Wrong-way driving detection method and wrong-way driving detection device

Abstract

To provide a technology capable of early or further reliably detecting a vehicle in a reverse driving.SOLUTION: A reverse driving determination method of a vehicle includes the steps of acquiring front images and rear images of a vehicle, recognizing surrounding targets of the vehicle using the front images and the rear images, calculating reverse driving likelihood based on the recognized information of the targets, and determining that the vehicle drives reversely when the reverse driving likelihood exceeds a threshold. A high value is used for the calculation factor of the reversing driving likelihood as compared with otherwise when in the calculating step of the reverse driving likelihood, recognition information of s guiding display surface on traffic control signs is included in the recognition information of the targets using the rear images.SELECTED DRAWING: Figure 8

Description

[Technical Field]

[0001] The present disclosure relates to a method and apparatus for determining "wrong-way driving," which is when a vehicle is traveling in the opposite direction to the direction indicated by the road or lane. [Background technology]

[0002] Japanese Patent Application Laid-Open Publication No. 2007-293390 discloses a wrong-way driving warning device. This conventional device recognizes the installation position of a road sign included in an image ahead of the vehicle acquired from a camera. The conventional device also performs wrong-way driving determination based on the relative position of the installation position with respect to the driving lane. Specifically, if it is determined that the installation position is located on the left side of the driving lane, it is determined that the vehicle is driving in the correct direction. On the other hand, if it is determined that the installation position is located on the right side of the driving lane, it is determined that the vehicle is driving in the wrong direction. If it is determined that the vehicle is driving in the wrong direction, the conventional device warns the driver that the vehicle is driving in the wrong direction. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2007-293390 Summary of the Invention [Problem to be solved by the invention]

[0004] The above-mentioned conventional device can determine whether a vehicle is traveling in the wrong direction based on the recognition information of the installation positions of road signs using a forward image. However, from the viewpoint of traffic safety, it is desirable to detect the state of a vehicle traveling in the wrong direction earlier or more reliably, and therefore further improvements are required.

[0005] One object of the present disclosure is to provide a technology that can detect wrong-way driving of a vehicle earlier or more reliably. [Means for solving the problem]

[0006] The first aspect is a method for determining whether a vehicle is running in the wrong direction, and has the following features. The method comprises: acquiring a rear image of the vehicle; Recognizing a target included in the rear image; determining whether the vehicle is traveling in the wrong direction based on the recognition information of the target; Includes: The step of determining reverse running includes: The method includes a step of determining whether or not the recognition information of the target includes recognition information of the guide display surface of a traffic sign. If it is determined that the recognition information of the target object includes the recognition information of the guidance display surface, it is determined that the vehicle is traveling in the wrong direction.

[0007] The second aspect is a method for determining whether a vehicle is running in the wrong direction, and has the following features. The method comprises: acquiring front and rear images of a vehicle; Recognizing objects around the vehicle using the front and rear images; calculating a likelihood of reverse running based on the target recognition information; determining that the vehicle is traveling in the wrong direction when the likelihood of the wrong direction is equal to or greater than a threshold; Includes: In the calculation step of the likelihood of wrong-way driving, if the recognition information of the target using the rear image includes recognition information of the guidance display surface of a traffic sign, a higher value is used as the calculation factor of the likelihood of wrong-way driving than if this is not the case.

[0008] The third aspect of the present invention is the same as the second aspect but further includes the following features. The target recognition information includes recognition information of the lane in which the vehicle is traveling. The recognition information of the guidance display surface includes recognition information of the lane corresponding to the guidance display surface. The calculation step of the likelihood of wrong-way driving further includes a step of determining, when it is determined that the recognition information of the guidance display surface is included in the recognition information of the target using the rear image, whether the lane in which the vehicle is traveling recognized in the rear image matches the lane corresponding to the guidance display surface recognized in the rear image. If it is determined that the lane in which the vehicle is traveling, as recognized in the rear image, matches the lane corresponding to the guidance display surface, as recognized in the rear image, a higher value is used for the calculation factor of the likelihood of wrong-way driving than if it is determined that this is not the case.

[0009] The fourth aspect of the present invention is the same as the second aspect, but further includes the following features. The target recognition information includes recognition information of the total number of lanes on the road on which the vehicle is traveling. The calculation step of the likelihood of wrong-way driving further includes a step of determining whether the total number of lanes is one when it is determined that the recognition information of the guidance display surface is included in the recognition information of the target using the rear image. If the total number of lanes is determined to be 1, a higher value is used for the calculation factor of the wrong-way running likelihood than if it is determined not to be 1.

