Driving assistance device and driving assistance method

Abstract

To reduce erroneous recognition of a person in a far infrared video image captured by a far infrared camera.SOLUTION: A driving support device 10 comprises: a video image acquisition unit 11 for acquiring a video image in which the outside of a vehicle is shot by a far infrared camera; a person detection unit 12 for detecting a person, from the video image acquired by the video image acquisition unit 11, by referring to the person recognition model; another vehicle detection unit 13 for detecting the other vehicle, from the video image acquired by the video image acquisition unit 11, by referring to a vehicle recognition model and for determining whether or not the detected other vehicle travels; and a determination unit 14 by which, in a case where it is determined that the other vehicle travels, when a detection range of the other vehicle detected by the other vehicle detection unit 13 at least partially overlaps a detection range of the person detected by the person detection unit 12, the person detected while overlapping the detection range of the other vehicle, which is determined to travel, is determined as erroneous detection.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a driving assistance device and a driving assistance method. [Background technology]

[0002] There is a known technology that uses a far-infrared camera to receive far-infrared rays emitted from a photographed object, generate a thermal image, and detect surrounding objects. Patent Document 1 describes that a driving assistance device detects vehicles, pedestrians, passengers, etc. from a thermal image obtained by a far-infrared camera installed in a vehicle using a recognition model (dictionary) created by machine learning. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2020-27380 Summary of the Invention [Problem to be solved by the invention]

[0004] Thermal images taken by far-infrared cameras often have low resolution, and the heat distribution of objects other than people, such as vehicles, may be mistaken for people. In particular, if a person is mistakenly recognized as being present in the lane in the direction of travel of the vehicle, even though there is no person actually present, the driver of the vehicle will visually search for the person based on the mistaken recognition result, and will neglect to check the vehicle's surroundings.

[0005] The present disclosure has been made in consideration of the above points, and provides a driving assistance device and a driving assistance method that can reduce erroneous recognition of people in far-infrared images acquired by a far-infrared camera. [Means for solving the problem]

[0006] The driving assistance device according to this embodiment includes an image acquisition unit that acquires an image of the exterior of the vehicle using a far-infrared camera; a person detection unit that detects people from the image acquired by the image acquisition unit by referring to a person recognition model; an other vehicle detection unit that detects other vehicles from the image acquired by the image acquisition unit by referring to a vehicle recognition model and determines whether the detected other vehicles are traveling; a judgment unit that, when it is determined that the other vehicle is traveling and at least a part of the detection range of the other vehicle detected by the other vehicle detection unit overlaps with the detection range of the person detected by the person detection unit, judges that the person detected and overlapping the detection range of the other vehicle determined to be traveling is a false detection; and a display control unit that displays, on a display device visible to the driver of the vehicle, a detection result image in which people detected by the person detection unit in the image acquired by the image acquisition unit are highlighted as people, excluding people that the judgment unit has determined to be a false detection.

[0007] The driving assistance method according to this embodiment includes the steps of: acquiring video of the exterior of the vehicle using a far-infrared camera; detecting people from the acquired video by referring to a person recognition model; detecting other vehicles from the acquired video by referring to a vehicle recognition model and determining whether the detected other vehicles are traveling; determining, when it is determined that the other vehicles are traveling and the detection range of the detected other vehicles overlaps at least partially with the detection range of the detected person, that the person detected and overlapping the detection range of the other vehicle determined to be traveling is a false detection; and displaying, on a display device visible to the driver of the vehicle, a detection result video in which all detected people in the video, excluding those determined to be a false detection, are highlighted as people. [Effects of the Invention]

[0008] According to the present disclosure, it is possible to reduce erroneous recognition of people in far-infrared images acquired by a far-infrared camera. [Brief explanation of the drawings]

