A driving assistance adaptive alarm triggering method and system

Abstract

The present invention discloses a method for adaptively triggering a driving assistance alarm, which includes the following steps: Step S10, obtaining the driver's line-of-sight direction and driver attention level information through a driver monitoring camera; Step S11, after the ADAS system receives the driver's line-of-sight direction information output by the driver monitoring camera, matching it with the current risk existence direction to obtain a driver's line-of-sight and risk matching degree coefficient; Step S12, obtaining a driver attention gain coefficient according to the driver's line-of-sight and risk matching degree coefficient and the driver attention level information; Step S13, using the driver attention gain coefficient to adjust the alarm trigger threshold of the ADAS system and performing alarm trigger judgment processing. The present invention also discloses a corresponding system. Implementing the present invention can adaptively adjust the alarm timing, thereby reducing false triggers and leaving enough reaction time for the driver, improving driving safety.

Description

Technical Field

[0001] The present invention relates to the technical field of driving assistance, and particularly to a driving assistance adaptive alarm triggering method and system. Background Art

[0002] In the field of passenger vehicle driving assistance, there have been many active safety driving assistance functions mass-produced on the market and the penetration rate is continuously increasing, such as forward collision warning (FCW), automatic emergency braking (AEB), lane departure warning (LDW), emergency lane keeping in blind spot (ELKA), etc. These systems are mainly composed of sensors for environmental detection such as cameras, millimeter wave radars, ultrasonic radars, signals obtained on the vehicle platform, controllers, and actuators. The monitoring of the driver's state by the system is limited to perceiving the driver's intention and state through sensors of the pedal and steering wheel. For example, when the accelerator / brake pedal is depressed deeply, the steering wheel rotation angle and speed are large, etc., it is considered that the driver is actively driving, and then the difference between triggering or suppressing the alarm / control intervention is made, but the specific triggering timing cannot be adaptively adjusted.

[0003] Among the existing technical solutions, the intervention and adjustment of the monitoring timing are mainly carried out in the following three ways:

[0004] Firstly, judging the driver's state through the positions of the brake pedal / throttle pedal, the steering wheel angle and rotation speed to suppress the alarm;

[0005] In the FCW system, suppression conditions are usually set for the positions of the brake pedal / throttle pedal, the steering wheel angle and rotation speed. For example, when any one of the conditions that the throttle pedal depth is greater than 85%, the steering wheel angle is greater than 120°, and the steering wheel angular velocity is greater than 105° / s is satisfied, FCW will not be triggered. The purpose of such setting is that when any of the above conditions is satisfied, it can be assumed that the driver is concentrating on driving at this time. On the contrary, when none of the above conditions is satisfied, FCW may be triggered when the TTC threshold is satisfied. However, the problem is that even if "none of the above conditions is satisfied", the driver may still be driving attentively, so there may inevitably be false triggers. And if the above conditions are narrowed (the threshold is decreased), then the problem of missed triggers may be brought. Here is a specific example:

[0006] Such as Figure 1As shown, a schematic diagram of an overtaking scenario in the prior art is presented. Assume there is a vehicle in front of and to the left front of the host vehicle, both with the same and constant speed \(v_t\), and the longitudinal distance between the two vehicles is \(d_x\). The speed of the host vehicle \(v_e > v_t\), and all three vehicles are driving in the center of the lane. In this case, overtaking and lane-changing by the host vehicle are relatively common driving conditions. However, different drivers have different driving habits. Some drivers (safety-conservative type) will choose to abandon overtaking and lane-changing when the longitudinal distance \(d_x\) between the two vehicles is relatively small, while some drivers (skilled and aggressive type) will still choose to overtake and lane-change when \(d_x\) is very small, and the overtaking and lane-changing routes selected by different drivers may also vary.

[0007] Then, when \(d_x\) is small and the driver is driving smoothly, the above conditions (the depth of the accelerator pedal is greater than 85%, the steering wheel angle is greater than 120°, and the angular velocity of the steering wheel is greater than 105° / s) will not be met. Then, when the TTC meets the threshold, the FCW may be triggered, and this is a false trigger for the driver at this time.