[0010] The fifth aspect of the present invention is the second aspect of the present invention, further comprising the following features: The recognition information of the guidance display surface includes information on the time when at least one of the front and rear images including the recognition information of the guidance display surface was acquired. The calculation step of the likelihood of reverse driving further includes a step of determining, when it is determined that the recognition information of the guidance display surface is included in the recognition information of the target using the rear image, whether or not recognition information of the guidance display surface of a traffic sign other than the guidance display surface is obtained in the recognition information of the target using the forward image until the elapsed time counted from the time the rear image is acquired exceeds a predetermined time. If it is determined that recognition information of a guidance display surface of a traffic sign other than the guidance display surface has been obtained before the elapsed time exceeds a predetermined time, a higher value is used for the calculation factor of the likelihood of wrong-way driving than if it is determined that this is not the case.

[0011] A sixth aspect is a device for determining whether a vehicle is running in the wrong direction, and has the following features. The device includes a memory in which a rear image of the vehicle is stored, and a processor. The processor: Recognizing targets around the vehicle using the rear image; Based on the recognition information of the target, it is determined that the vehicle is traveling in the wrong direction. In determining the wrong-way driving, the processor determining whether the recognition information of the target includes recognition information of a guidance display surface of a traffic sign; If it is determined that the recognition information of the target object includes the recognition information of the guidance display surface, it is determined that the vehicle is traveling in the wrong direction.

[0012] The seventh aspect is a device for determining whether a vehicle is running in the wrong direction, and has the following features. The device comprises a memory in which front and rear images of the vehicle are stored, and a processor. The processor: Recognizing targets around the vehicle using the front and rear images; Calculating a likelihood of reverse running based on the recognition information of the target; If the likelihood of wrong-way driving is equal to or greater than a threshold, it is determined that the vehicle is driving in the wrong direction. In calculating the likelihood of wrong-way driving, if the recognition information of the target using the rear image includes recognition information of the guidance display surface of a traffic sign, the processor uses a higher value as the calculation factor of the likelihood of wrong-way driving than if this is not the case.

[0013] The eighth aspect of the present invention is the seventh aspect further characterized by the following: The target recognition information includes recognition information of the lane in which the vehicle is traveling. The recognition information of the guidance display surface includes recognition information of the lane corresponding to the guidance display surface. The processor further includes, in calculating the likelihood of reverse running, if it is determined that the recognition information of the guidance display surface is included in the recognition information of the target using the rear image, it is determined whether or not the lane in which the vehicle is traveling recognized in the rear image coincides with the lane corresponding to the guidance display surface recognized in the rear image; If it is determined that the lane in which the vehicle is traveling, as recognized in the rear image, matches the lane corresponding to the guidance display surface, as recognized in the rear image, a higher value is used for the calculation factor of the likelihood of wrong-way driving than if it is determined that this is not the case.

[0014] The ninth aspect of the present invention is the seventh aspect further characterized by the following: The target recognition information includes recognition information of the total number of lanes on the road on which the vehicle is traveling. The processor further includes, in calculating the likelihood of reverse running, If it is determined that the recognition information of the guidance display surface is included in the recognition information of the target using the rear image, it is determined whether the total number of lanes is 1; If the total number of lanes is determined to be 1, a higher value is used as the calculation factor for the likelihood of wrong-way driving compared to when it is determined not to be 1.

[0015] The tenth aspect of the present invention is the seventh aspect further characterized by the following: The recognition information of the guidance display surface includes information on the time when at least one of the front and rear images including the recognition information of the guidance display surface was acquired. The processor further includes, in calculating the likelihood of reverse running, When it is determined that the recognition information of the guidance display surface is included in the recognition information of the target using the rear image, it is determined whether or not recognition information of a guidance display surface of a traffic sign other than the guidance display surface has been obtained in the recognition information of the target using the forward image until the elapsed time counted from the time the rear image was acquired exceeds a predetermined time, If it is determined that recognition information of a guidance display surface of a traffic sign other than the guidance display surface has been obtained before the elapsed time exceeds a predetermined time, a higher value is used for the calculation factor of the likelihood of wrong-way driving compared to when it is determined that this is not the case. [Effects of the Invention]

[0016] According to the first or fifth aspect, if the recognition information of the target using the rear image includes the recognition information of the guide display surface of a traffic sign, it can be determined that the vehicle is traveling in the wrong direction. Therefore, it becomes possible to detect early that the vehicle is traveling in the wrong direction.

[0017] According to the second or seventh aspect, when the recognition information of the target using the rear image includes the recognition information of the guide display surface of a traffic sign, a high value is used as the calculation factor of the wrong-way driving likelihood, so that the wrong-way driving likelihood is more likely to be determined to be equal to or greater than the threshold value, thereby enabling earlier detection of the vehicle traveling in the wrong direction.