[0009] [Figure 1] 1 is a block diagram showing a schematic configuration of a driving assistance device according to a first embodiment. [Figure 2] FIG. 2 is a flow diagram illustrating a driving assistance method according to the first embodiment. [Figure 3] FIG. 4 is a flowchart illustrating another example of the driving assistance method according to the first embodiment. [Figure 4] FIG. 10 is a block diagram showing a schematic configuration of a driving assistance device according to a second embodiment. [Figure 5] FIG. 10 is a flowchart illustrating a driving assistance method according to a second embodiment. [Figure 6] FIG. 10 is a diagram showing an example in which a person is erroneously recognized in a far-infrared image captured by a far-infrared camera. DETAILED DESCRIPTION OF THE INVENTION

[0010] Hereinafter, a driving assistance device and a driving assistance method according to an embodiment of the present invention will be described with reference to the drawings. However, the present disclosure is not limited to the following embodiments. In addition, the following description and drawings have been simplified as appropriate for clarity of explanation.

[0011] Embodiment 1. Fig. 1 is a block diagram showing a schematic configuration of a driving assistance device according to embodiment 1. The driving assistance system 100 includes a driving assistance device 10, a far-infrared camera 20, a storage device 21, and a display device 22. The driving assistance system 100 is attached to a vehicle such as a passenger car or a motorcycle. The driving assistance system 100 detects objects (other vehicles, people, etc.) around the vehicle to which it is attached.

[0012] The far-infrared camera 20 is attached to, for example, the front grille of the vehicle or its surrounding area. The far-infrared camera 20 continuously captures images of a predetermined imaging range (image angle of view) outside the vehicle at a predetermined imaging rate, and generates a far-infrared image composed of multiple pieces of time-series image data. Specifically, the far-infrared camera 20 captures far-infrared images (thermal images) around the vehicle, particularly in the traveling direction of the vehicle, and outputs the images to the driving assistance device 10.

[0013] The far-infrared camera 20 receives far-infrared rays generated by heat generated by the movement of other vehicles and heat emitted by people, and generates far-infrared images. Note that heat generated by the movement of a vehicle refers to, for example, heat radiation from the hood or front grille at the front of the vehicle, heat radiation from the exhaust pipe (exhaust port) at the rear of the vehicle, heat radiation from tires and their surroundings, etc. At night, heat radiation from headlights and taillights is also included.

[0014] The driving assistance device 10 detects other vehicles, people, etc. around the vehicle equipped with the driving assistance system 100 from far-infrared images captured by the far-infrared camera 20, and performs processing to notify the driver of the vehicle by video or audio as necessary. As shown in Fig. 1, the driving assistance device 10 includes an image acquisition unit 11, a person detection unit 12, another vehicle detection unit 13, a determination unit 14, and a display control unit 15.

[0015] The storage device 21 stores data of various trained models such as a vehicle recognition model and a person recognition model. These trained models are created by machine learning of far-infrared images of vehicles, people such as pedestrians, etc. The driving assistance device 10 uses these trained models when detecting other vehicles, people, etc. from far-infrared images. Note that normal image recognition processing can be used to detect other vehicles, people, etc.

[0016] The image acquisition unit 11 acquires far-infrared images of the exterior of the vehicle from the far-infrared camera 20. The person detection unit 12 detects people from the far-infrared images using a person recognition model read from the storage device 21. As described above, the images from which the person detection unit 12 recognizes people are far-infrared images, and sensors with low resolution are often used, which may result in the person being recognized as a heat distribution of a vehicle or the like. Therefore, the results detected by the person detection unit 12 may include false detections, that is, results in which a heat distribution in an area where no person actually exists is falsely detected as a person.

[0017] FIG. 6 is a diagram showing an example in which a person is erroneously detected in a far-infrared image acquired by a far-infrared camera. In this example, it is assumed that there is actually no person in front of another vehicle C1 traveling ahead in the traveling direction of the host vehicle. As shown in FIG. 6, heat distribution due to the movement of the other vehicle resembles the shape of a person, and as a result of the erroneous recognition that a person is present, a detection frame indicating the erroneously detected person H1 is displayed in the acquired far-infrared image to the driver of the host vehicle. In this case, the driver of the host vehicle will actually visually confirm the person H1 based on the erroneous recognition result, and will neglect to check the surroundings of the vehicle during that time.

