Driving assistance device, driving assistance method, and program

Abstract

To realize driving assistance that can prevent the host vehicle from approaching a traffic light in a situation where recognition of the traffic light by a camera is obstructed by a preceding vehicle, while reducing costs. 【Solution】The distance to the traffic light TL at the time when the traffic light TL is recognized is calculated, a virtual vehicle Cz whose vehicle height h Z is the upper limit of an assumed range is assumed, and the distance L to the virtual vehicle Cz at the time when the traffic light TL is recognized Z is calculated. At the time when the traffic light TL is recognized, if there is no difference between the measured value of the distance L to the actual vehicle C A A and the distance L to the virtual vehicle Cz Z , the vehicle height of the preceding vehicle is estimated as the vehicle height h Z , and then a recommended range of inter-vehicle distance is calculated. If there is a difference between the measured value of the distance L to the actual vehicle C A A and the distance L to the virtual vehicle Cz Z , the vehicle height h of the actual vehicle C A A is calculated, and then a recommended range of inter-vehicle distance is calculated.​​​

Description

[Technical Field]

[0001] This invention relates to a driver assistance device, a driver assistance method, and a program. [Background technology]

[0002] As a driver assistance device equipped with a function to calculate an appropriate following distance, there is a known device that calculates the following distance by considering the height of the preceding vehicle and the height of the traffic light, with the aim of preventing the vehicle from approaching a traffic light when the traffic light cannot be seen due to the height of the preceding vehicle (see, for example, Patent Documents 1 and 2). In the driver assistance devices described in Patent Documents 1 and 2, the height of the preceding vehicle is estimated by image recognition processing of a camera. [Prior art documents] [Patent Documents]

[0003] [Patent Document 1] Patent No. 5990941 [Patent Document 2] Patent No. 4852482 [Overview of the project] [Problems that the invention aims to solve]

[0004] When estimating the height of a preceding vehicle using camera image recognition processing, the accuracy depends heavily on the shooting environment, camera performance, and algorithmic design. Therefore, ensuring estimation accuracy requires using high-resolution or stereo cameras, employing auxiliary data to correct for viewpoint distortion and vehicle tilt, and incorporating correction processing that references standard vehicle height data for each vehicle type into the image recognition algorithm, all of which are costly.

[0005] In view of the above circumstances, the present invention aims to provide a driver assistance system that can prevent a vehicle from approaching a traffic light when the visibility of the traffic light by a camera is obstructed by a preceding vehicle, while keeping costs down. [Means for solving the problem]

[0006] The present invention provides a driver assistance device comprising: an image recognition unit that recognizes images from a camera that photographs the area in front of the vehicle; and a distance calculation unit that calculates a recommended range of distance between the vehicle and a preceding vehicle based on the images recognized by the image recognition unit, wherein the distance calculation unit calculates the distance to the traffic light at the time the traffic light is recognized by the image recognition unit, and the FOE (Focus of) of the camera image. Based on the relationship between the relative position of the Expansion and the traffic light, the installation height of the camera, and the height of the traffic light, the distance to the virtual vehicle at the time the traffic light is recognized by the video recognition unit is calculated based on the relationship between the distance to the traffic light at the time the traffic light is recognized by the video recognition unit, the height of the virtual vehicle, the installation height of the camera, and the height of the traffic light, and if there is no difference between the measured distance to the actual vehicle preceding the vehicle and the distance to the virtual vehicle at the time the traffic light is recognized by the video recognition unit, the recommended range of the inter-vehicle distance is calculated based on the relationship between the measured distance to the actual vehicle preceding the vehicle and the distance to the virtual vehicle at the time the traffic light is recognized by the video recognition unit. The system calculates the distance to the traffic light based on the relationship between the distance to the traffic light, the height of the virtual vehicle, the installation height of the camera, and the height of the traffic light. If there is a difference between the measured value and the distance to the virtual vehicle at the time the traffic light is recognized by the video recognition unit, the vehicle height of the actual vehicle is calculated based on the relationship between the distance to the traffic light at the time the traffic light is recognized by the video recognition unit, the measured value at the time the traffic light is recognized by the video recognition unit, the installation height of the camera, and the height of the traffic light. The system also calculates the recommended range of the inter-vehicle distance based on the relationship between the distance to the traffic light at the time the traffic light is recognized by the video recognition unit, the vehicle height of the actual vehicle, the installation height of the camera, and the height of the traffic light.