[0018] According to the third, fourth, fifth, eighth, ninth or tenth aspect, it is possible to avoid a problem in which a vehicle is erroneously determined to be traveling in the wrong direction even though it is actually traveling in the correct direction. In other words, it is possible to reliably detect that a vehicle is traveling in the wrong direction. [Brief explanation of the drawings]

[0019] [Figure 1] FIG. 1 is a diagram illustrating an outline of an embodiment of the present invention. [Figure 2] FIG. 1 is a diagram illustrating an outline of an embodiment. [Figure 3] 1 is a block diagram illustrating an example of the configuration of a wrong-way driving determination device according to an embodiment. [Figure 4] 4 is a block diagram illustrating an example of a functional configuration of the information processing device illustrated in FIG. 3. FIG. [Figure 5] 10A and 10B are diagrams illustrating a case where a determination is made based on recognition information of a lane corresponding to a guidance display surface. [Figure 6] FIG. 10 is a diagram illustrating a case where determination is made based on the total number of lanes on the road on which the vehicle is traveling. [Figure 7] 10A and 10B are diagrams illustrating a case where a determination is made based on the time when a rear image in which a guidance display surface of a traffic sign is recognized is acquired. [Figure 8] 4 is a flowchart showing a flow of processing by the information processing device shown in FIG. 3. DETAILED DESCRIPTION OF THE INVENTION

[0020] Hereinafter, a wrong-way driving detection method and wrong-way driving detection device according to an embodiment of the present invention will be described with reference to the drawings. The wrong-way driving detection method according to the embodiment is realized by computer processing performed in the wrong-way driving detection device. In addition, in each drawing, the same or corresponding parts are assigned the same reference numerals, and their description will be simplified or omitted.

[0021] 1. Overview of the embodiment 1-1. Example of a situation when vehicles are traveling in order Fig. 1 is a diagram illustrating an outline of an embodiment of the present invention. Fig. 1 depicts a vehicle M traveling on a lane L1. The direction DR_M in which the vehicle M travels is the same as the direction DR_L1 indicated by the lane L1. In other words, the vehicle M shown in Fig. 1 is traveling "forward" on the lane L1.

[0022] Vehicle M is equipped with a front camera 10. FIG. 1 schematically illustrates an angle of view A10 of the front camera 10, and a traffic sign TM1 is located in the range of this angle of view A10 on the left front side of vehicle M. A front image IMG_F illustrated in FIG. 1 is an example of an image acquired by the front camera 10. The front image IMG_F includes the guidance display surface (the surface displaying the speed limit) of traffic sign TM1 and the lane in which vehicle M is traveling (i.e., lane L1).

[0023] 1-2.Example of a situation when a vehicle is traveling in the wrong direction Fig. 2 is also a diagram for explaining an outline of an embodiment. Like Fig. 1, Fig. 2 depicts a vehicle M traveling on a lane L1. However, unlike Fig. 1, the direction DR_M in which the vehicle M travels is opposite to the direction DR_L1 indicated by the lane L1. In other words, the vehicle M shown in Fig. 2 is "traveling in the wrong direction" on the lane L1.

[0024] Vehicle M is equipped with a rear camera 20 in addition to the front camera 10 shown in FIG. 1. FIG. 2 schematically illustrates the angle of view A20 of the rear camera 20, and a traffic sign TM1 is located within the range of this angle of view A20. The rear image IMG_R illustrated in FIG. 2 is an example of an image acquired by the rear camera 20. The rear image IMG_R includes the guidance display surface of the traffic sign TM1 and the lane in which vehicle M is traveling (i.e., lane L1).

[0025] 1-3.Wrong-way driving detection In the embodiment, a "wrong-way driving determination" is performed to determine whether or not the vehicle M is driving in the wrong direction based on the recognition information of the object included in the forward image IMG_F and the rearward image IMG_R. In the wrong-way driving determination, it is determined whether or not the wrong-way driving likelihood H is equal to or greater than a threshold value TH. If it is determined that the wrong-way driving likelihood H is equal to or greater than the threshold value TH, it is determined that the vehicle M is driving in the wrong direction.

[0026] The wrong-way driving likelihood H is calculated, for example, by the following formula (1) which indicates the sum of a road type factor ηr and a wrong-way driving factor ηk (1≦k≦n). H=ηr+Σηk (1)

[0027] The road type factor ηr is a factor (calculation factor) used in calculating the wrong-way driving likelihood H, and is set according to the type of road on which the vehicle M is traveling (specifically, an ordinary road or an expressway). The type of road is estimated, for example, based on information about the lane markings on the road included in the forward image IMG_F, recognition information about structures existing around the road, and the like. For example, some lane markings on ordinary roads are yellow, while those on expressways are not. Therefore, the type of road can be estimated from color information about the lane markings. Furthermore, expressways are surrounded by structures unique to expressways, such as medians and sound barriers. Therefore, the type of road can be estimated from information about the structures. The type of road may also be estimated based on color information about the road included in the rearward image IMG_R, recognition information about structures existing around the road, and the like.

[0028] In this embodiment, a road type factor ηr is set when the type of road on which the vehicle M is traveling is estimated to be an expressway, and a road type factor ηr is set when the type of road on which the vehicle M is traveling is estimated to be an ordinary road, with the former set to a higher value than the latter, in order to detect wrong-way driving while traveling on an expressway at an earlier stage.