[0008] Second, provide the driver with several fixed warning timing options and let the driver choose by himself;

[0009] Continuing with the previous example, for some drivers (safety-conservative type), it is more appropriate to select the trigger timing as "far" / "early", that is, corresponding to a larger TTC (Time To Collision), and correspondingly, the alarm will be triggered earlier and in more scenarios in the same scenario. This is safer for him (more reaction time in case of a real emergency scenario), and because of his conservative driving style, there are fewer false triggers; while for some drivers (skilled and aggressive type), it is more appropriate to select the trigger timing as "near" / "late", otherwise there are many false trigger scenarios, which will instead interfere with their driving.

[0010] However, even for skilled and aggressive drivers, there will inevitably be times when they are distracted or fatigued. In this case, a relatively late trigger timing is actually unsafe, and the reaction time left for the driver may not be enough. Also, due to the need to balance the requirement that there should not be too many "false triggers", even for the "far" / "early" trigger timing, when the driver is really distracted or fatigued, the reserved reaction time may not be sufficient either.

[0011] Third, a driver fatigue monitoring system and its sensor (camera).

[0012] The driver fatigue monitoring system usually consists of a camera installed behind the steering wheel or on the A-pillar, etc., which takes pictures of the driver's face to monitor the driver's attention level, visual attention direction, fatigue level, etc., and can also implement the function of face-swiping to log in to the vehicle user through facial features. Thus, functions such as distraction reminder, fatigue reminder, and face-swiping login can be provided to the user.

[0013] However, the main drawbacks of these existing technical solutions are as follows: It is impossible to adaptively adjust the triggering timing and suppression conditions of active safety driving assistance functions in real time in combination with the driver's state. The contradiction between "reducing false triggering (delaying the triggering timing)" and "leaving enough reaction time for the driver (advancing the triggering timing)" cannot be fundamentally reconciled. Even if several freely selectable triggering timing gears are provided for the driver, it is impossible to cover different drivers or different attention levels of the same driver. Moreover, often in order to "reduce false triggering", for the gear with a "far" / "early" triggering timing, when the driver is really distracted or fatigued, the reserved reaction time is not enough, posing a safety hazard. Summary of the Invention

[0014] The technical problem to be solved by the present invention is to provide a driving assistance adaptive alarm triggering method and system, which can adaptively adjust the alarm timing in combination with a driver state monitoring system, so as to not only reduce false triggering but also leave enough reaction time for the driver, improving driving safety.

[0015] To solve the above technical problem, as one aspect of the present invention, a driving assistance adaptive alarm triggering method is provided, including the following steps:

[0016] Step S10, obtaining the driver's line-of-sight direction and driver attention level information through a driver monitoring camera;

[0017] Step S11, after the ADAS system receives the driver's line-of-sight direction information output by the driver monitoring camera, matching it with the current risk existence direction to obtain a driver's line-of-sight and risk matching degree coefficient;

[0018] Step S12, obtaining a driver attention gain coefficient according to the driver's line-of-sight and risk matching degree coefficient and the driver attention level information;

[0019] Step S13, using the driver attention gain coefficient to adjust the alarm triggering threshold of the ADAS system and performing alarm triggering judgment processing.

[0020] Among them, the step S10 further includes:

[0021] Obtaining an image of the driver through a driver monitoring camera, analyzing and obtaining the coordinate or area information corresponding to the driver's line-of-sight direction, determining it as the driver's line-of-sight direction information, and transmitting it to the ADAS system;

[0022] Obtain the driver's face image through the driver monitoring camera, judge the closing degree of the driver's eyelids, and combine the images in the past period of time to determine the driver's attention level information and transmit it to the ADAS system; wherein, the driver's attention level information is a real number between 0 and 1, or is a fatigue level.

[0023] Wherein, the step S11 further includes:

[0024] Calculate according to the driver's line of sight direction and the current risk existence direction through the following formula to obtain the coefficient of matching degree between the driver's line of sight and the risk:

[0025] Coefficient of matching degree between the driver's line of sight and the risk =

[0026] A_θ * [1 - tanh(abs(θ_h - θ_a) / 30°)] + (1 - A_θ) * [1 - tanh(abs(D_h - D_a) / 100m)]

[0027] Wherein, A_θ is the weight of the azimuth angle in the matching degree between the driver's line of sight and the risk, which is a predetermined calibration;

[0028] 1 - A_θ is the weight of the distance in the matching degree between the driver's line of sight and the risk;

[0029] θ_h is the angular value of the polar coordinates of the risk existence position;

[0030] θ_a is the angular value of the polar coordinates of the driver's line of sight position;

[0031] D_h is the distance value of the polar coordinates of the risk existence position;

[0032] D_a is the distance value of the polar coordinates of the driver's line of sight position.

