Vehicle control method, vehicle control system, and non-transitory storage medium

Abstract

A vehicle control method includes: identifying a signal candidate position in which a signal light is likely to exist around a vehicle using a sensor mounted on the vehicle, the vehicle being provided with a light configured to change a light distribution state; and executing a first light control process of controlling the light in response to the identification of the signal candidate position, wherein the first light control process includes at least one of: making irradiation light irradiated toward the signal candidate position stronger than before the identification of the signal candidate position, and making the irradiation light irradiated toward the signal candidate position stronger than irradiation light irradiated toward a region other than the signal candidate position.

Description

Technical Field

[0001] This disclosure relates to vehicle control methods, vehicle control systems, and non-transient storage media. Background Technology

[0002] Japanese Patent Application Publication No. 2021-002275 discloses a signal recognition system. The system detects traffic lights around a vehicle based on images captured by an onboard camera, acquiring signal detection information. The signal detection information shows at least the appearance of each of the detected portions of the traffic light. On the other hand, light pattern information shows the relative positional relationships between the multiple light portions of the traffic light and the appearance of each light portion when illuminated. The signal recognition system identifies the illumination state of the traffic light by comparing the signal detection information with the light pattern information. Summary of the Invention

[0003] Consider the scenario where vehicles use onboard cameras to recognize traffic lights. The accuracy of traffic light recognition may decrease in low-light conditions such as at night or at dusk.

[0004] For example, at night, not only are there signal lights, but also various other types of light. Even if red, yellow, or green light is identified, it may not actually be a signal light. To determine whether the identified light is a signal light, it is necessary to identify whether a signal light exists at the location of that light. To identify the presence of a signal light, for example, consider identifying the casing of the signal light. However, at night, non-luminous objects are difficult to see, and the casing of the signal light is also difficult to see. Therefore, the probability of accurately identifying the presence of a signal light decreases. That is, at night, the accuracy of signal light identification decreases.

[0005] This disclosure provides a technique that can improve the recognition accuracy of traffic lights in dark environments.

[0006] The first viewpoint is a vehicle control method comprising: identifying, using sensors mounted on the vehicle, a potential signal candidate location for a traffic light around the vehicle, the vehicle having lights configured to change their light distribution state; and performing a first light control process to control the lights in response to the identification of the signal candidate location, wherein the first light control process includes at least one of: making the illumination light illuminating the signal candidate location stronger than before the identification of the signal candidate location; and making the illumination light illuminating the signal candidate location stronger than the illumination light illuminating an area outside the signal candidate location.

[0007] The second viewpoint is a vehicle control system comprising one or more processors configured to perform the following processes: identifying potential signal candidate locations for traffic lights around the vehicle using sensors mounted on the vehicle, the vehicle having lights configured to change their light distribution state; and a first light control process performed in response to the identification of the signal candidate locations, wherein the first light control process includes at least one of: making the illumination light illuminating the signal candidate locations stronger than before the identification of the signal candidate locations; and making the illumination light illuminating the signal candidate locations stronger than the illumination light illuminating areas outside the signal candidate locations.

[0008] The third viewpoint is a non-transient storage medium that stores commands executable by one or more processors to perform the following functions: using sensors mounted on a vehicle to identify potential signal candidate locations for traffic lights around the vehicle, the vehicle having lights configured to change their light distribution; and in response to the identification of the signal candidate locations, performing a first light control process to control the lights, wherein the first light control process includes at least one of: making the illumination light illuminating the signal candidate locations stronger than before the identification of the signal candidate locations; and making the illumination light illuminating the signal candidate locations stronger than the illumination light illuminating areas outside the signal candidate locations.

[0009] According to this disclosure, potential signal candidate locations for traffic lights are identified around the vehicle. Then, a first light control process is executed in response to the identification of the signal candidate locations. The first light control process includes at least one of the following: making the illumination light directed towards the signal candidate location stronger than before the identification of the signal candidate location; and making the illumination light directed towards the signal candidate location stronger than the illumination light directed away from the signal candidate location. Through this first light control process, signal candidate locations are easily visible even in dark environments, increasing the probability of traffic light identification in dark environments. That is, the accuracy of traffic light identification is improved. Attached Figure Description

[0010] The features, advantages, and technical and industrial significance of exemplary embodiments of the present invention will be described with reference to the accompanying drawings, in which the same reference numerals denote the same parts, wherein: Figure 1 This is a conceptual diagram illustrating the outline of a vehicle control system applied to the vehicle involved in the implementation method. Figure 2 It is a conceptual diagram used to illustrate technical issues. Figure 3 This is a conceptual diagram illustrating a lamp mounted on a vehicle involved in the implementation method. Figure 4This is a conceptual diagram illustrating a lamp mounted on a vehicle involved in the implementation method. Figure 5 This is a conceptual diagram used to illustrate the candidate signal positions involved in the implementation method. Figure 6 This is a conceptual diagram illustrating an example of the lamp control processing involved in the implementation method. Figure 7 This is a conceptual diagram illustrating another example of the lamp control processing involved in the implementation method. Figure 8 This is a conceptual diagram illustrating yet another example of the lamp control processing involved in the implementation method. Figure 9 This is a conceptual diagram illustrating yet another example of the lamp control processing involved in the implementation method. Figure 10 This is a conceptual diagram used to illustrate the screening process involved in the implementation method. Figure 11 This is a conceptual diagram illustrating an example of a signal light that is the object of illumination in an implementation method. Figure 12 This is a conceptual diagram illustrating another example of a signal light that serves as the object of illumination in the implementation method. Figure 13 This is a block diagram illustrating an example of the configuration of a vehicle control system according to an embodiment. Figure 14 This is a block diagram illustrating an example of driving environment information involved in an implementation method. Figure 15 This is a flowchart illustrating the processes associated with the lamp control process involved in the implementation. Figure 16 This is a conceptual diagram used to illustrate the switching of the irradiation target involved in the implementation method. Figure 17 This is a conceptual diagram used to illustrate the switching of the irradiation target involved in the implementation method. Figure 18 This is a flowchart illustrating the process associated with switching the irradiated object according to the embodiment. Detailed Implementation

[0011] Embodiments of this disclosure will be described with reference to the accompanying drawings.

[0012] 1. Overview of Vehicle Control Systems Figure 1This is a conceptual diagram illustrating the outline of a vehicle control system 10 applied to the vehicle 1 according to this embodiment. The vehicle control system 10 controls the vehicle 1. Typically, the vehicle control system 10 is mounted on the vehicle 1. Alternatively, at least a portion of the vehicle control system 10 may be included in a remote system external to the vehicle 1, remotely controlling the vehicle 1. That is, the vehicle control system 10 may also be distributed between the vehicle 1 and the remote system.