[0029] The wrong-way driving factor ηk is a factor (calculation factor) used to calculate the wrong-way driving likelihood H, and is set based on events recognized while the vehicle is traveling (total number of events = n). An example of an event recognized while the vehicle is traveling is the recognition of the direction of traffic signs (road paint) painted on the lane. Here, traffic signs on the lane are painted in a direction that makes them easy for the driver of the vehicle to recognize. In other words, there is a certain standard for the direction in which traffic signs are painted on the lane. Therefore, if the recognized direction of a traffic sign is the same as the vehicle's direction of travel, it can be said to be an event recognized while the vehicle is traveling forward, and if the recognized direction of a traffic sign is opposite to the vehicle's direction of travel, it can be said to be an event recognized while the vehicle is traveling backward.

[0030] In this embodiment, a wrong-way driving factor ηk is set when a traffic sign on lane L1 is recognized from the forward image IMG_F and the drawing direction of this traffic sign is recognized to be reversed from the specified direction. Also, a wrong-way driving factor ηk corresponding to a traffic sign on lane L1 is set when a traffic sign is recognized from the rearward image IMG_R and the drawing direction of this traffic sign is recognized to be the same as the specified direction. The former and latter are set to, for example, the same value.

[0031] Another example of an event recognized while a vehicle is traveling is the recognition of the guidance display surface of a traffic sign installed around a lane. "Guidance display surface" is a general term for a surface that displays guidance, warnings, instructions, road traffic prohibitions, regulations, etc. for road users. For example, the recognition of the guidance display surface of a traffic sign (e.g., traffic sign TM1 shown in FIG. 1) located on the left side of lane L1 in the forward image IMG_F can be said to be an event recognized while the vehicle is traveling in the forward direction. On the other hand, the recognition of the back side of a traffic sign (i.e., the surface opposite to the guidance display surface) located on the right side of lane L1 in the forward image IMG_F can be said to be an event recognized while the vehicle is traveling in the wrong direction.

[0032] In this embodiment, a wrong-way driving factor ηk is set corresponding to the case where the back faces of traffic signs located on the right side of lane L1 are recognized in the forward image IMG_F. For example, the wrong-way driving factor ηk is set to a higher value as the number of recognized back faces of traffic signs located on the right side of lane L1 in the forward image IMG_F increases.

[0033] Regarding the guidance display surfaces of traffic signs installed around lane L1, recognizing a guidance display surface of "No Entry for Vehicles" included in the forward image IMG_F can be said to be an event recognized during wrong-way driving of a vehicle. In the embodiment, a wrong-way driving factor ηk is set corresponding to the case where a guidance display surface of "No Entry for Vehicles" located on the left side of lane L1 is recognized in the forward image IMG_F.

[0034] 1-4. Characteristics of wrong-way driving detection However, because traffic signs on lane L1 are drawn so that they are easy for the driver of vehicle M to recognize visually, it is expected that they may not be recognized in the forward image IMG_F. Furthermore, because the back of a traffic sign has few distinctive features for identifying it, it is expected that it will be difficult to recognize the back of a traffic sign located on the right side of lane L1 in the forward image IMG_F. In this regard, it can be said that the recognition information of the guidance display surface of "No Entry for Vehicles" is easily recognized in the forward image IMG_F. However, depending on the installation position of the No Entry for Vehicles sign and the orientation of this guidance display surface, it is expected that the guidance display surface of "No Entry for Vehicles" may not be recognized in the forward image IMG_F.

[0035] Therefore, in the embodiment, a new wrong-way driving factor ηk is set as a wrong-way driving factor ηk corresponding to the case where the guidance display surface of a traffic sign located on the left side of lane L1 is recognized in the rearward image IMG_R. For example, the wrong-way driving factor ηk is set to a higher value as the number of guidance display surfaces of traffic signs located on the left side of lane L1 recognized in the rearward image IMG_R increases.

[0036] Unlike traffic signs on lane L1 and the backsides of traffic signs placed around lane L1, the guidance display surfaces of traffic signs included in the rear image IMG_R are more easily recognized in the rear image IMG_R. Also, the number of guidance display surfaces of traffic signs located to the left of lane L1 in the rear image IMG_R is theoretically the same as the number of backsides of traffic signs located to the right of lane L1 in the forward image IMG_F. Therefore, the guidance display surfaces of traffic signs located to the left of lane L1 in the rear image IMG_R are more likely to be recognized than no entry signs located to the left of lane L1 in the forward image IMG_F.

[0037] Therefore, when the guidance display surface of such a traffic sign is recognized in the rear image IMG_R, a corresponding wrong-way driving factor ηk is set, which makes it easier to determine that the wrong-way driving likelihood H is equal to or greater than the threshold value TH. In this way, according to the embodiment, by increasing the opportunities to calculate the wrong-way driving factor ηk, it becomes possible to detect earlier that the vehicle M is driving in the wrong direction. This leads to improved traffic safety around the vehicle M.