[0007] The present invention provides a driving assistance method that uses a control device to perform an image recognition procedure to recognize an image from a camera that takes pictures of the front of the vehicle, and a distance calculation procedure to calculate a recommended range of distance between the vehicle and a preceding vehicle based on the image recognized by the image recognition procedure, wherein the distance calculation procedure uses the distance to the traffic light at the time the traffic light is recognized by the image recognition procedure to calculate the FOE (Focus of) in the camera image. Based on the relationship between the relative position of the Expansion and the traffic light, the installation height of the camera, and the height of the traffic light, the distance to the virtual vehicle at the time the traffic light is recognized by the video recognition procedure is calculated based on the relationship between the distance to the traffic light at the time the traffic light is recognized by the video recognition procedure, the height of the virtual vehicle, the installation height of the camera, and the height of the traffic light, and if there is no difference between the measured distance to the actual vehicle preceding the vehicle and the distance to the virtual vehicle at the time the traffic light is recognized by the video recognition procedure, the recommended range of the inter-vehicle distance is calculated based on the relationship between the measured distance to the actual vehicle preceding the vehicle and the distance to the virtual vehicle at the time the traffic light is recognized by the video recognition procedure. The distance to the traffic light is calculated based on the relationship between the distance to the traffic light, the height of the virtual vehicle, the installation height of the camera, and the height of the traffic light. If there is a difference between the measured value and the distance to the virtual vehicle at the time the traffic light is recognized by the video recognition procedure, the height of the actual vehicle is calculated based on the relationship between the distance to the traffic light at the time the traffic light is recognized by the video recognition procedure, the measured value at the time the traffic light is recognized by the video recognition procedure, the installation height of the camera, and the height of the traffic light. The recommended range for the inter-vehicle distance is calculated based on the relationship between the distance to the traffic light at the time the traffic light is recognized by the video recognition procedure, the height of the actual vehicle, the installation height of the camera, and the height of the traffic light.

[0008] The present invention is a program that causes a control device to execute an image recognition process that recognizes the image from a camera that takes pictures of the front of the vehicle, and a distance calculation process that calculates a recommended range of distance between the vehicle and a preceding vehicle based on the image recognized by the image recognition process, wherein the distance calculation process uses the distance to the traffic light at the time the traffic light is recognized by the image recognition process as the FOE (Focus of) of the camera image. Based on the relationship between the relative position of the Expansion and the traffic light, the installation height of the camera, and the height of the traffic light, the distance to the virtual vehicle at the time the traffic light is recognized by the video recognition process is calculated based on the relationship between the distance to the traffic light at the time the traffic light is recognized by the video recognition process, the height of the virtual vehicle, the installation height of the camera, and the height of the traffic light, and if there is no difference between the measured distance to the actual vehicle preceding the vehicle and the distance to the virtual vehicle at the time the traffic light is recognized by the video recognition process, the recommended range of the inter-vehicle distance is calculated based on the relationship between the measured distance to the actual vehicle preceding the vehicle and the distance to the virtual vehicle at the time the traffic light is recognized by the video recognition process. The distance to the traffic light is calculated based on the relationship between the distance to the traffic light, the height of the virtual vehicle, the installation height of the camera, and the height of the traffic light. If there is a difference between the measured value and the distance to the virtual vehicle at the time the traffic light is recognized by the video recognition process, the height of the actual vehicle is calculated based on the relationship between the distance to the traffic light at the time the traffic light is recognized by the video recognition process, the measured value at the time the traffic light is recognized by the video recognition process, the installation height of the camera, and the height of the traffic light. The recommended range for the inter-vehicle distance is calculated based on the relationship between the distance to the traffic light at the time the traffic light is recognized by the video recognition process, the height of the actual vehicle, the installation height of the camera, and the height of the traffic light. [Effects of the Invention]

[0009] According to the present invention, a driver assistance system can be implemented at a reduced cost that can prevent the vehicle from approaching a traffic light when the camera's view of the traffic light is obstructed by a preceding vehicle. [Brief explanation of the drawing]

[0010] [Figure 1] Figure 1 is a block diagram showing the configuration of a driver assistance system according to one embodiment of the present invention. [Figure 2] Figure 2 shows the relationship between the distance from the vehicle to the traffic light, the height of the traffic light, the installation height of the front camera, and the angle between the front camera's FOE and the front camera's line-of-view vector to the traffic light. [Figure 3] Figure 3 shows the relationship between the distance between your vehicle and the vehicle ahead and the visible distance of the traffic signal as seen by the front camera. [Figure 4] Figure 4 is a diagram illustrating how to calculate the recommended following distance. [Figure 5] Figure 5 is a diagram illustrating how to calculate the recommended following distance. [Figure 6] Figure 6 is a diagram illustrating how to calculate the recommended following distance. [Figure 7] Figure 7 is a flowchart illustrating the process for estimating the height of the preceding vehicle. [Figure 8] Figure 8 is a flowchart illustrating the process of providing advice on the distance between vehicles. [Modes for carrying out the invention]

[0011] The present invention will be described below in accordance with preferred embodiments. However, the present invention is not limited to the embodiments shown below, and can be modified as appropriate without departing from the spirit of the invention. Furthermore, in the embodiments shown below, some components are not illustrated or described; however, details of the omitted technologies can be appropriately applied from publicly known or well-known technologies, to the extent that they do not contradict the content described below.

[0012] Figure 1 is a block diagram showing the configuration of a driver assistance system 1 according to one embodiment of the present invention. The driver assistance system 1 shown in this figure includes an in-vehicle unit 10 that is installed in the vehicle C0 (see Figure 2) that is the target of the driver assistance. The in-vehicle unit 10 is an in-vehicle device that integrates a digital tachograph and a drive recorder, and includes a front camera 2.

[0013] In addition to general functions of a drive recorder and digital tachograph, such as video recording and operation recording functions, the in-vehicle unit 10 is equipped with a video recognition function and a driver assistance function. The video recognition function includes a function to recognize the preceding vehicle C1 (see Figure 3, etc.) and the traffic light TL (see Figure 2, etc.) from the video of the front camera 2. The driver assistance function includes a function to calculate the distance from the vehicle C0 to the traffic light TL, a function to calculate the recommended range of the distance between the vehicle C0 and the preceding vehicle C1, and a function to notify the driver of the vehicle C0 that the distance between the vehicle C0 and the preceding vehicle C1 should be optimized.