[0018] Therefore, in the driving assistance device 10 according to the first embodiment, the other vehicle detection unit 13 and the determination unit 14 perform the following processes. The other vehicle detection unit 13 detects other vehicles around the host vehicle from the far-infrared image using a vehicle recognition model read out from the storage device 21. The other vehicle detection unit 13 also determines whether the detected other vehicle is traveling.

[0019] For example, the other vehicle detection unit 13 determines whether or not another vehicle is traveling by referring to multiple frames of far-infrared images composed of multiple time-series image data. For example, it can determine that the other vehicle is traveling based on a change in the positional relationship between the detected other vehicle and surrounding objects (road surface, buildings, etc.) or a change in the relative positional relationship of the other vehicle based on the traveling speed of the host vehicle.

[0020] When it is determined that another vehicle is traveling, if the detection range of the person detected by the person detection unit 12 overlaps at least a part of the detection range of the other vehicle detected by the other vehicle detection unit 13, the determination unit 14 determines that the person detected overlapping the detection range of the traveling other vehicle is a false detection. This makes it possible to reduce false detection of people in far-infrared images.

[0021] When driving slowly or at a low speed in a traffic jam, a person may cross the road between other vehicles. For the driver of the vehicle, detecting such a person is a high priority. For this reason, it is desirable for the determination unit 14 to determine that another vehicle is traveling when the other vehicle is traveling at a predetermined speed or higher, for example, 10 km / h or higher. This makes it possible to detect a person jumping out from between other vehicles when the vehicle is traveling at a speed slower than the predetermined speed, and to reduce false detection of a person when the vehicle is traveling at a speed higher than the predetermined speed.

[0022] The display control unit 15 performs control to notify the driver of the vehicle of information about the detected person by video. For example, the display control unit 15 draws a frame (detection frame) indicating each person, among the people detected by the person detection unit 12, excluding people determined to be erroneously detected by the determination unit 14, in the far-infrared video acquired by the video acquisition unit 11.

[0023] The display control unit 15 outputs a detection result image, in which a detection frame is added to the acquired far-infrared image, to the display device 22. The display device 22 displays the detection result image transmitted from the display control unit 15 so that it can be visually confirmed by the driver of the vehicle. This allows the driver of the vehicle to recognize people who are actually present.

[0024] The driver of the vehicle may be notified of information about the detected person by voice. The voice may be output from the driving assistance device 10 or from a device external to the driving assistance device 10.

[0025] Next, the operation of the driving assistance device 10, i.e., the driving assistance method, will be described with reference to Fig. 2. Fig. 2 is a flow diagram illustrating the driving assistance method according to embodiment 1. When the driving assistance system 100 starts operating, the far-infrared camera 20 starts capturing far-infrared images in the traveling direction of the vehicle, and the image acquisition unit 11 acquires the far-infrared images (step S10).

[0026] Furthermore, when the far-infrared camera 20 starts capturing images, the person detection unit 12 starts detecting people from the far-infrared image with reference to the person recognition model (step S11). Similarly, the other vehicle detection unit 13 starts detecting other vehicles from the far-infrared image with reference to the vehicle recognition model (step S12). Next, the person detection unit 12 determines whether or not a person has been detected (step S13). If no person has been detected in step S13 (step S13, NO), the process proceeds to step S19.