[0038] The wrong-way driving detection method and wrong-way driving detection device according to the embodiment will be described in detail below.

[0039] 2. Wrong-way driving detection device 2-1.Configuration example Fig. 3 is a block diagram illustrating an example of the configuration of a wrong-way driving determination device according to an embodiment. In the example shown in Fig. 3, a wrong-way driving determination device 100 includes a front camera 10, a rear camera 20, an information processing device 30, and an HMI (Human Machine Interface) unit 40. The wrong-way driving determination device 100 is mounted on a vehicle M shown in Fig. 1 or 2.

[0040] The front camera 10 is a digital camera that captures an image in front of the vehicle M. The front camera 10 has a built-in imaging element such as a CCD (Charge Coupled Device) or a CIS (CMOS Image Sensor). The front camera 10 has a wide-angle lens or a fisheye lens, and captures an image with a horizontal angle of view A10 (for example, in the range of 140° to 220°). The front camera 10 transmits a front image IMG_F to the information processing device 30.

[0041] The rear camera 20 is a digital camera that captures an image behind the vehicle M. An example of the basic configuration of the rear camera 20 is the same as that of the front camera 10. The rear camera 20 captures an image with a horizontal angle of view A20 (for example, in the range of 140° to 220°). The rear camera 20 transmits a rear image IMG_R to the information processing device 30.

[0042] The information processing device 30 includes at least one processor 31 and at least one memory 32. The processor 31 executes various processes. Examples of the memory 32 include a volatile memory and a non-volatile memory. The memory 32 stores various data. The various data include a forward image IMG_F from the front camera 10 and a rearward image IMG_R from the rear camera 20. The processor 31 executes various computer programs to realize various functions. The various functions will be described later. The computer programs are stored in the memory 32 or recorded on a computer-readable recording medium.

[0043] The HMI unit 40 provides information to the driver of the vehicle M and also receives information from the driver. The HMI unit 40 includes, for example, an input device, a display device, a speaker, and a microphone. Examples of the input device include a touch panel, a keyboard, a switch, and a button. The information provided to the driver includes the driving status of the vehicle M and predetermined warnings. Information is provided to the driver using, for example, a display device and a speaker.

[0044] 2-2. Example of functional configuration of information processing device Fig. 4 is a block diagram illustrating an example of the functional configuration of the information processing device 30 shown in Fig. 3. In the example shown in Fig. 4, the information processing device 30 includes a target recognition unit 33, a wrong-way driving likelihood calculation unit 34, a wrong-way driving determination unit 35, and an attention calling unit 36. These functional blocks are realized by a processor 31 executing software for wrong-way driving determination, which is a computer program.

[0045] The target recognition unit 33 performs a process (target recognition process) to recognize targets around the vehicle M. The target recognition process is performed based on each of the front image IMG_F and the rear image IMG_R. Known methods such as machine learning are applied as a method for target recognition based on the front image IMG_F and the rear image IMG_R. The target recognition process recognizes static and dynamic targets that exist around the vehicle M. Examples of static targets include lane markings on which the vehicle M is traveling, guardrails, structures, etc. Examples of dynamic targets include pedestrians, bicycles, motorcycles, and automobiles.

[0046] Information on the targets recognized by the target recognition process (i.e., target recognition information) is transmitted to the reverse-running likelihood calculation unit 34. Here, the target recognition information includes recognition information on the road on which the vehicle M is traveling. This road recognition information may include information on the type of road on which the vehicle M is traveling (general road, expressway). The target recognition information may also include recognition information on traffic signs on the lane on which the vehicle M is traveling. This recognition information on the traffic signs on the lane may include information on the drawing direction of these traffic signs.

[0047] The recognition information of the target object may further include recognition information of the front and back of a guidance sign installed around the lane in which the vehicle M is traveling. The recognition information of the guidance sign may also include recognition information of a guidance sign surface indicating no entry for vehicles. Furthermore, the recognition information of the guidance sign surface may include information about the lane corresponding to this guidance sign surface. The "lane corresponding to the guidance sign surface" refers to the lane to which the information displayed on this guidance sign surface is transmitted. The "lane corresponding to the guidance sign surface" is estimated, for example, based on domestic standards regarding the installation method of traffic signs.

[0048] In Japan, most traffic signs are installed on the roadside in accordance with standards set by the Ministry of Land, Infrastructure, Transport and Tourism. Some traffic signs are installed as cantilevers (overhanging), gates (overhead), or attached signs (using crosswalks, utility poles, etc.). Therefore, when a guide display surface is recognized in the forward image IMG_F, the lane recognized to the right or below the recognition area of this guide display surface is presumed to be the "lane corresponding to the guide display surface." Also, when a guide display surface is recognized in the rearward image IMG_R, the lane recognized to the left or below the recognition area of this guide display surface is presumed to be the "lane corresponding to the guide display surface."