[0014] The in-vehicle unit 10 includes a CPU (Central Processing Unit) 11, memory 12, interfaces (hereinafter referred to as I / F) 13A, 13B, 13C, a card interface (hereinafter referred to as card I / F) 14, a communication module 15, a speaker 16, and a GPS (Global Positioning System) module 17. Video signals output from the front camera 2 and an optional camera (hereinafter referred to as optional camera) 3 are input to I / F 13A. Vehicle signals such as vehicle speed, engine speed, and operation of turn signals and reverse gear are input to I / F 13B. Signals output from external devices 4 such as a navigation system, switch unit, and ETC in-vehicle unit are input to I / F 13C. A recording medium 141 such as an SD card is installed in the card I / F 14.

[0015] A communication antenna 151 is connected to the communication module 15. The communication antenna 151 communicates with a management server (not shown) via a wireless communication network. The speaker 16 outputs voice notifications such as notifications to increase the distance between vehicles, various guidance and warnings. A GPS antenna 171 is connected to the GPS module 17 and receives GPS signals transmitted from GPS satellites (not shown). These GPS signals include GPS location information and GPS time information.

[0016] The CPU 11 controls various parts of the in-vehicle device 10, and in particular executes video recognition, driving assistance, video recording, and operation recording functions. The CPU 11 comprises a video recognition unit 11A, a driving assistance unit 11B, a video recording unit 11C, and an operation recording unit 11D.

[0017] The video recognition unit 11A recognizes traffic lights TL, preceding vehicles C1, pedestrians, lanes, signs, etc., from the video captured by the front camera 2. The driving support unit 11B determines sudden acceleration, sudden deceleration, sudden steering, speeding, etc., and outputs notification information, and calculates the recommended range for following distance, determines whether the current following distance is appropriate, and outputs notification information. The driving support unit 11B includes a following distance calculation unit 11B-1 and a notification unit 11B-2.

[0018] The inter-vehicle distance calculation unit 11B-1 calculates the recommended range of inter-vehicle distance between the vehicle C0 and the preceding vehicle C1. If the current inter-vehicle distance between the vehicle C0 and the preceding vehicle C1 is shorter than the recommended range of inter-vehicle distance calculated by the inter-vehicle distance calculation unit 11B-1, the notification unit 11B-2 outputs notification information to the speaker 16 or the like to notify the vehicle that it is necessary to increase the inter-vehicle distance.

[0019] The video recording unit 11C records video captured by the front camera 2 or optional camera 3 onto the recording medium 141 at all times or when an event occurs. The operation recording unit 11D records driving data such as vehicle speed and deceleration input to the I / F 13B, as well as events related to the operation of the vehicle C0, such as sudden acceleration, sudden deceleration, and sudden steering, onto the recording medium 141.

[0020] Memory 12 is a non-volatile memory that stores programs executed by the CPU 11. Examples of programs executed by the CPU 11 include programs for performing the aforementioned video recognition function, driving assistance function, video recording function, and operation recording function. Memory 12 also records the distance to the preceding vehicle C1 and traffic light TL recognized from the video of the front camera 2 by the video recognition unit 11A.

[0021] Front camera 2 is connected to the in-vehicle unit 10 and captures the area in front of the vehicle C0, outputting the video signal to I / F 13A. Optional camera 3 is connected to the in-vehicle unit 10 and captures the areas to the left and right of the vehicle C0, as well as the interior of the vehicle, outputting the video signal to I / F 13A.

[0022] Alternatively, a ranging device such as LiDAR (Light Detection and Ranging) or millimeter-wave radar may be connected to the on-board unit 10 to measure the distance from the vehicle C0 to the preceding vehicle C1, or the distance from the vehicle C0 to the traffic light TL.

[0023] Figure 2 shows the distance L from the vehicle C0 to the traffic light TL. S And the height of the traffic light TL h S And the installation height h of the front camera 2 C This figure shows the relationship between the angle α between the FOE of front camera 2 and the line-of-sight vector of front camera 2 to the traffic light TL.

[0024] Distance L from vehicle C0 to traffic light TL S It is calculated using the following formula (1). Note that distance L S This is the distance from front camera 2 to traffic light TL.

number

[0025] Figure 3 shows the distance L between vehicle C0 and the preceding vehicle C1. Vis a diagram showing the relationship with the recognizable distance of traffic light TL by front camera 2. In this diagram, the visible distance of front camera 2 is 8 to 50 m, host vehicle C0 is an ordinary passenger car with a lower vehicle height than a truck, and the vehicle height h of preceding vehicle C1 V is the legally upper limit (3.8 m). Note that inter-vehicle distance L V is the distance from front camera 2 to the rear end of preceding vehicle C1.