[0013] Vehicle 1 can be an autonomous vehicle. In this case, vehicle control system 10 performs autonomous driving control related to vehicle 1. For example, the level of autonomous driving is based on the premise that the driver does not need to be 100% focused on driving. Vehicle 1 can also be a driverless autonomous vehicle.

[0014] Consider the vehicle control system 10 recognizing the status of traffic lights around vehicle 1. For example, the vehicle control system 10 recognizes the traffic lights and their signal displays (red, green, yellow, arrows, etc.) in front of vehicle 1, and performs autonomous driving control based on the recognition results. As another example, the vehicle control system 10 may also notify the driver of the recognition results of the traffic lights or signal displays in front of vehicle 1 to assist the driver's driving. As yet another example, the vehicle control system 10 may also send the recognition results of the traffic lights or signal displays to an external management device (e.g., a map management device). In either case, it is desirable to recognize traffic lights with high accuracy.

[0015] The vehicle control system 10 uses a camera C mounted on the vehicle 1 to identify traffic lights around the vehicle 1. More specifically, the camera C acquires an image (IMG) showing the conditions around the vehicle 1. The vehicle control system 10 identifies traffic lights around the vehicle 1 based on the image (IMG) acquired by the camera C. For example, the vehicle control system 10 utilizes image recognition AI (Artificial Intelligence) obtained through machine learning to identify traffic lights in the image (IMG).

[0016] However, in dark environments such as nighttime and dusk, the recognition accuracy of traffic lights based on image IMG may decrease. Figure 2 This is a conceptual diagram used to illustrate the technical problem.

[0017] For example, at night, not only are there lights indicating traffic signals, but also various other types of light. Even if red, yellow, or green light is identified, it may not actually be a traffic signal. For example, there is the red light of a vehicle's taillights. As another example, there is the yellow light of a vehicle's turn signal. As yet another example, there is the green or yellow light of a taxi's top light. In situations where there are lights similar to those indicating traffic signals, judging traffic signals solely based on light recognition is unsatisfactory. For example, during autonomous driving control, it is possible that the vehicle control system 10 interprets red light as a red light signal and automatically slows down the vehicle 1, only to find that there is actually no red light signal. This means that the accuracy of autonomous driving control decreases, which is unsatisfactory.

[0018] To determine whether the detected light is the signal light, it's necessary to identify whether a signal light exists at the location of that light. To identify the presence of a signal light, consider, for example, identifying the signal light's casing. During the day, the signal light casing is relatively easy to identify. However, at night, non-luminous objects are difficult to see, and the signal light casing is also difficult to see. Therefore, the probability of accurately identifying (detecting) the presence of a signal light decreases. That is, at night, the accuracy of signal light identification decreases.

[0019] As another example, consider the arrow signal accompanying the main signal. There are also situations where the main signal displays "red (stop)" but the arrow signal displays "green (go)". However, since the light intensity of the arrow signal is less than that of the main signal, it is possible that the arrow signal display may not be accurately recognized before vehicle 1 approaches the traffic light to some extent. Automatically decelerating vehicle 1 at a point when only the red light is recognized and the arrow signal display is not recognized would lead to a decrease in the accuracy of autonomous driving control. To prevent such preemptive autonomous driving control, it is desirable to first identify whether the arrow signal itself exists. However, at night, the arrow signal's casing is also difficult to see. Therefore, the probability of accurately recognizing the presence of the arrow signal decreases. That is, at night, the recognition accuracy of traffic lights decreases.

[0020] The same technical problem exists regarding "obstruction signals" that cannot be seen before approaching to a certain extent.

[0021] Regarding the aforementioned technical issues, the problems become more pronounced the further away from the traffic light. On the other hand, to achieve sufficient margin for autonomous driving control, it is also necessary to accurately identify traffic lights from a distance of approximately 100 meters. Ideally, the recognition accuracy of traffic lights should be maximized even at a distance from the light itself.

[0022] Improving the recognition accuracy of traffic lights based on image IMGs is also useful in the case of remote assistance for vehicle 1. In the case of remote assistance for vehicle 1, an image IMG acquired by camera C is sent to a remote operator terminal on the remote operator side. The remote operator terminal has a display device that displays the received image IMG. The remote operator views the image IMG displayed on the display device, understands the situation around vehicle 1, and remotely assists the actions of vehicle 1. Examples of remote assistance performed by the remote operator include recognition assistance, judgment assistance, and remote driving. In any case, if the traffic lights in the image IMG are difficult to see, the accuracy of the remote assistance performed by the remote operator may decrease. Therefore, making the traffic lights in the image IMG easy to see is also preferred from the viewpoint of the accuracy of remote assistance.

[0023] Based on the above viewpoints, this embodiment proposes a technology that improves the recognition accuracy of traffic lights in dark environments such as nighttime and dusk. Specifically, the vehicle control system 10 involved in this embodiment utilizes lights 50 mounted on the vehicle 1 to improve the recognition accuracy of traffic lights.

[0024] Figure 3 This is a concept diagram illustrating a lamp 50 mounted on vehicle 1. The lamp 50 illuminates the exterior of vehicle 1. Furthermore, the light distribution of the lamp 50 is variable. That is, the lamp 50 is configured to freely change the illumination range and intensity.

[0025] More specifically, the maximum illumination range of lamp 50 is set wide, allowing for free adjustment of the illumination range within this maximum range. In other words, lamp 50 is configured to selectively illuminate only a portion of the maximum illumination range. For example, lamp 50 may contain multiple light sources. For instance, if the light source is an LED (Light Emitting Diode), lamp 50 may contain an LED array consisting of multiple LEDs. Multiple light sources can be independently controlled to turn on / off. By independently controlling the on / off of multiple light sources, the illumination range can be freely adjusted within the maximum illumination range. Furthermore, as... Figure 3 As shown, the irradiation range can be changed in both the horizontal (XY direction) and vertical (Z direction) directions.

[0026] In addition, the illumination intensity (output, power) of multiple light sources can be freely changed. Typically, to suppress power consumption, the default value of the illumination intensity of each light source is set to be less than the maximum illumination intensity. However, the illumination intensity of each light source can be temporarily increased or decreased as needed.

[0027] For example, lamp 50 is incorporated into the headlights of vehicle 1. In particular, lamp 50 here is a high beam (low beam is provided separately). The high beam can illuminate the position at the height of the traffic lights. In addition, the high beam can illuminate at least about 100m in front of vehicle 1. For example, the adaptive high beam system (AHS) installed in vehicle 1 in recent years can be used as lamp 50.