[0049] The wrong-way driving likelihood calculation unit 34 calculates the wrong-way driving likelihood H based on the target recognition information received from the target recognition unit 33. The wrong-way driving likelihood H is calculated based on, for example, a calculation formula (see formula (1) above) with a road type factor ηr and a wrong-way driving factor ηk (1≦k≦n) as variables.

[0050] As already explained, the road type factor ηr is set corresponding to the type of road on which the vehicle M is traveling (specifically, an ordinary road or an expressway). The wrong-way driving factor ηk is set based on events recognized while the vehicle is traveling. The road type factor ηr and the wrong-way driving factor ηk are stored in a predetermined database, for example, as map data. The wrong-way driving likelihood calculation unit 34 identifies the road type factor ηr and the wrong-way driving factor ηk corresponding to the recognition information received from the target recognition unit 33 by referring to map data that uses this recognition information. The wrong-way driving likelihood calculation unit 34 then calculates the wrong-way driving likelihood H based on the road type factor ηr and the wrong-way driving factor ηk obtained by referring to this map data and a calculation formula that uses these as variables.

[0051] The wrong-way driving likelihood calculation unit 34 may additionally perform the following determination when identifying the road type factor ηr and the wrong-way driving factor ηk corresponding to the recognition information of the target object.

[0052] An example of a first additional determination is a determination based on recognition information of a lane corresponding to a guidance display screen. FIG. 5 is a diagram illustrating a case where a determination is made based on recognition information of a lane corresponding to a guidance display screen. In the example shown in FIG. 5, vehicle M is traveling on lane L2. The direction DR_M in which vehicle M is traveling is the same as the direction DR_L2 indicated by lane L2. In other words, vehicle M shown in FIG. 5 is traveling "forward" on lane L2. In the example shown in FIG. 5, lane L3 is also present near lane L2. Lane L3 is, for example, a lane that forms a Y-shaped intersection with lane L2. The direction DR_L3 indicated by lane L3 is the opposite direction to direction DR_L2.

[0053] 5, traffic sign TM2 is present within the range of angle of view A20. Therefore, the target recognition information received from the target recognition unit 33 includes recognition information of the guidance display surface of traffic sign TM2 in the rearward image IMG_R. However, the destination of the information displayed on the guidance display surface of this traffic sign TM2 is lane L3, not lane L2. Therefore, if only recognition information of the guidance display surface of traffic sign TM2 in the rearward image IMG_R is received, there is a possibility that the wrong-way driving factor ηk will be identified by referring to map data based on this recognition information.

[0054] In the first additional determination, if the target recognition information received from the target recognition unit 33 includes information about the lane corresponding to the guidance display surface of the traffic sign TM2 in the rear image IMG_R (i.e., lane L3), this information is used. Then, it is determined whether the lane corresponding to the guidance display surface of the traffic sign TM2 matches the lane in which the vehicle M is traveling (i.e., lane L2). Note that in the example shown in FIG. 5, it is determined that the two do not match, so the wrong-way driving factor ηk is not identified. Furthermore, it is assumed that the recognition information of the lane in which the vehicle M is traveling is included in the target recognition information received from the target recognition unit 33.

[0055] An example of a second additional determination is a determination based on the total number of lanes on the road on which vehicle M is traveling. FIG. 6 is a diagram illustrating a case where a determination based on the total number of lanes on the road on which vehicle M is traveling is made. In the example shown in FIG. 6, vehicle M is traveling on lane L4. The direction DR_M in which vehicle M is traveling is the same as the direction DR_L4 indicated by lane L4. In other words, vehicle M shown in FIG. 6 is traveling "forward" on lane L4. In the example shown in FIG. 6, there is also a lane L5 adjacent to lane L4. Lane L5 is, for example, an oncoming lane of lane L4. The direction DR_L5 indicated by lane L5 is the opposite direction to direction DR_L4.

[0056] 6, traffic sign TM3 is present within the range of angle of view A20. Therefore, the target recognition information received from the target recognition unit 33 includes recognition information of the guidance display surface of traffic sign TM3 in the rear image IMG_R. However, the information displayed on the guidance display surface of this traffic sign TM3 is transmitted to lane L5, not lane L4. Therefore, if only recognition information of the guidance display surface of traffic sign TM3 in the rear image IMG_R is received, there is a possibility that the wrong-way driving factor ηk will be identified by referring to map data based on this recognition information.

[0057] In the second additional determination, if the target recognition information received from the target recognition unit 33 includes information regarding the total number of lanes (i.e., two) of the road on which vehicle M is traveling, this information is used. Then, it is determined whether the total number of lanes of the road on which vehicle M is traveling is one. Note that in the example shown in FIG. 6, since it is determined that the total number of lanes is not one, the wrong-way driving factor ηk is not identified. Furthermore, it is assumed that the information regarding the total number of lanes of the road on which vehicle M is traveling is included in the target recognition information received from the target recognition unit 33.