[0026] On the left side of Figure 3, there is shown a scenario where host vehicle C0 approaches traffic light TL in a state where traffic light TL is never visually recognized by front camera 2. In this scenario, traffic light TL 8 m ahead from front camera 2 is hidden by preceding vehicle C1 and cannot be visually recognized by front camera 2. The inter-vehicle distance L at which this scenario occurs V is calculated by the following equation (2). h V =3.8 m, h C =1.3 m, L S =8 m, h S =5 m, the inter-vehicle distance L V is 5.4 m. {Equation}[

[0027] On the right side of Figure 3, there is shown a scenario where traffic light TL 50 m ahead from front camera 2 is hidden by preceding vehicle C1 and cannot be visually recognized by front camera 2. The inter-vehicle distance L at which this scenario occurs V is calculated by the above equation (2). Note that h V =3.8 m, h C =1.3 m, L S =50 m, h S =5 m, the inter-vehicle distance L V is 33.8 m.

[0028] Figures 4 to 6 are diagrams for explaining the method of calculating the recommended range of inter-vehicle distance L V . As shown in Figures 4 to 6, in the present embodiment, a virtual vehicle Cz whose vehicle height h Z is the legally specified upper limit, and a preceding vehicle whose inter-vehicle distance is actually measured (hereinafter referred to as actual vehicle) C ADepending on the relative position, the height h of the preceding vehicle C1 V Estimate the vehicle height h V Based on the estimated value, the following distance L V The recommended range is calculated. Note that this is for actual vehicle C. A Distance L A This is measured by a ranging device such as LiDAR or millimeter-wave radar, or by image recognition processing from the forward camera 2.

[0029] Figures 4 to 6 show the situation at the point when the front camera 2 begins recognizing the traffic light TL. Here, the distance L from the vehicle C0 to the traffic light TL is... S If the recognition of the traffic light TL by the front camera 2 begins when the distance is 50m or less, it means that the traffic light TL was hidden by the preceding vehicle C1 in the field of view of the front camera 2 up to that point.

[0030] Vehicle C0 is traveling ahead of a vehicle with height h Z Assuming a virtual vehicle Cz with a legal upper limit, the distance L to the virtual vehicle Cz is... Z This can be calculated using the following equation (3): The distance L to the traffic light TL at the time when the front camera 2 starts recognizing the traffic light TL. S 40m, traffic light TL height h S The installation height of the front camera 2 is 5m, h C The height of the virtual vehicle Cz is 1.3m. Z If we set it to 3.8m, then distance L Z It will be 27m.

number

[0031] Figure 4 shows the distance L to the virtual vehicle Cz. Z and actual vehicle C A Distance L A A situation where the difference between them is 0 is shown. In this situation, the virtual vehicle Cz and the real vehicle C A This matches the preceding vehicle C1, which was obscuring the signal light TL. Actual vehicle C A Distance L A This has been measured by the video recognition unit 11A, etc.

[0032] Therefore, the height of the virtual vehicle Cz h Z However, the height of the preceding vehicle C1, which was obscuring the traffic light TL, h V This is the actual vehicle C. A Distance L A (Measured value) and distance L to virtual vehicle Cz Z The following distance L V This becomes the threshold for the recommended range. Therefore, in this situation, the vehicle height h of the preceding vehicle C1 V It is estimated to be 3.8m, following distance L V The recommended range is calculated to be less than 27m.

[0033] Figure 5 shows the distance L to the virtual vehicle Cz. Z and actual vehicle C A Distance L A The difference is shown to be less than the overall length of a typical passenger car or light truck (for example, less than 6m). In this situation, actual vehicle C A This matches the preceding vehicle C1, which was obscuring the signal light TL. Therefore, the actual vehicle C A Distance L A Actual vehicle C calculated based on A Vehicle height h A However, the height of the preceding vehicle C1, which was obscuring the traffic light TL, h V It is estimated that the actual vehicle C A Distance L A However, the distance between cars is L V This will be the recommended threshold range.

[0034] Actual vehicle C A Vehicle height h A This can be calculated using the following equation (4). Actual vehicle C A Distance L A The distance is 24m, and the distance to the traffic light TL is L. S If the distance is 40m, then the actual vehicle C A Vehicle height h A It is estimated to be 3.5m, following distance L V The recommended range is calculated to be less than 24m (h C = 1.3m, h S (=5m).

number

[0035] FIG. 6 shows the distance L to virtual vehicle Cz Z and the distance L to real vehicle C A A is a situation where the difference between them is equal to or greater than the total length of a typical ordinary passenger car or small truck (for example, 6 m or more). In this situation, real vehicle C A may have a preceding vehicle C1 that hides traffic signal TL between itself and virtual vehicle Cz. Therefore, real vehicle C A distance L to A the vehicle height h of real vehicle C calculated based on A A does not necessarily match the vehicle height h of preceding vehicle C1 that was hiding traffic signal TL V . However, since there is no possibility that a preceding vehicle C1 with vehicle height h V equal to or greater than vehicle height h A exists between host vehicle C0 and traffic signal TL, the vehicle height h of preceding vehicle C1 V is estimated as vehicle height h A , and the inter-vehicle distance L calculated based on the estimated value V does not cause any problem even if it is used as the threshold value for the recommended range of inter-vehicle distance L V . Real vehicle C A distance L to A is 17 m, and the distance L to traffic signal TL S is 40 m, the vehicle height h of preceding vehicle C1 V is estimated to be 2.9 m, and the recommended range of inter-vehicle distance L V is calculated to be less than 14 m.