[0033] Wherein, the step S12 further includes:

[0034] Step S120, calculate to obtain the intermediate value G of the ADAS alarm driver attention gain coefficient through the following formula DALtemp ):

[0035] G DALtemp = Driver attention level × Coefficient of matching degree between the driver's line of sight and the risk

[0036] Step S121, pre-determine the upper limit value G of the driver attention gain coefficient DALMax and the lower limit value G DALMin ), and use the conversion function: f_coefficient conversion() to perform conversion to obtain the final value G of the ADAS alarm driver attention gain coefficient DAL :

[0037] G DAL= f_coefficient conversion(G DALMiddle ) = G DALMax *(1 - G DALtemp ) + G DALMin *G DALtemp

[0038] wherein, the upper limit value G DALMax of the driver attention gain coefficient is a real number greater than or equal to 1, which characterizes the degree of sensitivity of the system expected to be adjusted; the lower limit value G DALMin of the driver attention gain coefficient is a real number greater than 0 and less than or equal to 1, which characterizes the degree of sluggishness of the system expected to be adjusted.

[0039] Wherein, the step S13 further includes:

[0040] Multiplying the parameter threshold of the ADAS alarm trigger judgment by the ADAS alarm driver attention gain coefficient to obtain an adjusted alarm trigger threshold;

[0041] Performing alarm trigger judgment processing using the adjusted alarm trigger threshold.

[0042] Correspondingly, on the other hand, the present invention further provides a driving assistance adaptive alarm trigger system, which includes:

[0043] A driver original state acquisition unit, configured to obtain driver line-of-sight orientation and driver attention level information through a driver monitoring camera;

[0044] A matching degree coefficient acquisition unit, configured to match with the current risk existence orientation after the ADAS system receives the driver line-of-sight orientation information output by the driver monitoring camera, and obtain a driver line-of-sight and risk matching degree coefficient;

[0045] An attention gain coefficient acquisition unit, configured to obtain a driver attention gain coefficient according to the driver line-of-sight and risk matching degree coefficient and the driver attention level information;

[0046] An alarm trigger threshold adjustment processing unit, configured to adjust the alarm trigger threshold of the ADAS system using the driver attention gain coefficient and perform alarm trigger judgment processing.

[0047] Wherein, the driver original state acquisition unit further includes:

[0048] A driver line-of-sight orientation information acquisition unit, configured to obtain an image of the driver through a driver monitoring camera, analyze and obtain coordinate or area information corresponding to the driver line-of-sight orientation, determine it as driver line-of-sight orientation information, and transmit it to the ADAS system;

[0049] A driver attention level information acquisition unit, which is used to obtain a driver's face image through a driver monitoring camera, judge the degree of closure of the driver's eyelids, and combine the images in the past period of time to determine the driver attention level information and transmit it to the ADAS system; wherein, the driver attention level information is a real number between 0 and 1, or a fatigue level.

[0050] Wherein, the matching degree coefficient acquisition unit further includes:

[0051] A calculation unit, which is used to calculate according to the driver's line of sight orientation and the current risk existence orientation through the following formula to obtain the matching degree coefficient between the driver's line of sight and the risk:

[0052] Matching degree coefficient between driver's line of sight and risk =

[0053] A_θ * [1 - tanh(abs(θ_h - θ_a) / 30°)] + (1 - A_θ) * [1 - tanh(abs(D_h - D_a) / 100m)]

[0054] Wherein, A_θ is the weight of the azimuth angle in the matching degree between the driver's line of sight and the risk, and it is a predetermined calibration;

[0055] 1 - A_θ is the weight of the distance in the matching degree between the driver's line of sight and the risk;

[0056] θ_h is the angular value of the polar coordinates of the risk existence position;

[0057] θ_a is the angular value of the polar coordinates of the driver's line of sight position;

[0058] D_h is the distance value of the polar coordinates of the risk existence position;

[0059] D_a is the distance value of the polar coordinates of the driver's line of sight position.