[0027] If a person is detected in step S13 (step S13, YES), the determination unit 14 determines whether or not there is another vehicle that overlaps the area where the person is detected, based on the detection result of the other vehicle detection unit 13 (step S14). The determination performed in step S14 involves identifying the detection area of ​​the detected person in the far-infrared image and identifying the detection area of ​​the detected other vehicle in the far-infrared image, and determining whether at least a portion of the person's detection area overlaps the detection area of ​​the other vehicle. For example, if the overlapping ratio of the person's detection area with the detection area of ​​the other vehicle is relatively small, the person is detected based on a heat distribution that is different from the heat distribution of the detected other vehicle, and therefore the detected person is likely to be properly detected. If the overlapping ratio of the person's detection area with the detection area of ​​the other vehicle is relatively large, there is a high possibility that the heat distribution of the detected other vehicle has been erroneously detected as a person. Therefore, the determination performed in step S14 may be performed on the condition that a predetermined ratio or more, for example, 70% or more, of the detected person's detection area overlaps the detection area of ​​the other vehicle.

[0028] In step S14, if it is determined that there is no other vehicle overlapping the person detection range (step S14, NO), the process proceeds to step S17. If it is determined that there is another vehicle with at least a part of the person detection range overlapping the other vehicle, the process determines whether or not the other vehicle with the overlapping person detection range is traveling based on the detection result of the other vehicle detection unit 13 (step S15). As described above, it is preferable to determine whether or not the other vehicle is traveling when the other vehicle is traveling at, for example, 10 km / h or more. The process of step S15 may be included in the process of step S14. In that case, the process of step S14 determines whether or not there is another vehicle traveling that overlaps the person detection range.

[0029] The other vehicle detection unit 13 starts detecting other vehicles in step S12, but may detect other vehicles when a person is detected in step S13. The range for detecting other vehicles in the far-infrared image is not limited to the entire far-infrared image, but may be limited to the area around the person detected in step S13.

[0030] In step S15, if it is not determined that another vehicle is traveling (step S15, NO), that is, if the other vehicle is stopped or traveling at a speed less than a predetermined speed, the process proceeds to step S17. Also, in step S15, if it is determined that another vehicle is traveling (step S15, YES), the process proceeds to step S16.

[0031] In step S16, the determination unit 14 determines that the detected person is a false detection based on the determination results of steps S14 and S15. That is, the determination unit 14 determines that the person detected overlapping the detection range of another traveling vehicle is a false detection. In addition, in step S17, the determination unit 14 determines that the detected person is not a false detection based on the determination results of step S14 or step S15.

[0032] Next, based on the result of the determination in step S16 or step S17, the display control unit 15 displays a detection frame in the far-infrared image acquired by the image acquisition unit 11. Specifically, among the people detected by the person detection unit 12, the display control unit 15 does not display a detection frame for a person that the determination unit 14 determines to be a false detection, and displays a detection frame for a person that the determination unit 14 determines not to be a false detection (step S18).

[0033] For the person for whom a detection frame is displayed in step S18, a tracking process is performed for each frame of the far-infrared image, and the detection frame continues to be displayed until the person is no longer included in the far-infrared image or until there is no longer any need to display the detection frame.

[0034] Next, the driving assistance device 10 determines whether or not to end the process (step S19). Ending the process means that the conditions for ending image capture by the far-infrared camera 20 are met, or that the engine or power of the vehicle to which the driving assistance system 100 is attached is turned off. If it is determined that the process should be ended (step S19, YES), the process of FIG. 2 is ended. If it is determined that the process should not be ended (step S19, NO), the process proceeds to step S13.

[0035] Fig. 3 is a flow diagram illustrating another example of the driving assistance method according to embodiment 1. In Fig. 3, the same processes as those in Fig. 2 are denoted by the same reference numerals. Fig. 3 differs from Fig. 2 in that step S20 is provided between step S15 and step S16.

[0036] As shown in FIG. 3, if it is determined in step S15 that another vehicle is traveling (YES in step S15), the process proceeds to step S20. In step S20, it is determined whether the movements of the detected other vehicle and the person are linked. For example, the other vehicle detection unit 13 can determine that the other vehicle and the person are linked if the distance and / or direction of movement over time are equal, using the detection position (detection frame or ground position) of the other vehicle and the detection position (detection frame or ground position) of the person detected by the person detection unit 12. Alternatively, the other vehicle detection unit 13 determines that the movements of the other vehicle and the person are linked if the movement vectors over time of the coordinates of the far-infrared image indicating the range of the detected person and the coordinates of the far-infrared image indicating the range of the detected other vehicle are identical or almost identical.