[0058] A third additional determination may be a determination based on the time when a rearward image IMG_R in which the guidance display surface of a traffic sign is recognized is acquired. FIG. 7 is a diagram illustrating a case in which a determination is made based on the time when a rearward image IMG_R in which the guidance display surface of a traffic sign is recognized is acquired. In the example shown in FIG. 7, vehicle M is traveling on lane L6. Lane L6 is a lane that allows two-way traffic. Therefore, vehicle M shown in FIG. 7 is traveling "forward" on lane L6.

[0059] 7, traffic sign TM4 is present within the range of angle of view A20. Therefore, the recognition information of the target received from the target recognition unit 33 includes recognition information of the guidance display surface of the traffic sign TM4 in the rear image IMG_R. However, the destination of the information displayed on the guidance display surface of this traffic sign TM4 is a vehicle traveling in lane L6 in the opposite direction to the traveling direction DR_M of vehicle M. Therefore, if only recognition information of the guidance display surface of traffic sign TM4 in the rear image IMG_R is received, there is a possibility that the wrong-way driving factor ηk will be identified by referring to map data based on this recognition information.

[0060] In the third additional determination, if the target recognition information received from the target recognition unit 33 includes information regarding the time when the rearward image IMG_R in which the guidance display surface of a traffic sign was recognized was acquired, this information is used. Then, until the elapsed time counted from this time exceeds a predetermined time, it is determined whether information regarding the time when the forward image IMG_F in which the guidance display surface of a traffic sign other than the guidance display surface was recognized was received from the target recognition unit 33. Note that in the example shown in FIG. 7, since recognition information of the guidance display surface of the traffic sign TM5 located ahead of the vehicle M is acquired, the result of this determination is negative, and the wrong-way driving factor ηk is not identified. Furthermore, it is assumed that information regarding the time when the rearward image IMG_R in which the guidance display surface of a traffic sign was recognized was acquired and the time when the rearward image IMG_R in which a different guidance display surface was recognized was acquired is included in the target recognition information received from the target recognition unit 33.

[0061] The wrong-way driving determination unit 35 determines whether the vehicle M is driving in the wrong direction. The wrong-way driving determination unit 35 determines whether the wrong-way driving likelihood H received from the wrong-way driving likelihood calculation unit 34 is equal to or greater than a threshold value TH. If it is determined that the wrong-way driving likelihood H is equal to or greater than the threshold value TH, it is determined that the vehicle M is driving in the wrong direction. If it is determined that the vehicle M is driving in the wrong direction, the wrong-way driving determination unit 35 transmits a signal indicating this wrong-way driving determination to the attention alert unit 36.

[0062] When the warning unit 36 receives a signal indicating a wrong-way driving determination from the wrong-way driving determination unit 35, it issues a predetermined warning to the driver of the vehicle M. The predetermined warning is issued, for example, by displaying a warning message such as "You are driving in the wrong direction" on the HMI unit 40 (display device). As another example, the warning is issued by playing a voice message saying "You are driving in the wrong direction" from the HMI unit 40 (speaker).

[0063] 2-3.Examples of information processing by information processing device Fig. 8 is a flowchart showing the flow of processing by the information processing device 30 (processor 31) shown in Fig. 3. The flowchart shown in Fig. 8 is repeatedly executed at a predetermined control period.

[0064] 8, first, various information is acquired (step S11). Examples of this information include a front image IMG_F and a rear image IMG_R. The various information may include information on the time when the front image IMG_F and the rear image IMG_R were acquired.

[0065] Following the processing of step S11, target recognition is performed (step S12). The target recognition is performed based on each of the forward image IMG_F and the rearward image IMG_R. A known method such as machine learning is applied to the target recognition based on the forward image IMG_F and the rearward image IMG_R. According to the target recognition processing, static and dynamic targets existing around the vehicle M are recognized.

[0066] Following the processing of step S12, the wrong-way driving likelihood H is calculated (step S13). The processing of step S13 is performed based on the target recognition information obtained by the processing of step S12. As already explained, this target recognition information includes information for identifying the road type factor ηr and the wrong-way driving factor ηk. Therefore, in the processing of step S13, the road type factor ηr and the wrong-way driving factor ηk corresponding to this recognition information are identified by referring to map data using the target recognition information. Then, the wrong-way driving likelihood H is calculated based on the identified road type factor ηr and wrong-way driving factor ηk and a calculation formula (the above formula (1)) that uses these as variables.

[0067] In the process of step S13, if additional information is included in the target recognition information obtained by the process of step S12, first to third additional determinations may be performed using this additional information. By performing the first to third additional determinations, it becomes possible to avoid the problem of identifying the reverse running factor ηk.

[0068] Following the processing of step S13, it is determined whether the wrong-way running likelihood H is equal to or greater than a threshold value TH (step S14). If the determination result of step S14 is affirmative, a predetermined warning is issued (step S15). Examples of the warning are as already described.