[0060] Wherein, the attention gain coefficient acquisition unit further includes:

[0061] An intermediate value calculation unit, which is used to calculate the intermediate value G of the ADAS alarm driver attention gain coefficient through the following formula DALtemp ):

[0062] G DALtemp = Driver attention level × Matching degree coefficient between driver's line of sight and risk

[0063] A final value acquisition unit, which is used to pre-calibrate the upper limit value G DALMax and the lower limit value G DALMin ) of the driver attention gain coefficient, and use the conversion function: f_coefficient conversion() to perform conversion to obtain the final value G of the ADAS alarm driver attention gain coefficient DAL:

[0064] G DAL = f_coefficient conversion(G DALMiddle ) = G DALMax *(1 - G DALtemp ) + G DALMin *G DALtemp

[0065] Wherein, the upper limit value G of the driver attention gain coefficient DALMax is a real number greater than or equal to 1, which represents the degree to which the system to be adjusted is expected to become sensitive; the lower limit value G of the driver attention gain coefficient DALMin is a real number greater than 0 and less than or equal to 1, which represents the degree to which the system to be adjusted is expected to become sluggish.

[0066] Wherein, the alarm trigger threshold adjustment processing unit further includes:

[0067] An adjustment unit, configured to multiply the parameter threshold of the ADAS alarm trigger judgment by the ADAS alarm driver attention gain coefficient to obtain an adjusted alarm trigger threshold;

[0068] An alarm trigger judgment unit, configured to perform alarm trigger judgment processing by using the adjusted alarm trigger threshold.

[0069] Implementing the embodiments of the present invention has the following beneficial effects:

[0070] The present invention provides a driving assistance adaptive alarm trigger method and system. It can adaptively adjust the alarm timing in combination with the driver state monitoring system, thereby reducing false triggers and leaving enough reaction time for the driver, improving driving safety.

[0071] In the embodiments of the present invention, by introducing the combination of the driver state monitoring system and the active safety driving assistance system, the trigger timing and suppression conditions of the alarm / control intervention are adaptively adjusted according to different attention states of the driver: when it is monitored that the driver is in a concentrated attention state and the perspective can better notice the dangerous working conditions, the trigger timing of the alarm is appropriately postponed, so as to ensure sufficient reaction time for the driver and reduce false alarms at the same time; when it is monitored that the driver is distracted or the attention is not in the direction where the dangerous working conditions occur, the trigger timing of the alarm / control intervention is maintained or appropriately advanced, so as to reserve more time for the driver to recover attention and ensure sufficient reaction time, thereby improving driving safety. Description of the Drawings

[0072] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, obtaining other drawings based on these drawings still belongs to the scope of the present invention.

[0073] Figure 1 It is a schematic diagram of overtaking in the prior art;

[0074] Figure 2 It is a schematic diagram of the main process of an embodiment of a driving assistance adaptive alarm triggering method provided by the present invention;

[0075] Figure 3 For Figure 2 It is a schematic diagram of the principle of obtaining the driver's line-of-sight azimuth information involved in

[0076] Figure 4 For Figure 3 It is a schematic diagram corresponding to a scene;

[0077] Figure 5 For Figure 3 It is a schematic diagram of another corresponding scene;

[0078] Figure 6 It is a schematic diagram of the structure of an embodiment of a driving assistance adaptive alarm triggering system provided by the present invention;

[0079] Figure 7 For Figure 6 It is a schematic diagram of the structure of the driver's original state acquisition unit in

[0080] Figure 8 For Figure 6 It is a schematic diagram of the structure of the attention gain system acquisition unit in

[0081] Figure 9 It is a schematic diagram of the structure of the alarm trigger threshold adjustment processing unit. Detailed implementation manners

[0082] To make the objectives, technical solutions, and advantages of the present invention clearer, the following will further describe the present invention in detail with reference to the drawings.

[0083] As Figure 2 shown, it shows a schematic diagram of the main process of an embodiment of a driving assistance adaptive alarm triggering method provided by the present invention, and in combination with Figure 5 shown. In this embodiment, the driving assistance adaptive alarm triggering method includes the following steps:

[0084] Step S10, obtain the driver's line-of-sight direction and driver attention level information through a driver monitoring camera;

[0085] Among them, the step S10 further includes:

[0086] Step S100, obtain the driver's image through a driver monitoring camera (DMS), analyze and obtain the coordinate or area information corresponding to the driver's line-of-sight direction, determine it as the driver's line-of-sight direction information, and transmit it to the ADAS system;

[0087] Example 1:

[0088] As Figure 3 shown, the driver's line-of-sight coordinates are output with reference to the vehicle's coordinate system (the front of the vehicle is the positive X-axis, the driver's left hand direction is the positive Y-axis, and the coordinate origin is the center of the rear axle of the vehicle), and the specific coordinates of the driver's current gaze are output. For example, in the following figure, the driver of this vehicle is looking at the middle of the rear of the gray car in front. At this time, the driver's line-of-sight coordinates (polar coordinates) are (0°, 35m). If the driver is looking at the pedestrian crossing the road on the left, then the driver's line-of-sight coordinates at this time are (30°, 20.5m).