[0037] If the determination unit 14 determines that the detected position of the other vehicle and the detected position of the person are moving in conjunction with each other (step S20, YES), the process proceeds to step S16. That is, the determination unit 14 determines that the person moving in conjunction with the movement of the other vehicle is an erroneous detection (step S16).

[0038] 3, when it is determined that another vehicle is traveling, and the detection range of the other vehicle detected by the other vehicle detection unit overlaps at least partly with the detection range of the person detected by the person detection unit, if it is determined that the detection position of the other vehicle and the detection position of the person are moving in conjunction with each other, the person detected and overlapping the detection range of the other vehicle determined to be traveling is determined to be a false detection. This makes it possible to more accurately detect people in far-infrared video.

[0039] If it is determined that the detected position of the other vehicle and the detected position of the person are not linked (step S20, NO), the process proceeds to step S17. In other words, it is determined that a person whose movement is not linked to the movement of the other vehicle is not an erroneous detection (step S17).

[0040] Embodiment 2. Fig. 4 is a block diagram showing a schematic configuration of a driving assistance device according to embodiment 2. In Fig. 4, the same components as those in Fig. 1 are given the same reference numerals, and descriptions thereof will be omitted as appropriate. In addition to the configuration in Fig. 1, embodiment 2 further includes a lane detection unit 16.

[0041] The lane detection unit 16 detects lanes from the far-infrared image. For example, the lane detection unit 16 performs edge detection processing on the far-infrared image, and detects lane markings by performing smoothing processing or Hough transform processing on the detected edge components, and detects lanes based on the positions of the detected lane markings. Note that various known detection processes can be used to detect lane markings and lanes, etc.

[0042] The lane detection unit 16 detects the lane in which the host vehicle is traveling based on the positions of the detected lane markings. The lane in which the host vehicle is traveling may be the lane in which the host vehicle is traveling, or, if there are multiple lanes traveling in the same direction, the lane in which the host vehicle is traveling may include the lane in which the host vehicle is traveling as well as the multiple lanes traveling in the same direction. When the determination unit 14 determines that another vehicle is traveling in the lane in which the host vehicle is traveling, defined by the lane markings detected by the lane detection unit 16, the determination unit 14 determines that a person detected as overlapping the detection range of the other vehicle is an erroneous detection. In other words, the determination unit 14 does not determine that a person detected as overlapping the detection range of a vehicle other than the other vehicle traveling in the lane is an erroneous detection.

[0043] Fig. 5 is a flow diagram illustrating a driving assistance method according to the second embodiment. In Fig. 5, the same processes as those in Figs. 2 and 3 are denoted by the same reference numerals. Fig. 5 differs from the example shown in Fig. 3 in that step S30 is provided between step S12 and step S13, and step S31 is provided between step S14 and step S20. Note that step S20 may be omitted in Fig. 5.

[0044] In step S30, the lane detection unit 16 detects lane markings from the far-infrared image and starts detecting the lane in which the host vehicle is traveling, which is defined by the detected lane markings. Then, in step S14, if another vehicle is traveling (YES), it is determined whether the lane in which the other vehicle is traveling is the lane in which the host vehicle is traveling, detected by the lane detection unit 16 (step S31).

[0045] In step S31, if it is determined that the lane on which the other vehicle is traveling is the lane on which the host vehicle is traveling, detected by the lane detection unit 16 (step S31, YES), the process proceeds to step S20. Alternatively, if step S20 is omitted, the process proceeds to step S16. In step S31, if it is determined that the lane on which the other vehicle is traveling is not the lane on which the host vehicle is traveling, detected by the lane detection unit 16 (step S31, NO), the process proceeds to step S17.