[0036] FIG. 7 is a flowchart for explaining the process of estimating the vehicle height h of preceding vehicle C1 V . The process shown in this flowchart relates to the distance L to traffic signal TL S ​​is started when the value falls below a predetermined value. The predetermined value is equal to or greater than the visible distance of the front camera 2 (for example, 50 m), is 1 to 10 times the visible distance of the front camera 2, and is preferably 1 to several times the visible distance of the front camera 2. The CPU 11 identifies the position of the traffic light TL by collating GPS position information with map information, and determines the distance L between the coordinates of the current location of the host vehicle C0 and the coordinates of the traffic light TL S is calculated.

[0037] In the processing shown in the flowchart of Fig.7, first, the image recognition unit 11A performs a process of recognizing the traffic light TL from the image of the front camera 2, and a process of recognizing the actual vehicle C from the image of the front camera 2 A and the actual vehicle C A The distance L to A is started (step S1). The process of measuring the distance to the actual vehicle C A The distance L to A may be measured by a distance measuring device such as LiDAR or a millimeter wave radar.

[0038] Next, the driving support unit 11B assigns an ID (identifier) to the actual vehicle C recognized by the image recognition unit 11A A and records it in the memory 12 (step S2). Next, the driving support unit 11B determines whether or not the vehicle speed of the host vehicle C0 is greater than 0 (step S3). When the vehicle speed of the host vehicle C0 is 0 (NO in step S3), the driving support unit 11B repeats the processing of step S3. In contrast, when the vehicle speed of the host vehicle C0 is greater than 0 (YES in step S3), the driving support unit 11B determines whether or not the traffic light TL has been recognized from the image of the front camera 2 by the image recognition unit 11A (step S4).

[0039] When the traffic light TL is not recognized from the image of the front camera 2 by the image recognition unit 11A (NO in step S4), the process returns to step S3. In contrast, when the traffic light TL is recognized from the image of the front camera 2 by the image recognition unit 11A (YES in step S4), the driving support unit 11B determines whether or not the traffic light TL has not been recognized for a predetermined time (for example, 10 seconds) before the time point when the traffic light TL was recognized (step S5).

[0040] If the traffic light TL was recognized during a predetermined time period prior to the time the traffic light TL was recognized (NO in step S5), the process returns to step S3. Conversely, if the traffic light TL was not recognized during a predetermined time period prior to the time the traffic light TL was recognized (YES in step S5), the inter-vehicle distance calculation unit 11B-1 calculates the distance L from the image of the front camera 2 to the traffic light TL at the time the traffic light TL was recognized. S This is calculated using the above formula (1) (step S6).

[0041] Next, the inter-vehicle distance calculation unit 11B-1 calculates the distance L to the virtual vehicle Cz at the time the traffic light TL is recognized from the image of the front camera 2. Z The value is calculated using the above formula (3) (step S7). The vehicle height h of the virtual vehicle Cz. Z The height of the traffic light TL is 3.8m, h s The height is 5m, and the installation height of the front camera 2 is h C It is 1.3m. Distance L to traffic light TL S This value represents the point at which the traffic light TL was recognized from the image of the front camera 2.

[0042] The inter-vehicle distance calculation unit 11B-1 calculates the distance L to the virtual vehicle Cz. Z and actual vehicle C A Distance L A It is determined whether the difference with is 0 or not (step S8). The inter-vehicle distance calculation unit 11B-1 calculates the distance L to the virtual vehicle Cz. Z and actual vehicle C A Distance L A If the difference is 0 (YES in step S8), the vehicle height h of the preceding vehicle C1 V The height of the virtual vehicle Cz h Z Set to (step S9). In step S9, virtual vehicle Cz and real vehicle C A Based on the determination that this matches the preceding vehicle C1 which was obscuring the traffic light TL, the vehicle height h of the preceding vehicle C1 was determined V The height of the virtual vehicle Cz h Z This is confirmed. The process moves from step S9 to step S21 in the flowchart of Figure 8.

[0043] The inter-vehicle distance calculation unit 11B-1 calculates the distance L to the virtual vehicle Cz. Z and actual vehicle C A Distance L A If the difference is not 0 (NO in step S8), the distance L to the virtual vehicle Cz is Z and actual vehicle C A Distance L A The system determines whether the difference is less than the total length of a regular passenger car or a light truck (less than 6m in this embodiment) (step S10). The inter-vehicle distance calculation unit 11B-1 calculates the distance L to the virtual vehicle Cz. Z and actual vehicle C A Distance L A If the difference is less than the overall length of a regular passenger car or light truck (YES in step S10), then actual vehicle C A Vehicle height h A The value is calculated using the above formula (4) and the vehicle height h of the preceding vehicle C1 is calculated. V Set to (step S11). In step S11, actual vehicle C A Based on the determination that this matches the preceding vehicle C1 which was obscuring the traffic light TL, the vehicle height h of the preceding vehicle C1 was determined V Actual vehicle C A Vehicle height h A This is confirmed. The process moves from step S11 to step S21 in the flowchart of Figure 8.

[0044] The inter-vehicle distance calculation unit 11B-1 calculates the distance L to the virtual vehicle Cz. Z and actual vehicle C A Distance L A If the difference is not less than the total length of a regular passenger car or light truck (NO in step S10), the distance L to the virtual vehicle Cz is Z and actual vehicle C A Distance L A Step S12 determines whether the difference is greater than or equal to the overall length of a regular passenger car or a light truck (6m or more in this embodiment).