[0089] Example 2:

[0090] After the driver's line of sight is divided into areas according to the requirements of the ADAS system, the area is output. That is, the driver's gaze area is divided according to the front windshield, interior rearview mirror, left exterior rearview mirror, right exterior rearview mirror, instrument panel, center control screen, etc., and then the DMS outputs the driver's current gaze area to the ADAS.

[0091] Step S101, obtain the driver's face image through a driver monitoring camera, judge the closing degree of the driver's eyelids and combine the images in the past period of time, determine the driver attention level information and transmit it to the ADAS system; among them, the driver attention level information is a real number between 0 and 1, 0 represents no attention, such as extreme fatigue, and 1 represents very alert and concentrating on driving; or use a fatigue level (such as the KSS fatigue level) and convert it into a real number between 0 and 1.

[0092] It can be understood that the driver monitoring camera can also output the effective states corresponding to these two parameters. When the specific values and effectiveness of the parameters cannot be determined (such as when the camera is blocked and the driver cannot be monitored), the parameter invalid should be output.

[0093] Step S11, after the ADAS system receives the driver's line-of-sight direction information output by the driver monitoring camera, match it with the current risk existence direction to obtain the driver's line-of-sight and risk matching degree coefficient;

[0094] Among them, the step S11 further includes:

[0095] Calculate according to the driver's line-of-sight direction and the current risk position through the following formula to obtain the coefficient of matching degree between the driver's line of sight and the risk:

[0096] Coefficient of matching degree between the driver's line of sight and the risk =

[0097] A_θ * [1 - tanh(abs(θ_h - θ_a) / 30°)] + (1 - A_θ) * [1 - tanh(abs(D_h - D_a) / 100m)]

[0098] Where A_θ is the weight of the azimuth angle in the matching degree between the driver's line of sight and the risk, which is a predetermined calibration;

[0099] 1 - A_θ is the weight of the distance in the matching degree between the driver's line of sight and the risk;

[0100] θ_h is the angular value of the polar coordinates of the risk position;

[0101] θ_a is the angular value of the polar coordinates of the driver's line-of-sight position;

[0102] D_h is the distance value of the polar coordinates of the risk position;

[0103] D_a is the distance value of the polar coordinates of the driver's line-of-sight position.

[0104] Step S12, obtain the driver attention gain coefficient according to the coefficient of matching degree between the driver's line of sight and the risk and the driver attention level information;

[0105] Example 3:

[0106] As Figure 4 shown, referring to the working condition in Example 1, where the pedestrian is walking to the left and the vehicle in front is braking suddenly, and the ADAS system detects a collision risk with the vehicle in front, then the risk position at this time is directly in front, or using the polar coordinate system, the azimuth is (θ, D) = (0°, 35m).

[0107] As Figure 5 shown, referring to the working condition in Example 1, where the pedestrian is walking across the road to the right and the vehicle in front is driving forward normally, and the ADAS system detects a collision risk with the pedestrian crossing, then the risk position at this time is the front left, or using the polar coordinate system, the azimuth is (θ, D) = (30°, 20.5m).

[0108] If the driver's line-of-sight azimuth outputs (30°, 20.5 m) and the current risk existence azimuth is also (30°, 20.5 m), the matching degree coefficient between the driver's line of sight and the risk should be relatively high, which is a real number in the range of [0, 1] and close to 1. At this time, there is a collision risk between the vehicle and the pedestrian crossing the road, and the driver is exactly looking at the pedestrian. If the driver's line-of-sight azimuth outputs (0°, 35 m) and the current risk existence azimuth is (30°, 20.5 m), the matching degree coefficient between the driver's line of sight and the risk should be relatively low, which is a real number in the range of [0, 1] and close to 0. At this time, there is a collision risk between the vehicle and the pedestrian crossing the road, but the driver is looking at the vehicle ahead.