[0046] In this way, in the second embodiment, when it is determined that another vehicle is traveling, if the detection range of the other vehicle detected by the other vehicle detection unit overlaps at least partly with the detection range of the person detected by the person detection unit, and if it is determined that the other vehicle is traveling in a lane defined by lane markings detected by the lane detection unit 16 in which the vehicle is traveling, the person detected overlapping the detection range of the other vehicle determined to be traveling is determined to be a false detection. This makes it possible to prevent false detection of people in the traveling lane of the vehicle, which is something that the driver of the vehicle should be most careful about.

[0047] The configuration of the driving assistance device 10 is not limited to the above, and it is also possible to integrate a plurality of devices, for example, the driving assistance device 10 and the storage device 21, into a driving assistance device having a storage unit. It is also possible to integrate all of the components of the driving assistance system 100 into a driving assistance device having a far-infrared camera, a storage unit, and a display unit.

[0048] Note that some components of the driving assistance device 10 may be substituted by devices outside the driving assistance system 100 that are connected via communication means. For example, the person detection unit 12, the other vehicle detection unit 13, and the lane detection unit 16 may be substituted by a server outside the driving assistance system 100 that is connected via communication means.

[0049] In addition to a form in which a part or all of the driving assistance device is installed in the vehicle, a form in which the driving assistance device is installed in the vehicle so as to be portable or retrofittable may also be used. Note that, although the above description has been given assuming that the driving assistance system 100 is installed in an automobile, the driving assistance system 100 may also be installed in a vehicle other than an automobile.

[0050] In the above example, the person detection unit 12 and the other vehicle detection unit 13 use image recognition that uses a model created by machine learning images of vehicles, people, etc. However, the present invention is not limited to this. For example, other image recognition such as pattern matching using templates of other vehicles, people, etc. may be performed.

[0051] The present invention has been specifically described above based on the embodiments, but it goes without saying that the present invention is not limited to the above embodiments and can be modified in various ways without departing from the spirit of the invention. Two or more of the above embodiments can also be combined as appropriate.

[0052] Each functional block that performs various processes of the driving assistance device 10 shown in the drawings can be configured in hardware with a processor, memory, and other circuits. The above-described processes can also be realized by having a processor execute a program. Therefore, these functional blocks can be realized in various forms, such as hardware only, software only, or a combination thereof, and are not limited to any one of these.

[0053] The above-described program can be stored and supplied to a computer using various types of non-transitory computer-readable media. Non-transitory computer-readable media include various types of tangible storage media. Examples of non-transitory computer-readable media include semiconductor memory (e.g., mask ROM, PROM (Programmable ROM), EPROM (Erasable PROM), flash ROM, and RAM (Random Access Memory)). The program may also be supplied to a computer by various types of transitory computer-readable media. Examples of transitory computer-readable media include electrical signals, optical signals, and electromagnetic waves. The transitory computer-readable medium can supply the program to a computer via a wired communication path such as an electric wire or optical fiber, or via a wireless communication path. [Explanation of symbols]

[0054] 100 Driver Assistance Systems 10 Driving assistance devices 11 Video acquisition unit 12 Person detection unit 13 Other vehicle detection unit 14 Judgment Department 15 Display control unit 16 Lane detection unit 20 Far-infrared camera 21 Storage device 22 Display device H1 person C1, C2, C3 and other vehicles

Claims

1. an image acquisition unit that acquires an image of the outside of the vehicle captured by a far-infrared camera; a person detection unit that detects a person from the video acquired by the video acquisition unit by referring to a person recognition model; an other vehicle detection unit that detects other vehicles from the image acquired by the image acquisition unit by referring to a vehicle recognition model and determines whether the detected other vehicles are traveling; a determination unit that, when it is determined that the other vehicle is traveling and at least a part of the detection range of the other vehicle detected by the other vehicle detection unit overlaps with the detection range of the person detected by the person detection unit, determines that the person detected overlapping the detection range of the other vehicle determined to be traveling is an erroneous detection; Equipped with When a predetermined percentage or more of the detection range of the person overlaps with the detection range of the other vehicle, the determination unit determines that the person detected overlapping with the detection range of the other vehicle determined to be traveling is a false detection. Driving assistance device.