[0045] The inter-vehicle distance calculation unit 11B-1 calculates the distance L to the virtual vehicle Cz. Z and actual vehicle C A Distance L AIf the difference is greater than or equal to the overall length of a regular passenger car or light truck (YES in step S12), then actual vehicle C A Vehicle height h A The value is calculated using the above formula (4) and the calculated value is the actual vehicle C A Vehicle height h A Set it temporarily (Step S13). In Step S13, actual vehicle C A Based on the judgment that there is a possibility that a preceding vehicle C1 that was hiding the traffic light TL was present between and the virtual vehicle Cz, the vehicle height h of the preceding vehicle C1 V Actual vehicle C A Vehicle height h A This is set as a temporary value. From step S13, proceed to step S21 in the flowchart of Figure 8.

[0046] The inter-vehicle distance calculation unit 11B-1 calculates the distance L to the virtual vehicle Cz. Z and actual vehicle C A Distance L A If the difference is not greater than the overall length of a regular passenger car or light truck (NO in step S12), the vehicle height h of the preceding vehicle C1 V The process of estimating is terminated (step S14). In step S14, the process is terminated based on the determination that the height of the signal light TL is not relevant, or that the preceding vehicle C1 is a vehicle with a special shape. The process returns from step S14 to step S1.

[0047] Figure 8 shows the following distance L. V This flowchart illustrates the process of providing advice. The process shown in this flowchart begins after steps S9, S11, and S13 in Figure 7.

[0048] First, the inter-vehicle distance calculation unit 11B-1 calculates the vehicle height h of the preceding vehicle C1, which was set in steps S9, S11, and S13. V Based on the above equation (2), the following distance L V The recommended range is calculated (step S21). Here, the distance L S This is the distance to the traffic light TL when the traffic light TL is hidden by a preceding vehicle (identical vehicle Cz) whose vehicle height is within the legal limit, and in this embodiment it is 40m. In addition, in this embodiment the installation height h of the front camera 2.C The height of the traffic light TL is 1.3m, h S It is 5m.

[0049] Under the above conditions, the vehicle height h of the preceding vehicle C1 V The height of the virtual vehicle Cz h Z If set to (step S9), the vehicle height h V The distance between vehicles becomes 3.8m, L V The recommended threshold range is 27m. Also, the vehicle height of the preceding vehicle C1 is h V Actual vehicle C A Vehicle height h A If set to (step S11), the vehicle height h V The distance between vehicles becomes 3.5m, L V The recommended threshold range is 24m. Furthermore, the vehicle height of the preceding vehicle C1 is h V Actual vehicle C A Vehicle height h A If it is temporarily set to (Step S13), the vehicle height h V The distance between vehicles becomes 2.9m, L V The recommended threshold range is 14m.

[0050] Next, the notification unit 11B-2 indicates the following distance L V The recommended threshold range is the current actual vehicle C A The distance between vehicles up to L V It is determined whether the distance L is greater than or less (step S22). The notification unit 11B-2 determines the distance L between vehicles. V The recommended threshold range is the current actual vehicle C A The distance between vehicles up to L V If the value is greater (YES in step S22), the process executes to output notification information instructing the vehicle to increase its following distance (step S23). One example of the process in step S23 is the process of notifying the vehicle by voice via speaker 16.

[0051] The driver assistance unit 11B controls the following distance L V The recommended threshold range is the current actual vehicle C A distance L to the vehicle ahead V If it is smaller (NO in step S22), or after performing the process in step S23, the actual vehicle C AIt is determined whether the ID has changed (step S24). The driver assistance unit 11B determines whether the ID of the actual vehicle C A While the ID of is not changing (NO in step S24), the process from steps S22 to S24 is repeated, and the actual vehicle C A If the ID changes (YES in step S24), proceed to step S1 of the flowchart in Figure 7.

[0052] As described above, in the driver assistance system 1 according to this embodiment, a video recognition unit 11A and a vehicle distance calculation unit 11B-1 are provided in the CPU 11. The video recognition unit 11A recognizes the video from the front camera 2 that takes pictures of the area in front of the vehicle C0. The vehicle distance calculation unit 11B-1 calculates a recommended range of vehicle distance between the vehicle C0 and the preceding vehicle C1 based on the video recognized by the video recognition unit 11A.

[0053] The inter-vehicle distance calculation unit 11B-1 performs the following processes (1) to (4). (1) The distance L to the traffic light TL at the time the traffic light TL is recognized by the video recognition unit 11A S This refers to the relative position of FOE and traffic light TL in the image from front camera 2, and the installation height h of front camera 2. C And the height of the traffic light TL h S It is calculated based on the relationship with [the other factor]. (2) Vehicle height h Z Assuming a virtual vehicle Cz is the upper limit of the assumed range, the distance L to the virtual vehicle Cz at the time the traffic light TL is recognized by the video recognition unit 11A is assumed to be the upper limit of the assumed range. Z The distance L to the traffic light TL at the time the traffic light TL is recognized by the video recognition unit 11A. S And the ride height of the virtual vehicle Cz h Z And the installation height h of the front camera 2 C And the height of the traffic light TL h S It is calculated based on the relationship with [the other factor]. (3) When the signal light TL is recognized by the video recognition unit 11A, the actual vehicle C preceding the own vehicle C0 A Distance L A Actual measured value and distance L to virtual vehicle Cz Z If there is no difference between the two, the following distance LV The recommended range is the distance L to the traffic light TL at the time the traffic light TL is recognized by the video recognition unit 11A. S And the ride height of the virtual vehicle Cz h Z And the installation height h of the front camera 2 C And the height of the traffic light TL h S It is calculated based on the relationship with [the other factor]. (4) When the signal light TL is recognized by the video recognition unit 11A, the actual vehicle C A Distance L A Actual measured value and distance L to virtual vehicle Cz Z If there is a difference between the two, then the actual vehicle C A Vehicle height h A The distance L to the traffic light TL at the time the traffic light TL is recognized by the video recognition unit 11A. S And the actual vehicle C at the time the traffic light TL is recognized by the video recognition unit 11A. A Distance L A The measured value and the installation height h of the front camera 2 C And the height of the traffic light TL h S Calculated based on the relationship, following distance L V The recommended range is the distance L to the traffic light TL at the time the traffic light TL is recognized by the video recognition unit 11A. S And, actual vehicle C A Vehicle height h A And the installation height h of the front camera 2 C And the height of the traffic light TL h S It is calculated based on the relationship with [the other factor].