[0028] like Figure 4 As shown, vehicle 1 can also be equipped with a left lamp 50L and a right lamp 50R. It can also perform "focusing" to make the illumination light from the left lamp 50L overlap with the illumination light from the right lamp 50R. By focusing the light, the illumination intensity for a portion of the illumination range can be increased.

[0029] The vehicle control system 10 of this embodiment performs "light control processing" to control the lights 50 mounted on the vehicle 1. In particular, the vehicle control system 10 performs light control processing in a manner that improves the recognition accuracy of traffic lights in dark environments. Hereinafter, the light control processing performed by the vehicle control system 10 of this embodiment will be described in more detail.

[0030] 2. Light control processing First, refer to Figure 5 The "signal candidate position SC" used in the light control processing will be explained. A signal candidate position SC is a location around vehicle 1 where a traffic light may be present. Here, "position" encompasses the concepts of "area" and "space." Furthermore, "position" can be a location in actual space or a location within an image IMG. The size of a signal candidate position SC is finite. The size of a single signal candidate position SC within an image IMG is smaller than the size of the entire image IMG.

[0031] The vehicle control system 10 uses a sensor group 20 mounted on the vehicle 1 to identify (acquire) signal candidate locations SC around the vehicle 1.

[0032] For example, sensor group 20 includes an identification sensor 21 for recognizing the conditions around vehicle 1. Examples of identification sensors 21 include a camera C and a LiDAR (Laser Imaging Detection and Ranging) system. Vehicle control system 10 identifies a candidate signal location SC based on the recognition results obtained from the identification sensor 21. At this time, it is not necessary to confirm the presence of a traffic light at the candidate signal location SC. If light, objects, etc., associated with a traffic light are detected, that location can also be obtained as the candidate signal location SC.

[0033] As another example, sensor group 20 includes a position sensor 23 for obtaining the position of vehicle 1. An example of position sensor 23 is a GPS (Global Positioning System) sensor. There is a high probability that traffic lights are installed at intersections. Therefore, vehicle control system 10 can also obtain the positions of intersections surrounding vehicle 1 based on vehicle 1's position information and map information, and identify the positions of these intersections as candidate signal locations SC.

[0034] Various other methods can also be considered as a means of identifying candidate signal locations (SCs). Examples of various methods for identifying candidate signal locations (SCs) will be described later.

[0035] The vehicle control system 10 identifies (acquires) one or more signal candidate positions SC around the vehicle 1. There are also cases where multiple signal candidate positions SC are identified simultaneously. Then, the vehicle control system 10 performs lighting control processing in response to the identification of the signal candidate positions SC. Various examples of lighting control processing will be described below.

[0036] 2-1. Brightening treatment Figure 6 This is a conceptual diagram illustrating a "brightness enhancement process" as an example of lamp control processing. Brightness enhancement processing is the process of increasing the illumination light (output of lamp 50) directed towards a signal candidate position SC in response to a certain trigger. The trigger here is the recognition of the signal candidate position SC. That is, the vehicle control system 10, in response to the recognition of the signal candidate position SC, makes the illumination light directed towards the signal candidate position SC stronger than before the signal candidate position SC was recognized.

[0037] exist Figure 6 In the example shown, at time ts, the intensity of the illumination light illuminating the candidate signal position SC increases. Time ts is the identification time of the candidate signal position SC, or immediately after that identification time. On the other hand, the intensity of the illumination light illuminating other candidate signal positions remains unchanged. Typically, after time ts, the illumination light illuminating the candidate signal position SC becomes stronger than the illumination light illuminating other candidate signal positions SC.

[0038] For example, before time ts, light 50 (high beam) is off. Afterwards, in response to the identification of a signal candidate position SC, the vehicle control system 10 partially turns on light 50 so that it only illuminates the identified signal candidate position SC. That is, the vehicle control system 10 selectively illuminates signal candidate positions SC, excluding those other than the identified ones. This type of light control process can also be referred to as "selective lighting."

[0039] As another example, lamp 50 can also be illuminated before time ts. Typically, the illumination intensity before time ts is the default value. Afterward, in response to the identification of the signal candidate position SC, the vehicle control system 10 can also make the illumination intensity directed at the identified signal candidate position SC higher than the default value.

[0040] Furthermore, when objects such as traffic lights are illuminated with excessively strong light, they may become difficult to see due to factors such as reflected light and glare. Therefore, it is not necessarily necessary to increase the illumination intensity illuminating the candidate signal position SC to the maximum intensity. A suitable target illumination intensity can be pre-calculated through simulation, experimentation, etc. In this case, the vehicle control system 10 performs light control processing in a manner that ensures the illumination intensity illuminating the candidate signal position SC is a suitable target illumination intensity.

[0041] The effects obtained by the brightening treatment described above are as follows.

[0042] First, the illumination light shining on the candidate signal location SC becomes stronger, increasing the brightness (illuminance) of SC. Therefore, if a signal light actually exists at the candidate signal location SC, the brightness of the signal light's casing also increases, making it easier to see visually. Thus, even in dark environments, the presence of the signal light can be easily identified (detected) based on the image IMG. As a result, the probability of recognizing (detecting) a signal light based on the image IMG increases. That is, the accuracy of signal light recognition is improved.

[0043] Furthermore, the high beams can illuminate at least 100 meters ahead of vehicle 1. Therefore, even if vehicle 1 is about 100 meters away from the traffic light, the traffic light can still be brightened, improving visibility. In other words, the recognition accuracy of the traffic light can be improved even at a distance. This is preferable from the viewpoint of achieving autonomous driving control with sufficient margin.

[0044] Furthermore, when the illumination light directed at the signal candidate position SC becomes stronger than the illumination light directed at locations other than SC, the contrast of the image IMG increases. By applying filtering or other methods to the image IMG to remove noise light from locations other than the signal candidate position SC, the signal candidate position SC can be made clearer. This also helps increase the recognition probability (detection probability) of traffic lights based on the image IMG. In addition, since noise light is removed, the possibility of misidentifying noise light as traffic lights is suppressed. In other words, the recognition accuracy of traffic lights is improved.

[0045] Improving the recognition accuracy of traffic lights contributes to improving the accuracy of autonomous driving control based on traffic light recognition results.

[0046] Furthermore, the brightening of the signal candidate position SC is also useful in remote assistance provided by a remote operator. In the case of remote assistance, the image IMG acquired by camera C is sent to the remote operator terminal on the remote operator's side. The remote operator views the image IMG to understand the situation around vehicle 1. At this time, if the signal candidate position SC in the image IMG brightens, the remote operator can easily identify whether a traffic light is present. In addition, even if the signal display is not visible, the remote operator can easily identify whether there are arrow signals or obstructed signals. These features are preferred from the viewpoint of accuracy of remote assistance.