2. an image acquisition unit that acquires an image of the outside of the vehicle captured by a far-infrared camera; a person detection unit that detects a person from the video acquired by the video acquisition unit by referring to a person recognition model; an other vehicle detection unit that detects other vehicles from the image acquired by the image acquisition unit by referring to a vehicle recognition model and determines whether the detected other vehicles are traveling; a determination unit that, when it is determined that the other vehicle is traveling and at least a part of the detection range of the other vehicle detected by the other vehicle detection unit overlaps with the detection range of the person detected by the person detection unit, determines that the person detected overlapping the detection range of the other vehicle determined to be traveling is an erroneous detection; Equipped with When the determination unit determines that the detected position of the other vehicle and the detected position of the person are moving in conjunction with each other, the determination unit determines that the person detected overlapping the detection range of the other vehicle determined to be moving is an erroneous detection. Driving assistance device.

3. a lane detection unit that detects lane markings and a lane in which the vehicle is traveling based on the lane markings from the image acquired by the image acquisition unit, When it is determined that the other vehicle is traveling in a lane in which the vehicle is traveling, which is defined by lane markings detected by the lane detection unit, the determination unit determines that a person detected overlapping a detection range of the other vehicle is a false detection. The driving assistance device according to claim 1 or 2.

4. a display control unit that displays, on a display device visible to a driver of the vehicle, a detection result image in which people detected by the person detection unit, excluding people determined to be erroneously detected by the determination unit, are highlighted as people in the image acquired by the image acquisition unit; The driving assistance device according to claim 1 or 2.

5. The determination unit determines that another vehicle is traveling when the detected other vehicle is traveling at a predetermined speed or faster. The driving assistance device according to claim 1 .

6. an image acquisition step of acquiring an image of the exterior of the vehicle by a far-infrared camera; a person detection step of detecting a person from the acquired video by referring to a person recognition model; a vehicle detection step of detecting other vehicles from the acquired video by referring to a vehicle recognition model and determining whether the detected other vehicles are traveling; a determining step of determining that the person detected overlapping the detection range of the other vehicle determined to be traveling is an erroneous detection when at least a part of the detection range of the other vehicle and the detection range of the detected person overlap; a display control step of displaying, on a display device visible to a driver of the vehicle, a detection result image in which detected people in the image are highlighted as people, excluding people determined to be erroneously detected; Including, In the determination step, if a predetermined percentage or more of the detection range of the person overlaps with the detection range of the other vehicle, the person detected overlapping with the detection range of the other vehicle determined to be moving is determined to be a false detection. A driving assistance method executed by a driving assistance device.

7. an image acquisition step of acquiring an image of the exterior of the vehicle by a far-infrared camera; a person detection step of detecting a person from the acquired video by referring to a person recognition model; a vehicle detection step of detecting other vehicles from the acquired video by referring to a vehicle recognition model and determining whether the detected other vehicles are traveling; a determining step of determining that the person detected overlapping the detection range of the other vehicle determined to be traveling is an erroneous detection when at least a part of the detection range of the other vehicle and the detection range of the detected person overlap; a display control step of displaying, on a display device visible to a driver of the vehicle, a detection result image in which detected people in the image are highlighted as people, excluding people determined to be erroneously detected; Including, In the determination step, when it is determined that the detected position of the other vehicle and the detected position of the person are moving in conjunction with each other, the person detected to overlap with the detection range of the other vehicle determined to be moving is determined to be an erroneous detection. A driving assistance method executed by a driving assistance device.

Images