[0054] As a result, the vehicle height h of the preceding vehicle C1 V Without requiring processing to estimate the following distance L by recognizing the image from the front camera 2, V The recommended range can be calculated. Therefore, a driver assistance system can be implemented at a reduced cost that avoids the vehicle C0 approaching the traffic light TL when the recognition of the traffic light TL by the front camera 2 is obstructed by the preceding vehicle C1.

[0055] Furthermore, in the driving support system 1 according to this embodiment, the notification unit 11B-2 receives the calculated inter-vehicle distance L VIt outputs notification information to inform the driver of the recommended range for following a vehicle. V It can recognize it.

[0056] In particular, the notification unit 11B-2 detects the actual vehicle C from the moment the signal light TL is recognized by the video recognition unit 11A. A Distance L A The actual measured value is the calculated following distance L V If the following distance becomes smaller than the recommended range, V It outputs notification information to inform the driver to widen the following distance L. V We can recognize the need to expand it.

[0057] Furthermore, in the driving support system 1 according to this embodiment, the vehicle height h of the virtual vehicle Cz is Z However, it is set to the legal upper limit (3.8m). As a result, the recognition of the traffic light TL by the front camera 2 is affected by the vehicle height h V This enables driver assistance that can prevent the vehicle C0 from approaching the traffic light TL when it is obstructed by a preceding vehicle C1 below the legal upper limit.

[0058] Furthermore, in the driving support system 1 according to this embodiment, the distance L between the vehicle C0 and the traffic light TL is S Only after the distance becomes less than or equal to a predetermined value that is at least one times the visible distance of the front camera 2, the inter-vehicle distance calculation unit 11B-1 calculates the inter-vehicle distance L between the vehicle C0 and the preceding vehicle C1 based on the video recognized by the video recognition unit 11A. V The recommended range is calculated. This allows for reduced CPU usage, lower power consumption, and reduced load on CPU11.

[0059] Although the present invention has been described above based on the embodiments described above, the present invention is not limited to the embodiments described above, and modifications may be made, or publicly known or well-known technologies may be combined as appropriate, without departing from the spirit of the present invention.

[0060] For example, in the above embodiment, the in-vehicle device 10 has a distance L between vehicles.V The recommended range was calculated. However, the following distance L V The calculation of the recommended range may be performed on a management server that can communicate with the in-vehicle unit 10. [Explanation of Symbols]

[0061] 1: Driver assistance system (driver assistance device) 2: Front camera (camera) 11: CPU (Control Unit) 11A: Image recognition unit 11B-1: Inter-vehicle distance calculation section 11B-2:Notification section C0: Own vehicle C1: Leading vehicle C A Actual vehicle Cz: Virtual Vehicle h A : Vehicle height (actual vehicle height) h C Installation height (camera installation height) h S Height (height of traffic lights) h Z : Vehicle height (vehicle height of the virtual vehicle) L A : Distance (distance to the actual vehicle) L S : Distance (distance to the traffic light) L V : Following distance L Z : Distance (distance to the virtual vehicle) TL:Traffic light

Claims

1. A video recognition unit that recognizes the image from a camera that takes pictures of the front of the vehicle, A distance calculation unit calculates a recommended distance range between the vehicle and the preceding vehicle based on the video recognized by the video recognition unit. A driver assistance device equipped with, The aforementioned inter-vehicle distance calculation unit, The distance to the traffic light at the time the traffic light is recognized by the image recognition unit is calculated based on the relationship between the FOE (Focus of Expansion) in the camera's image, the relative position of the traffic light, the installation height of the camera, and the height of the traffic light. Assuming a hypothetical vehicle whose ride height is at the upper limit of the assumed range, The distance to the virtual vehicle at the time the traffic light is recognized by the video recognition unit is calculated based on the relationship between the distance to the traffic light at the time the traffic light is recognized by the video recognition unit, the height of the virtual vehicle, the installation height of the camera, and the height of the traffic light. When the traffic light is recognized by the video recognition unit, if there is no difference between the measured distance to the actual vehicle preceding the vehicle and the distance to the virtual vehicle, the recommended range of the following distance is calculated based on the relationship between the distance to the traffic light at the time the traffic light is recognized by the video recognition unit, the height of the virtual vehicle, the installation height of the camera, and the height of the traffic light. If there is a difference between the measured value and the distance to the virtual vehicle at the time the traffic light is recognized by the video recognition unit, the vehicle height of the actual vehicle is calculated based on the relationship between the distance to the traffic light at the time the traffic light is recognized by the video recognition unit, the measured value at the time the traffic light is recognized by the video recognition unit, the installation height of the camera, and the height of the traffic light. The recommended range for the following distance is calculated based on the relationship between the distance to the traffic light at the time the traffic light is recognized by the video recognition unit, the vehicle height of the actual vehicle, the installation height of the camera, and the height of the traffic light. Driving assistance system.