[0047] 2-2. Light Reduction Treatment Figure 7 This is a conceptual diagram illustrating "dimming processing," another example of lamp control processing. Dimming processing is the process of reducing the illumination light (output of lamp 50) illuminating beyond the signal candidate position SC in response to a certain trigger. The trigger here is the identification of the signal candidate position SC. That is, the vehicle control system 10, in response to the identification of the signal candidate position SC, makes the illumination light illuminating beyond the signal candidate position SC weaker than before the identification of the signal candidate position SC.

[0048] exist Figure 7 In the example shown, before time ts, lamp 50 (high beam) is on. Typically, the illumination intensity before time ts is the default value. Afterward, at time ts, the intensity of the illumination light illuminating areas outside the signal candidate position SC is reduced. Lamp 50 can also be partially turned off so that areas outside the signal candidate position SC are not illuminated. On the other hand, the intensity of the illumination light targeting the signal candidate position SC remains unchanged. As a result, compared with... Figure 6 The same applies to the example shown, where the illumination light directed at the signal candidate position SC becomes stronger than the illumination light directed outside the signal candidate position SC.

[0049] The effect obtained through this light reduction process is as follows.

[0050] Because the illumination light directed at the signal candidate position SC becomes stronger than the illumination light directed at locations other than SC, the contrast of the image IMG increases. By applying filtering to the image IMG to remove noise light from locations other than SC, the signal candidate position SC becomes clearer. This also helps increase the recognition probability (detection probability) of traffic lights based on the image IMG. Furthermore, because noise light is removed, the misidentification of noise light as traffic lights is suppressed. In other words, the recognition accuracy of traffic lights is improved.

[0051] Furthermore, reflectors are often attached to structures on roads and roadsides. For signal recognition, reflected light from such reflectors becomes noise light. By reducing the illumination light directed beyond the signal candidate position SC, the generation of noise light can be suppressed. This improves the recognition accuracy of traffic lights.

[0052] Furthermore, the illumination light shining outside the signal candidate position SC weakens, reducing the brightness (illuminance) at that location. Consequently, objects outside the signal candidate position SC become difficult to see. As a result, the probability of misidentifying objects other than traffic lights as traffic lights decreases. This also contributes to improving the accuracy of traffic light recognition.

[0053] Furthermore, by partially reducing the amount of light emitted, power consumption can be reduced.

[0054] 2-3. Combination of brightening and dimming treatments Figure 8 The combination of the aforementioned "brightness enhancement" and "brightness reduction" is illustrated. In response to the recognition of a signal candidate position SC, the vehicle control system 10 makes the illumination light directed towards the signal candidate position SC stronger than before the recognition of the signal candidate position SC, and makes the illumination light directed away from the signal candidate position SC weaker than before the recognition of the signal candidate position SC. Thus, it is possible to obtain both the effects of the brightness enhancement and the effects of the brightness reduction.

[0055] 2-4. Attenuation treatment following brightening treatment Figure 9 This is a conceptual diagram illustrating the "attenuation processing" following the brightening processing. As described above, through the brightening processing, the illumination light shining on the signal candidate position SC becomes stronger, and the illuminance of the signal candidate position SC increases. Then, vehicle 1 moves closer to the signal candidate position SC. With the intensity of the illumination light output from vehicle 1's lamp 50 remaining constant, the illuminance of the signal candidate position SC increases as vehicle 1 approaches. However, when objects such as signal lights are illuminated more intensely than necessary, they may become difficult to see due to factors such as reflected light and glare.

[0056] Therefore, in order to prevent the signal candidate position SC from being illuminated too intensely as needed, the vehicle control system 10 may also perform attenuation processing after the brightness enhancement processing. Specifically, after the brightness enhancement processing, the vehicle control system 10 reduces the illumination light illuminating the signal candidate position SC as the vehicle 1 moves.

[0057] The attenuation rate of the illumination light intensity during the attenuation process can be dynamically set according to the speed of vehicle 1. In this case, it is set such that the attenuation rate of the illumination light intensity increases as the speed of vehicle 1 increases.

[0058] The intensity of the illumination light during the attenuation process can also be set according to the distance from vehicle 1 to the signal candidate position SC. In this case, it is set such that the intensity of the illumination light decreases as the distance from vehicle 1 to the signal candidate position SC decreases.

[0059] Alternatively, the illumination intensity from the lamp 50 on the output side can be set in a way that keeps the illuminance of the signal candidate position SC on the receiving side constant during the attenuation process. A constant value is an illuminance that allows for appropriate signal recognition.

[0060] Through the attenuation processing described above, the situation where the signal candidate position SC is illuminated more intensely than necessary after the brightness enhancement process is suppressed. As a result, the situation where objects such as traffic lights become difficult to see is suppressed. This also helps to improve the recognition accuracy of traffic lights.

[0061] 2-5. Effects As explained above, according to this embodiment, a potential signal candidate position SC for a traffic light is identified around vehicle 1. Then, in response to the identification of the signal candidate position SC, the aforementioned light control process is executed. The light control process includes at least one of the following: a process that makes the illumination light directed towards the signal candidate position SC stronger than before the identification of the signal candidate position SC, and a process that makes the illumination light directed towards the signal candidate position SC stronger than the illumination light directed towards areas other than the signal candidate position SC. Through this light control process, the signal candidate position is easily visible even in a dark environment, increasing the probability of traffic light identification (detection probability) in dark environments. That is, the accuracy of traffic light identification is improved.

[0062] 3. Screening Process As described above, light control processing is performed in response to the identification of signal candidate positions SC. For convenience, this light control processing is referred to as the "first light control processing." Through the first light control processing, the recognition accuracy of signal lights based on the image IMG is increased. As a result, there are cases where it is determined that a signal light actually exists at the signal candidate position SC, or that its existence is highly probable. Conversely, there are also cases where it is determined that a signal light does not exist at the signal candidate position SC, or that its existence is highly probable. Therefore, the signal candidate positions SC can be further narrowed down after the first light control processing. In other words, the screening of signal candidate positions SC becomes possible.

[0063] Figure 10 This is a conceptual diagram illustrating the filtering process. One or more signal candidate positions SC are captured in the image IMG obtained by camera C. After the first light control process, the vehicle control system 10 further performs signal recognition based on the image IMG to reduce the number of signal candidate positions SC. At this point, the vehicle control system 10 can reduce both the "number" and the "size" of the signal candidate positions SC.

[0064] For example, if the vehicle control system 10 determines that a traffic light is absent at a certain candidate signal position SC, or that the probability of its presence is less than a threshold, it excludes that candidate signal position SC from future candidate signal positions SC. As another example, if the vehicle control system 10 determines that a portion of a certain candidate signal position SC contains no traffic light, or that the probability of its presence is less than a threshold, it excludes that portion of the candidate signal position SC from the candidate signal position SC.

