer to Figures 3 to 5 To achieve the aforementioned function (alarm function) of the Driver Assistance System (DAS), the programs PR1 and PR2 executed by the CPU (hereinafter referred to as "CPUn") of each vehicle Vn and the CPU (hereinafter referred to as "CPUsc") of the Driver Assistance Server SV are described below. CPUn executes programs PR1 and PR3 at a predetermined cycle. Furthermore, CPUsc executes program PR2 at a predetermined cycle.

[0046] (Program PR1)

[0047] CPUn starts executing program PR1 from step 100 and proceeds to step 101.

[0048] In step 101, CPUn acquires various information INF from the vehicle-mounted sensor 20n. Then, CPUn proceeds to step 102.

[0049] In step 102, CPUn applies various information INF to various neural networks DNN1 (backbone networks) to extract various feature quantities FA. Then, CPUn proceeds to step 103.

[0050] In step 103, CPUn applies various feature quantities FA to the neural network DNN2 (head) to generate a mapping map M1n. Then, CPUn proceeds to step 104.

[0051] In step 104, CPUn obtains its current position Pn from navigation system 24n. Then, CPUn proceeds to step 105.

[0052] In step 105, CPUn sends the mapping map M1n and the current position Pn to the driver assistance server SV (CPUsc). Then, CPUn proceeds to step 106.

[0053] CPUn terminates the execution of program PR1 in step 106.

[0054] (Program PR2)

[0055] Program PR2 includes processes for generating mapping map M2 (steps 201 to 206) and processes for sending collision risk information to vehicle Vn when the collision risk between vehicle Vn and other target object OBm is high (steps 207 to 210).

[0056] CPUsc starts executing program PR2 from step 200 and proceeds to step 201.

[0057] In step 201, CPUsc obtains the maximum number of vehicles Vn (the number of vehicles Vn that can communicate at the current time) located within the coverage area of ​​the base station where CPUsc is set. Then, CPUsc proceeds to step 202.

[0058] In step 202, CPUsc initializes the index n, which serves as the identification number for the identified vehicle Vn, to "0". Then, CPUsc proceeds to step 203.

[0059] In step 203, CPUsc increments index n. Then, CPUsc proceeds to step 204.

[0060] In step 204, CPUsc retrieves and stores various information (mapping map M1n and current position Pn) from vehicle Vn (the vehicle corresponding to index n). Then, CPUsc proceeds to step 205.

[0061] In step 205, CPUsc determines whether index n matches the number of vehicles nmax. That is, CPUsc determines whether the mapping map M1n and the current position Pn have been obtained from all vehicles Vn that can communicate at the current time. If CPUsc determines that index n matches the number of vehicles nmax (205: Yes), it proceeds to step 206. Conversely, if CPUsc does not determine that index n matches the number of vehicles nmax (205: No), it returns to step 203.

[0062] In step 206, CPUsc generates map M2 based on all stored maps M1n and the current position Pn. Then, CPUsc proceeds to step 207.

[0063] In step 207, CPUsc initializes index n to "0". Then, CPUsc proceeds to step 208.

[0064] In step 208, CPUsc increments index n. Then, CPUsc proceeds to step 209.

[0065] In step 209, CPUsc obtains (calculates) the collision margin time TTCnx of vehicle Vn and other target objects OBx based on the mapping map M2 (position, direction of movement, and speed of vehicle Vn and other target objects OBx), and selects the smallest collision margin time TTCnx from the obtained collision margin times TTCnx. Then, CPUsc determines whether the selected collision margin time TTCnx is below the threshold TTCth. If CPUsc determines that the selected collision margin time TTCnx is below the threshold TTCth (209: Yes), it proceeds to step 210. On the other hand, if it does not determine that the selected collision margin time TTCnx is below the threshold TTCth (209: No), it proceeds to step 211.

[0066] In step 210, CPUsc sends collision risk information to vehicle Vn (the vehicle identified as having a high collision risk in step 209). Then, CPUsc proceeds to step 211.

[0067] In step 211, CPUsc determines whether index n is consistent with the number of vehicles nmax. That is, CPUsc determines whether the collision risk for all vehicles Vn that can communicate at the current time has been determined. If CPUsc determines that index n is consistent with the number of vehicles nmax (211: Yes), it proceeds to step 212. On the other hand, if CPUsc determines that index n is inconsistent with the number of vehicles nmax (211: No), it returns to step 208.

[0068] CPUsc terminates the execution of program PR2 in step 212.

[0069] (Program PR3)

[0070] CPUn starts executing program PR3 from step 300 and proceeds to step 301.

[0071] In step 301, CPUn determines whether it has received collision risk information from the driver assistance server SV. If CPUn determines that it has received collision risk information (301: Yes), it proceeds to step 302. On the other hand, if CPUn does not determine that it has received collision risk information (301: No), it proceeds to step 303.