[0069] 3.Effects According to the embodiment described above, the wrong-way driving factor ηk is calculated based on target recognition information using the rearward image IMG_R in addition to target recognition information using the forward image IMG_F. That is, the opportunities for calculating the wrong-way driving factor ηk are increased. Therefore, it becomes possible to detect that the vehicle M is driving in the wrong direction at an earlier stage. This leads to improved traffic safety around the vehicle M.

[0070] Furthermore, according to the embodiment, when the first to third additional determinations are made, it is possible to avoid the problem of the wrong-way driving factor ηk being identified. Therefore, it is possible to reliably detect that the vehicle M is running in the wrong direction.

[0071] 4. Modified version of wrong-way driving detection In the above embodiment, the wrong-way driving likelihood H is calculated based on the recognition information of the target included in the forward image IMG_F and the backward image IMG_R. However, the wrong-way driving likelihood H may also be calculated based only on the recognition information of the target included in the backward image IMG_R. Furthermore, in this case, the calculation of the wrong-way driving likelihood H does not need to be performed. For example, it is determined whether or not the recognition information of the target included in the backward image IMG_R includes recognition information of the guidance display surface of a traffic sign. Then, if it is determined that the recognition information of this guidance display surface is included in the recognition information of the target included in the backward image IMG_R, it may be determined that the vehicle M is driving in the wrong direction. [Explanation of symbols]

[0072] 10. Front camera 20 Rear camera 30 Information processing equipment 31 processors 32 memory 40 HMI units 100 Wrong-way driving detection device M vehicle L1~L6 lane TM1~TM5 Traffic signs DR_M Direction in which the vehicle is moving DR_L1~DR_L6 Direction indicated by lane IMG_F Front image IMG_R Rear image

Claims

1. acquiring front and rear images of a vehicle; Recognizing objects around the vehicle using the front and rear images; calculating a likelihood of reverse running based on the target recognition information; determining that the vehicle is traveling in the wrong direction when the likelihood of the wrong direction is equal to or greater than a threshold; Including, the target recognition information includes recognition information of a lane in which the vehicle is traveling; In the calculation of the wrong-way driving likelihood, the wrong-way driving likelihood is calculated using a sum of values of wrong-way driving factors that are set based on events recognized while the vehicle is traveling; the wrong-way driving factor includes a guide display surface of a traffic sign located to the left of the lane in which the vehicle recognized in the rear image is traveling, When the recognition information of the target using the rear image includes recognition information of the guidance display surface, a higher value is used as the recognition number of the guidance display surface increases, and the recognition information of the guidance display surface includes information on a time when at least one of the front image and the rear image including the recognition information of the guidance display surface was acquired; The calculation step of the wrong-way running likelihood further includes a step of determining, when it is determined that the recognition information of the guidance display surface is included in the recognition information of the target using the rear image, whether or not recognition information of a guidance display surface of a traffic sign other than the guidance display surface is obtained in the recognition information of the target using the forward image until the elapsed time counted from the time the rear image is acquired exceeds a predetermined time, When it is determined that recognition information of a guidance display surface of a traffic sign other than the guidance display surface is not obtained before the elapsed time exceeds a predetermined time, the guidance display surface recognized in the rear image is identified as the wrong-way driving factor. A method for determining whether a vehicle is running in the wrong direction.

2. a memory storing images of the front and rear of the vehicle; a processor; A wrong-way driving determination device comprising: The processor: Recognizing targets around the vehicle using the front and rear images; Calculating a likelihood of reverse running based on the recognition information of the target; If the likelihood of wrong-way driving is equal to or greater than a threshold, it is determined that the vehicle is driving in the wrong direction; the target recognition information includes recognition information of a lane in which the vehicle is traveling; the calculation of the wrong-way driving likelihood is performed using a sum of values of wrong-way driving factors that are set based on events recognized while the vehicle is traveling; the wrong-way driving factor includes a guide display surface of a traffic sign located to the left of the lane in which the vehicle recognized in the rear image is traveling, The processor, in calculating the reverse running likelihood, When the recognition information of the target using the rear image includes recognition information of the guidance display surface, a higher value is used as the recognition number of the guidance display surface increases, and the recognition information of the guidance display surface includes information on a time when at least one of the front image and the rear image including the recognition information of the guidance display surface was acquired; The processor further includes, in calculating the likelihood of reverse running, When it is determined that the recognition information of the guidance display surface is included in the recognition information of the target using the rear image, it is determined whether or not recognition information of a guidance display surface of a traffic sign other than the guidance display surface has been obtained in the recognition information of the target using the forward image until the elapsed time counted from the time the rear image was acquired exceeds a predetermined time, When it is determined that recognition information of a guidance display surface of a traffic sign other than the guidance display surface is not obtained before the elapsed time exceeds a predetermined time, the guidance display surface recognized in the rear image is identified as the wrong-way driving factor. A wrong-way driving determination device characterized by:

Images