[0065] The positions excluded from the candidate signal positions SC through the screening process will be referred to as "excluded positions EX". The probability of a signal light existing at excluded positions EX is low. On the other hand, the probability of a signal light existing at the selected candidate signal positions SC is high.

[0066] By narrowing down the candidate signal locations (SC) to those with a high probability of containing a traffic light, traffic lights can be identified more accurately and quickly. In other words, this filtering process improves both the accuracy and speed of traffic light identification.

[0067] The vehicle control system 10 may also perform lamp control processing after the screening process. For convenience, the lamp control processing performed after the screening process will be referred to as the "second lamp control process" below. The second lamp control process is triggered by the execution of the screening process.

[0068] For example, there is no need to continuously illuminate the excluded position EX, which has been eliminated from the signal candidate position SC. Continuously illuminating the excluded position EX could actually generate unnecessary noise light and cause objects other than the signal light to be misidentified as signal lights. Therefore, in the second light control processing, the vehicle control system 10 performs light reduction processing on the excluded position EX (see reference). Figure 7 , Figure 8 In other words, the vehicle control system 10 reduces the illumination light directed towards the exclusion position EX compared to before the screening process. This light reduction process achieves the aforementioned effect, further improving the recognition accuracy of the traffic lights.

[0069] As another example, in the second light control processing, the vehicle control system 10 can also further enhance the brightness of the signal candidate position SC (see reference). Figure 6 , Figure 8 In other words, the vehicle control system 10 can also make the illumination light directed at the signal candidate position SC stronger than before the screening process. At this time, the vehicle control system 10 can perform... Figure 4 The diagram shows the use of "focusing" light from the left lamp 50L and the right lamp 50R to intensify the illumination light directed at the signal candidate position SC. This light enhancement process achieves the aforementioned effect, further improving the recognition accuracy of the traffic lights.

[0070] Furthermore, the vehicle control system 10 can also perform attenuation processing after the enhancement processing (see reference). Figure 9 Therefore, the situation where the signal candidate position SC is illuminated too intensely than necessary after the brightness enhancement process is suppressed. As a result, the situation where objects such as traffic lights become difficult to see is suppressed. This also helps to improve the recognition accuracy of traffic lights.

[0071] 4. Examples of objects to be illuminated Figure 11 This is a conceptual diagram illustrating an example of a traffic light as the object of illumination in this embodiment. First lane L1 is the lane in which vehicle 1 is traveling. First traffic light SG1 is a traffic light for first lane L1. In particular, first traffic light SG1 is located in front of vehicle 1. Vehicle 1 travels according to the signal display of first traffic light SG1. This first traffic light SG1 is at least the object of illumination. The signal candidate position SC contains locations where the first traffic light SG1 may be present.

[0072] Figure 12 This is a conceptual diagram illustrating another example of a traffic light as the object of illumination in this embodiment. The second lane L2 intersects with the first lane L1. Other vehicles 2 are present in the second lane L2. The second traffic light SG2 is a traffic light for the second lane L2. Other vehicles 2 drive according to the signal display of the second traffic light SG2. The second traffic light SG2 can also be the object of illumination. That is, the signal candidate position SC can include not only positions where the first traffic light SG1 might be present, but also positions where the second traffic light SG2 might be present. The vehicle control system 10 of vehicle 1 in the first lane L1 also illuminates the second traffic light SG2 in the second lane L2, thereby making it easy for other vehicles 2 in the second lane L2 to identify the second traffic light SG2. In other words, the vehicle control system 10 of vehicle 1 can assist other vehicles 2 in signal identification.

[0073] 5. Examples of vehicle control systems 5-1. Example of composition Figure 13 This is a block diagram illustrating an example configuration of the vehicle control system 10 according to this embodiment. The vehicle control system 10 includes a sensor group 20, a driving device 30, a communication device 40, lights 50, and a control device 100.

[0074] The sensor group 20 is mounted on the vehicle 1. The sensor group 20 includes an identification sensor 21, a vehicle status sensor 22, a position sensor 23, etc.

[0075] The identification sensor 21 identifies (detects) the conditions around the vehicle 1. The identification sensor 21 includes a camera C. The identification sensor 21 may also include LIDAR (Laser Imaging Detection and Ranging), radar, etc.

[0076] The vehicle status sensor 22 detects the status of the vehicle 1. For example, the vehicle status sensor 22 includes a speed sensor, an acceleration sensor, a yaw rate sensor, a steering angle sensor, etc.

[0077] Position sensor 23 detects the position and orientation of vehicle 1. An example of position sensor 23 is a GPS (Global Positioning System) sensor.

[0078] The driving mechanism 30 includes a steering mechanism, a drive mechanism, and a braking mechanism. The steering mechanism steers the wheels. For example, the steering mechanism includes an electric power steering (EPS) system. The drive mechanism is the power source that generates driving force. Examples of drive mechanisms include engines, electric motors, and in-wheel motors. The braking mechanism generates braking force.

[0079] The communication device 40 communicates with the outside of the vehicle 1. For example, the communication device 40 communicates with a management device outside the vehicle 1. Examples of management devices include a map management device that manages map information and an automatic driving management device that manages the automatic driving of the vehicle 1. As another example, the communication device 40 may also communicate with a remote operator terminal that provides remote assistance to the vehicle 1.

[0080] Lamp 50 is mounted on vehicle 1 and provides external illumination to vehicle 1. Furthermore, the light distribution of lamp 50 is variable. That is, lamp 50 is configured to freely change the illumination range and intensity. For example, lamp 50 includes multiple light sources. When the light source is an LED, lamp 50 includes an LED array composed of multiple LEDs. Each of the multiple light sources can be controlled independently.

[0081] The control device 100 controls the vehicle 1. The control device 100 includes one or more processors 110 (hereinafter simply referred to as processor 110) and one or more storage devices 120 (hereinafter simply referred to as storage device 120). The processor 110 performs various processes. For example, the processor 110 includes a CPU (Central Processing Unit). The storage device 120 stores various information. Examples of storage devices 120 include volatile memory, non-volatile memory, HDD (Hard Disk Drive), SSD (Solid State Drive), etc. The control device 100 may also include one or more ECUs (Electronic Control Units). A portion of the control device 100 may also be an external information processing device to the vehicle 1. In this case, a portion of the control device 100 communicates with the vehicle 1 and remotely controls the vehicle 1.

[0082] The vehicle control program PROG is a computer program used to control the vehicle 1. Various processes performed by the control device 100 are implemented by the processor 110 through the execution of the vehicle control program PROG. The vehicle control program PROG is stored in the storage device 120. Alternatively, the vehicle control program PROG may also be recorded on a computer-readable recording medium.