2. The driving assistance device according to claim 1, further comprising a notification unit that outputs notification information to notify the driver of the recommended range of the distance between vehicles calculated by the distance between vehicles calculation unit.

3. The driver assistance device according to claim 2, wherein the notification unit outputs notification information to notify the driver to widen the distance between vehicles if the measured value becomes smaller than the recommended range of the distance between vehicles calculated by the distance between vehicles calculation unit after the time when the traffic light is recognized by the image recognition unit.

4. The driving assistance device according to claim 1 or 2, wherein the height of the virtual vehicle is the legally stipulated upper limit.

5. The inter-vehicle distance calculation unit calculates a recommended range for the inter-vehicle distance between the vehicle and the preceding vehicle based on the image recognized by the image recognition unit, but only after the distance between the vehicle and the traffic light becomes less than or equal to a predetermined value that is at least one times the visible distance of the camera. The driving support device according to claim 1 or 2.

6. A video recognition procedure for recognizing the image from a camera that takes pictures of the front of the vehicle, A distance calculation procedure for calculating the recommended distance between the vehicle and the preceding vehicle based on the video recognized by the aforementioned video recognition procedure, and A driving assistance method that is performed using a control device, The procedure for calculating the distance between vehicles is as follows: The distance to the traffic light at the time it is recognized by the aforementioned video recognition procedure is calculated based on the relationship between the Focus of Expansion (FOE) in the camera's video, the relative position of the traffic light, the camera's installation height, and the height of the traffic light. Assuming a hypothetical vehicle whose ride height is at the upper limit of the assumed range, The distance to the virtual vehicle at the time the traffic light is recognized by the aforementioned video recognition procedure is calculated based on the relationship between the distance to the traffic light at the time the traffic light is recognized by the aforementioned video recognition procedure, the height of the virtual vehicle, the installation height of the camera, and the height of the traffic light. If, at the time the traffic light is recognized by the video recognition procedure, there is no difference between the measured distance to the actual vehicle preceding the vehicle and the distance to the virtual vehicle, the recommended range of the following distance is calculated based on the relationship between the distance to the traffic light at the time the traffic light is recognized by the video recognition procedure, the height of the virtual vehicle, the installation height of the camera, and the height of the traffic light. If there is a difference between the measured value and the distance to the virtual vehicle at the time the traffic light is recognized by the video recognition procedure, the vehicle height of the actual vehicle is calculated based on the relationship between the distance to the traffic light at the time the traffic light is recognized by the video recognition procedure, the measured value at the time the traffic light is recognized by the video recognition procedure, the camera's installation height, and the traffic light's height. The recommended range for the following distance is calculated based on the relationship between the distance to the traffic light at the time the traffic light is recognized by the video recognition procedure, the vehicle height of the actual vehicle, the camera's installation height, and the traffic light's height. Driving assistance methods.

7. Image recognition processing that recognizes the image from a camera that takes pictures of the front of the vehicle, A distance calculation process that calculates a recommended distance range between the vehicle and the preceding vehicle based on the image recognized by the aforementioned image recognition process. A program to cause the control device to execute, The aforementioned process for calculating the distance between vehicles is: The distance to the traffic light at the time it is recognized by the aforementioned image recognition process is calculated based on the relationship between the Focus of Expansion (FOE) in the camera's image, the relative position of the traffic light, the camera's installation height, and the height of the traffic light. Assuming a hypothetical vehicle whose ride height is at the upper limit of the assumed range, The distance to the virtual vehicle at the time the traffic light is recognized by the aforementioned video recognition process is calculated based on the relationship between the distance to the traffic light at the time the traffic light is recognized by the aforementioned video recognition process, the height of the virtual vehicle, the installation height of the camera, and the height of the traffic light. If, at the time the traffic light is recognized by the aforementioned video recognition process, there is no difference between the measured distance to the actual vehicle preceding the vehicle and the distance to the virtual vehicle, the recommended range of the following distance is calculated based on the relationship between the distance to the traffic light at the time the traffic light is recognized by the aforementioned video recognition process, the height of the virtual vehicle, the installation height of the camera, and the height of the traffic light. If there is a difference between the measured value and the distance to the virtual vehicle at the time the traffic light is recognized by the video recognition process, the vehicle height of the actual vehicle is calculated based on the relationship between the distance to the traffic light at the time the traffic light is recognized by the video recognition process, the measured value at the time the traffic light is recognized by the video recognition process, the camera's installation height, and the traffic light's height. The recommended range for the following distance is calculated based on the relationship between the distance to the traffic light at the time the traffic light is recognized by the video recognition process, the vehicle height of the actual vehicle, the camera's installation height, and the traffic light's height. program.

Images