[0072] In step 302, CPUn controls the notification device 30n to issue a prescribed alarm to the driver. Then, CPUn proceeds to step 303.

[0073] CPUn terminates the execution of program PR3 in step 303.

[0074] (Effect)

[0075] In the driver assistance system (DAS) of this embodiment, target information acquired by the onboard sensors 20n of each vehicle Vn is processed by the ECU 10n of that vehicle Vn to extract various feature quantities FA. Furthermore, the feature quantities FA are processed by the ECU 10n to generate a mapping map M1n. Then, the feature quantities FA or the mapping map M1n, along with the position of the vehicle Vn (current position Pn), are sent to the driver assistance server SV. Typically, the data size of this information (feature quantities FA, mapping map M1n, and current position Pn) is smaller than the data size of the information output from the onboard sensors 20n (e.g., the reflection point data of radio waves output from the millimeter-wave radar 21n, or the image data output from the camera 22n). Therefore, the time for transmitting this information from the ECU 10n of each vehicle Vn to the driver assistance server SV is short. The driver assistance server SV generates a mapping map M2 and calculates the collision risk between each vehicle Vn and the target object OBx based on this mapping map M2. Then, the driver assistance server SV sends collision risk information (trigger information for issuing an alert to the driver in the vehicle) to the ECU 10n of the specific vehicle Vn with a high risk of collision with the target object OBx. The data size of the collision risk information is relatively small. Therefore, the time for transmitting the collision risk information from the driver assistance server SV to the ECU 10n is short. As a result, the data size transmitted and received between the ECU 10n and the driver assistance server SV is relatively small, thus enabling collision risk information to be provided at the appropriate time.

[0076] (Variation Example 1)

[0077] In the driver assistance system (DAS) described above, the driver assistance server (SV) is located at the mobile phone base station, but it may also be located only at a specific base station. For example, the driver assistance server (SV) may be located only at a base station covering a pre-selected area (e.g., an intersection) with high traffic volume. Furthermore, the driver assistance server (SV) may be located in other facilities instead of at the mobile phone base station. For example, the functionality of the driver assistance server (SV) may be programmed as part of the functionality of another server computer (e.g., a web server) connected to the mobile phone base station via a wired communication line.

[0078] (Variation Example 2)

[0079] In the above embodiment, in each vehicle Vn, various information obtained from the on-board sensor 20n is applied to a neural network DNN1 to extract various feature quantities FA, and these feature quantities FA are further applied individually to a neural network DNN2. A mapping map M1n is generated. Then, instead of the driving assistance server SV, the driving assistance server SV obtains the mapping map M1n and the current position Pn from each vehicle Vn and generates a mapping map M2 based on this information.

[0080] Symbol Explanation

[0081] DAS - Driver Assistance System, 10n - ECU, 20n - Onboard Sensors, 30n - Notification Device, SV - Driver Assistance Server.

Claims

1. A driving assistance system, characterized in that, have: Multiple position sensors are mounted on multiple vehicles and acquire position information representing the position of each vehicle. Multiple ambient sensors are mounted on the multiple vehicles respectively, and each sensor acquires target information related to target objects located around each vehicle. A plurality of first processors, each mounted on one of the plurality of vehicles, are capable of extracting feature quantities related to targets located around the vehicle based on target information obtained from the surrounding sensors of the vehicle, and of generating a mapping map containing the position, direction of movement, and speed of the target relative to the vehicle based on the feature quantities; and A second processor, connected to the plurality of first processors via a wireless communication line, is configured to: acquire the feature quantity or the mapping map from the plurality of first processors and acquire the position of the vehicle from the plurality of first processors; generate a wide-area mapping map based on the acquired information, containing information related to the position, direction of movement, and speed of the plurality of vehicles and surrounding objects; estimate the collision risk between the vehicle and the objects based on the wide-area mapping map; and send collision risk information indicating a high collision risk to vehicles estimated to have a high collision risk. The first processor is configured to, upon obtaining the collision risk information from the second processor, control the vehicle's notification device to issue a prescribed warning to the driver.

2. The driving assistance system according to claim 1, characterized in that, The wireless communication line is a line set up for transmitting and receiving radio waves from mobile phones. The second processor is located at a base station that relays the wireless communication line and the wired communication line.

3. The driving assistance system according to claim 2, characterized in that, The second processor is only configured in specific base stations among the multiple base stations that cover a pre-selected area as a high-traffic area.

4. The driving assistance system according to claim 1, characterized in that, If the second processor determines that the collision risk is high when it detects that the time required for a specific vehicle among the plurality of vehicles to collide with the target object, as estimated based on the wide-area mapping, i.e., the collision margin time, is below a threshold, the second processor determines that the collision risk is high.

Images