[0083] 5-2. Driving Environment Information The control device 100 uses the sensor group 20 to acquire driving environment information 200, which shows the driving environment of the vehicle 1. The driving environment information 200 is stored in the storage device 120.

[0084] Figure 14 This is a block diagram illustrating an example of driving environment information 200. Driving environment information 200 includes surrounding condition information 210, vehicle status information 220, vehicle location information 230, and map information 240.

[0085] Surrounding conditions information 210 is information showing the conditions around vehicle 1. Control device 100 uses recognition sensor 21 to recognize the conditions around vehicle 1 and obtain surrounding conditions information 210. For example, surrounding conditions information 210 includes an image (IMG) captured by camera C. As another example, surrounding conditions information 210 includes point cluster information obtained by LIDAR.

[0086] The surrounding environment information 210 also includes object information 212 related to objects around vehicle 1. Examples of objects include pedestrians, bicycles, two-wheeled vehicles, other vehicles (driving vehicles, parked vehicles, etc.), white lines, traffic lights, structures (e.g., utility poles, pedestrian overpasses), signs, obstacles, etc. Object information 212 shows the relative position and relative speed of the object relative to vehicle 1. For example, by analyzing the image IMG obtained by camera C, objects can be identified and their relative positions calculated. For example, control device 100 uses image recognition AI obtained through machine learning to identify objects such as traffic lights in the image IMG. In addition, based on point group information obtained by LIDAR, objects can also be identified and their relative positions and relative speeds obtained.

[0087] Vehicle status information 220 indicates the status of vehicle 1, including vehicle speed, acceleration, yaw rate, steering angle, etc. Control device 100 obtains vehicle status information 220 from vehicle status sensor 22. Vehicle status information 220 can also indicate the driving status of vehicle 1 (automatic driving / manual driving).

[0088] Vehicle position information 230 indicates the current position of vehicle 1. Control device 100 obtains vehicle position information 230 based on the detection results obtained from position sensor 23. Alternatively, control device 100 can also obtain high-precision vehicle position information 230 by utilizing well-known localization processing based on object information 212 and map information 240.

[0089] Map information 240 includes a typical navigation map. Map information 240 may also show lane configurations and road shapes. Map information 240 may also include location information for structures, traffic lights, signs, etc. Control device 100 retrieves the map information 240 for the desired area from a map database. The map database may be stored in storage device 120 or in a map management device external to vehicle 1. In the latter case, control device 100 communicates with the map management device via communication device 40 to obtain the required map information 240. Map information 240 is stored in storage device 120.

[0090] 5-3. Vehicle Driving Control The control device 100 performs "vehicle driving control" to control the movement of the vehicle 1. Vehicle driving control includes steering control, acceleration control, and deceleration control. The control device 100 performs vehicle driving control by controlling the driving mechanism 30. Specifically, the control device 100 performs steering control by controlling the steering mechanism. Furthermore, the control device 100 performs acceleration control by controlling the drive mechanism. Additionally, the control device 100 performs deceleration control by controlling the braking mechanism.

[0091] Furthermore, the control device 100 performs autonomous driving control based on the driving environment information 200. More specifically, the control device 100 generates a driving plan for the vehicle 1 based on the driving environment information 200. Examples of driving plans include maintaining the current driving lane, changing lanes, making left and right turns, and avoiding obstacles. Then, based on the driving environment information 200, the control device 100 generates a target trajectory required for the vehicle 1 to travel according to the driving plan. The target trajectory includes a target position and a target speed. The control device 100 then performs vehicle driving control in a manner that causes the vehicle 1 to follow the target trajectory.

[0092] 5-4. Communication Processing The control device 100 communicates with the outside of the vehicle 1 via the communication device 40. For example, the control device 100 communicates with a management device outside the vehicle 1 via the communication device 40. Examples of management devices include a map management device that manages map information and an automatic driving management device that manages the automatic driving of the vehicle 1. The control device 100 can also communicate with other vehicles via the communication device 40.

[0093] Furthermore, when vehicle 1 is the object of remote assistance performed by a remote operator, control device 100 communicates with a remote operator terminal on the remote operator side via communication device 40. Examples of remote assistance performed by the remote operator include recognition assistance, judgment assistance, and remote driving. When control device 100 determines that remote assistance from a remote operator is required, it sends an assistance request to the remote operator terminal. Additionally, control device 100 sends at least a portion of the driving environment information 200 to the remote operator terminal. In particular, control device 100 sends an image IMG obtained by camera C to the remote operator terminal. The remote operator terminal displays the received image IMG on a display device. The remote operator views the image IMG displayed on the display device, understands the situation around vehicle 1, and remotely assists in the actions of vehicle 1. Control device 100 receives instruction information from the remote operator terminal indicating the instructions performed by the remote operator. Then, control device 100 performs vehicle driving control according to the instruction information.

[0094] 5-5. Lighting Control Processing The control device 100 performs lamp control processing to control the lamp 50. As described above, the light distribution state of the lamp 50 is variable. That is, the lamp 50 is configured to freely change its illumination range and intensity. The control device 100 can control the lamp 50 to change its illumination range and intensity.

[0095] The control device 100 described in this embodiment performs light control processing in a manner that improves the recognition accuracy of signal lights in dark environments. Figure 15 This is a flowchart illustrating the processes associated with such lamp control processing.

[0096] 5-5-1. Signal Candidate Location Identification Processing (Step S110) In step S110, the control device 100 identifies potential signal candidate locations SC around the vehicle 1 where traffic lights may exist (refer to...). Figure 5 The identification of the candidate signal location SC uses sensor group 20. Various examples of methods for identifying the candidate signal location SC are described below. Furthermore, confirmation that a signal light is present at the candidate signal location SC is also possible.

[0097] <First Example> The control device 100 acquires an image IMG obtained by the camera C. Then, the control device 100 analyzes the image IMG and temporarily identifies traffic lights within the image IMG. For example, image recognition AI obtained through machine learning can be used. Alternatively, a group of traffic lights at an intersection can be temporarily identified. Then, the control device 100 sets the surrounding area (within a certain range) of the temporarily identified traffic lights as candidate signal positions SC. The relative position of the candidate signal position SC with respect to vehicle 1 is calculated based on the image IMG.

[0098] <Second Example> The control device 100 acquires an image IMG obtained by the camera C. Then, the control device 100 analyzes the image IMG and identifies the light sources of the signal display colors (green, yellow, red) within the image IMG. Next, the control device 100 sets the positions (within a certain range) around the light sources of the signal display colors as signal candidate positions SC. The relative position of the signal candidate position SC with respect to the vehicle 1 is calculated based on the image IMG.

[0099] <Example 3> When the traffic light in front of vehicle 1 is red, there may be a vehicle stopped below it. The taillights of the stopped vehicle are red. Therefore, control device 100 analyzes the image IMG and identifies the red light of the vehicle in front within the image IMG. Then, control device 100 sets the position (within a certain range) above the identified red light as a candidate signal position SC. The relative position of the candidate signal position SC with respect to vehicle 1 is calculated based on the image IMG.

[0100] <Example 4> There is a high probability that traffic lights are installed at the intersection. Therefore, the control device 100 analyzes the image IMG and identifies the intersection in the image IMG. Alternatively, the control device 100 can also obtain the positions of intersections around vehicle 1 based on vehicle position information 230 obtained by position sensor 23 and map information 240. Then, the control device 100 sets the position (within a certain range) above the intersection as the signal candidate position SC. The relative position of the signal candidate position SC with respect to vehicle 1 is calculated based on the image IMG, or based on vehicle position information 230 and map information 240.

[0101] <Example 5> Traffic lights are sometimes also installed on structures such as utility poles and pedestrian overpasses. Hereinafter, such structures associated with traffic lights will be referred to as "traffic light associated structures." The control device 100 identifies traffic light associated structures around vehicle 1 based on the identification results obtained from the identification sensor 21 (e.g., camera, LIDAR). That is, the control device 100 identifies traffic light associated structures based on the surrounding environment information 210 (image IMG, object information 212). As another example, map information 240 with the locations of structures can also be prepared. In this case, the control device 100 can obtain the locations of traffic light associated structures around vehicle 1 based on the vehicle position information 230 obtained by the position sensor 23 and the map information 240. Then, the control device 100 sets the locations (within a certain range) around the traffic light associated structures as candidate signal positions SC. The relative position of the candidate signal position SC with respect to vehicle 1 is calculated based on the object information 212, or based on the vehicle position information 230 and the map information 240.

[0102] 5-5-2. First Lamp Control Process (Step S120) In step S120 following step S110, the control device 100 performs a "first lamp control process" to control the lamp 50 based on the signal candidate position SC. As described in section 2 above, the first lamp control process includes at least one of "brightness enhancement processing" and "brightness reduction processing" (see [reference]). Figures 6-8 The first lamp control process can also include a "attenuation process" following the brightening process (see [reference]). Figure 9 ).

[0103] The relative position of the signal candidate position SC with respect to the vehicle 1 is calculated as described above. The installation position of the lamp 50 on the vehicle 1, the installation direction of the lamp 50, and the optical axis direction of each light source included in the lamp 50 are known information. Therefore, the control device 100 can control the lamp 50 to selectively illuminate the signal candidate position SC.

[0104] 5-5-3. Screening Process (Step S130) In step S130, following step S120, the control device 100 performs a "filtering process" to reduce the candidate signal positions SC (refer to section 3 above). For example, the control device 100 uses image recognition AI to identify traffic lights in the image IMG. As a result of the first light control process in step S120, the recognition accuracy of the traffic lights is improved. Therefore, in this step S130, the control device 100 is able to identify traffic lights in the image IMG with higher accuracy.

[0105] If the control device 100 determines that a signal candidate position SC does not contain a traffic light, or that the probability of its presence is less than a threshold, it excludes that signal candidate position SC from the list of future signal candidate positions SC. As another example, if the control device 100 determines that at least a portion of a signal candidate position SC does not contain a traffic light, or that the probability of its presence is less than a threshold, it excludes a portion of that signal candidate position SC from the list of signal candidate positions SC. The position excluded from the signal candidate position SC is called the exclusion position EX (refer to...). Figure 10 ).

[0106] 5-5-4. Second Lamp Control Process (Step S140) In step S140, following step S130, the control device 100 performs a "second lamp control process" to control the lamp 50 based on the signal candidate position SC or the exclusion position EX. As described in section 3 above, the second lamp control process includes at least one of a "brightness enhancement process" and a "brightness reduction process." The second lamp control process may also include a "brightness reduction process" following the brightness enhancement process.

[0107] Furthermore, the second light control process can be omitted.

[0108] 6. Switching the target of illumination at intersections Figure 16 and Figure 17 This is a conceptual diagram illustrating the switching of the illuminated object at an intersection. Here, we consider a scenario where multiple traffic lights are continuously installed at an intersection ahead of vehicle 1. These multiple traffic lights include a first traffic light SGa and a second traffic light SGb. The second traffic light SGb is farther from vehicle 1 than the first traffic light SGa. That is, from the perspective of vehicle 1, the first traffic light SGa is located closer to the front, and the second traffic light SGb is located further in.

[0109] The image IMG acquired by camera C contains both the first traffic light SGa and the second traffic light SGb. Control device 100 identifies both the first traffic light SG1 and the second traffic light SGb based on the image IMG. However, control device 100 does not simultaneously treat both the first traffic light SG1 and the second traffic light SGb as the illumination target of light 50. Control device 100 switches the illumination target of light 50.

[0110] More specifically, in such Figure 16 When vehicle 1 is far from the intersection, it is sufficient to recognize the signal display of the first traffic light SGa. There is no need to recognize the signal display of the second traffic light SGb. Therefore, the control device 100 sets the first traffic light SGa as the target of illumination and excludes the second traffic light SGb from the target. In other words, the control device 100 controls the lights 50 by illuminating the first traffic light SGa but not the second traffic light SGb. As a result, excess reflected light (noise light) from the second traffic light SGb is suppressed. Furthermore, power consumption is reduced.

[0111] After that, in such Figure 17 As vehicle 1 approaches the intersection to some extent, the light source switches from the first traffic light SGa to the second traffic light SGb. That is, the control device 100 controls the light 50 to illuminate the second traffic light SGb while illuminating the first traffic light SGa. This suppresses excess reflected light (noise light) from the first traffic light SGa. Furthermore, power consumption is reduced.

[0112] Figure 18 This is a flowchart illustrating the process associated with switching the irradiated object.

[0113] In step S200, the control device 100 uses camera C to identify the traffic lights in front of vehicle 1. If multiple traffic lights located at the same intersection are identified (step S200; yes), the process proceeds to step S210.

[0114] In step S210, the control device 100 sets the first signal light SGa on the near front side as the target of illumination, and excludes the second signal light SGb on the depth side from the target of illumination. That is, the control device 100 controls the lamp 50 in a manner that illuminates the first signal light SGa but not the second signal light SGb.

[0115] In step S220, the control device 100 predicts the time until the vehicle 1 reaches the position of the first traffic light SGa. Alternatively, the stop line near the first traffic light SGa can be used instead of the position of the first traffic light SGa (see reference). Figure 16The distance from vehicle 1 to the first traffic light SGa or the stop line is obtained based on object information 212. The speed and acceleration of vehicle 1 are obtained based on vehicle state information 220. Therefore, based on object information 212 and vehicle state information 220, control device 100 can calculate the predicted time until vehicle 1 reaches the position of the first traffic light SGa.

[0116] In step S230, the control device 100 determines whether the switching condition is met. The switching condition includes at least "the predicted time is less than a threshold". The switching condition may also include "vehicle 1 is going straight at the intersection". Whether vehicle 1 is going straight at the intersection can be determined, for example, based on the target trajectory of vehicle 1 in the autonomous driving control. If the predicted time is greater than the threshold, the switching condition is not met (step S230: No). In this case, the process returns to step S210. On the other hand, when the switching condition is met (step S230: Yes), the process proceeds to step S240.

[0117] In step S240, the control device 100 switches the target of illumination from the first signal light SGa to the second signal light SGb. That is, the control device 100 controls the light 50 in a manner that illuminates the second signal light SGb but not the first signal light SGa.

Claims

1. A vehicle control method, characterized in that, include: The vehicle uses sensors mounted on the vehicle to identify potential signal candidate locations around the vehicle where traffic lights may be present, and the vehicle is equipped with lights configured to change their light distribution. A first lamp control process is performed in response to the identification of the signal candidate position, wherein the first lamp control process includes at least one of the following: The illumination light directed at the candidate signal location is stronger than that before the candidate signal location was identified; The intensity of the illumination light directed toward the candidate signal location is greater than the intensity of the illumination light directed toward areas other than the candidate signal location. Images of the area surrounding the vehicle are acquired using a camera mounted on the vehicle. After the first light control processing, a filtering process is performed to reduce the candidate signal positions based on the image; and After the filtering process, a second lamp control process is performed to control the lamp. The second lamp control process includes reducing the illumination light directed toward the exclusion position compared to before the filtering process, wherein the exclusion position is a position that was excluded from the signal candidate position by the filtering process.

2. The vehicle control method according to claim 1, characterized in that, The first lamp control process includes: making the illumination light illuminating the signal candidate position stronger than before the signal candidate position was identified, and illuminating only the signal candidate position, without illuminating areas other than the signal candidate position.

3. The vehicle control method according to claim 1, characterized in that, The first light control process includes: making the illumination light illuminating the signal candidate position stronger than before the signal candidate position was identified, and then weakening the illumination light illuminating the signal candidate position as the vehicle approaches the signal candidate position.

4. The vehicle control method according to claim 3, characterized in that, The first lamp control process further includes: making the illumination light illuminating the area outside the signal candidate location weaker than before the signal candidate location was identified.

5. The vehicle control method according to claim 1, characterized in that, The second lamp control process includes making the illumination light directed at the signal candidate position stronger than before the screening process.

6. The vehicle control method according to claim 5, characterized in that, The second light control process includes: making the illumination light illuminating the signal candidate position stronger than before the screening process, and then weakening the illumination light illuminating the signal candidate position as the vehicle approaches the signal candidate position.

7. The vehicle control method according to any one of claims 1 to 6, characterized in that, The candidate signal locations include: There may be a position of the traffic light for the first lane in which the vehicle is traveling, and There may be a location for the traffic light for the second lane that intersects with the first lane.

8. The vehicle control method according to any one of claims 1 to 6, characterized in that, The sensors mounted on the vehicle include: Identification sensors, configured to identify the conditions around the vehicle; and A position sensor configured to acquire the position of the vehicle. The identification sensor includes a camera configured to acquire images of the area around the vehicle. The identification includes at least one of the following: Temporarily identify traffic lights in the image, and set the positions around the temporarily identified traffic lights as candidate signal positions; Identify the light source of the signal display color in the image, and set the positions around the light source of the signal display color as the signal candidate positions; Identify the red light of the vehicle ahead in the image, and set the position above the red light as the signal candidate position; Identify intersections in the image, or obtain the locations of intersections around the vehicle based on the vehicle's location and map information, and set the location above the intersection as the signal candidate location; and The identification sensor is used to identify traffic light-related structures around the vehicle, or the location of the traffic light-related structures is obtained based on the vehicle's location and map information, and the locations around the traffic light-related structures are set as the candidate locations for the traffic lights.

9. The vehicle control method according to any one of claims 1 to 6, characterized in that, Also includes: Using a camera mounted on the vehicle, multiple traffic lights at an intersection located in front of the vehicle are identified, wherein the multiple traffic lights include a first traffic light and a second traffic light that is farther away from the vehicle than the first traffic light; Calculate the predicted time until the vehicle reaches the position of the first traffic light; If the predicted time is above a threshold, the first traffic light is set as the target of illumination, and the second traffic light is excluded from the target of illumination; and If the switching condition, including the predicted time being less than the threshold, is met, the illuminated object is switched from the first signal light to the second signal light.

10. A vehicle control system, characterized in that, It has one or more processors configured to perform the following processes: The process of using sensors mounted on the vehicle to identify potential signal candidate locations around the vehicle, the vehicle having lights configured to change their light distribution state; A first lamp control process for controlling the lamp in response to the recognition of the signal candidate position, wherein the first lamp control process includes at least one of the following: The illumination light directed at the candidate signal location is stronger than that before the candidate signal location was identified; The illumination light directed at the candidate signal location is made more intense than the illumination light directed at areas outside the candidate signal location. Images of the area surrounding the vehicle are acquired using a camera mounted on the vehicle. After the first light control processing, a filtering process is performed to reduce the candidate signal positions based on the image; and After the filtering process, a second lamp control process is performed to control the lamp. The second lamp control process includes reducing the illumination light directed toward the exclusion position compared to before the filtering process, wherein the exclusion position is a position that was excluded from the signal candidate position by the filtering process.

11. A non-transitory storage medium storing commands executable by one or more processors and causing said one or more processors to perform the following functions, characterized in that, The functions include: The vehicle uses sensors mounted on the vehicle to identify potential signal candidate locations around the vehicle where traffic lights may be present, and the vehicle is equipped with lights configured to change their light distribution. A first lamp control process is performed in response to the identification of the signal candidate position, wherein the first lamp control process includes at least one of the following: The illumination light directed at the candidate signal location is stronger than that before the candidate signal location was identified; The illumination light directed at the candidate signal location is made more intense than the illumination light directed at areas outside the candidate signal location. Images of the area surrounding the vehicle are acquired using a camera mounted on the vehicle. After the first light control processing, a filtering process is performed to reduce the candidate signal positions based on the image; and After the filtering process, a second lamp control process is performed to control the lamp. The second lamp control process includes reducing the illumination light directed toward the exclusion position compared to before the filtering process, wherein the exclusion position is a position that was excluded from the signal candidate position by the filtering process.

Images