[0109] Among them, the step S12 further includes:

[0110] Step S120, if the driver's attention level has been output as a real number in the range of [0, 1], calculate the intermediate value G of the ADAS alarm driver attention gain coefficient through the following formula DALtemp ):

[0111] G DALtemp = driver's attention level × matching degree coefficient between the driver's line of sight and the risk

[0112] Step S121, pre-determine the upper limit value G DALMax and the lower limit value G DALMin ) of the driver attention gain coefficient, and use the conversion function: f_coefficient conversion() to perform conversion to obtain the final value G of the ADAS alarm driver attention gain coefficient DAL :

[0113] G DAL = f_coefficient conversion(G DALMiddle ) = G DALMax *(1 - G DALtemp ) + G DALMin *G DALtemp

[0114] Among them, the upper limit value G DALMax of the driver attention gain coefficient is a real number greater than or equal to 1, which represents the degree of sensitivity of the system to be adjusted as expected; the lower limit value G DALMin of the driver attention gain coefficient is a real number greater than 0 and less than or equal to 1, which represents the degree of dullness of the system to be adjusted as expected; the above two values can be obtained by calibrating different vehicle models.

[0115] Step S13, use the driver attention gain coefficient to adjust the alarm trigger threshold of the ADAS system, and perform alarm trigger judgment processing.

[0116] Among them, the step S13 further includes:

[0117] Multiply the parameter threshold for ADAS alarm triggering judgment by the ADAS alarm driver attention gain coefficient to obtain an adjusted alarm triggering threshold;

[0118] Use the adjusted alarm triggering threshold to perform alarm triggering judgment processing.

[0119] Taking Example 3 in Step S12 as an example, a pedestrian is walking across the road to the right, and the vehicle in front is driving forward normally. The ADAS system detects a collision risk with the crossing pedestrian. The driver is looking at the gray car in front. Then the driver's attention does not match the risk position very well, and the coefficient of the driver's line of sight matching the risk is 0.2. At the same time, the driver's attention is average at this time, and the driver's attention level is 0.5. Then G DALtemp = 0.5 × 0.2 = 0.1, and the selected G DALMax and G DALMin are 1.5 (maximum sensitivity adjustment of 50%) and 0.8 (maximum dullness adjustment of 20%) respectively. Then G DAL = 1.5 × (1 - 0.1) + 0.8 × 0.1 = 1.43. Then, if the TTC for triggering an alarm for a crossing pedestrian before introducing driver state adaptive adjustment is 0.8 s, after introducing driver state adaptive adjustment, multiply the two (0.8 × 1.43) to obtain an adjusted TTC of 1.144 s. Then, in this case where the driver's attention is average and the driver does not notice the risk, the alarm time can be advanced by 43%.

[0120] In addition to using the ADAS alarm driver attention gain coefficient for the most critical parameter threshold of TTC, other parameter thresholds (such as the accelerator pedal depth being greater than 85%, the steering wheel angle being greater than 120°, and the steering wheel angular velocity being greater than 105° / s) can also be adjusted in this way, and the achieved effect is to advance the alarm timing.

[0121] It can be understood that the method provided by the present invention combines the DMS system and the ADAS system, and adaptively adjusts the alarm timing and suppression conditions of the ADAS system according to the driver state, so as to appropriately advance the alarm timing when the driver's attention is low (such as fatigue) or inattentive (such as operating the central control screen), and reduce the probability that the alarm is suppressed by the driver's vehicle control actions; conversely, the alarm timing can be appropriately postponed to increase the probability that the alarm is suppressed by the driver's vehicle control actions.

[0122] The advantages of the present invention are as follows: It reconciles the contradiction between "reducing false triggers (delaying the trigger timing)" and "leaving enough reaction time for the driver (advancing the trigger timing)". In fact, whether the trigger timing and suppression conditions of the ADAS system's alarm / control intervention are appropriate is highly related to the driver's state. However, the existing systems cannot monitor the driver's state well, so the trigger timing and suppression conditions for the system to trigger the alarm / control intervention cannot achieve a good balance between the two. Adaptive adjustment of the trigger timing and suppression conditions of the alarm / control intervention according to the driver's different attention states can achieve this: when it is detected that the driver is in a concentrated attention state and the perspective can better notice the dangerous working conditions, the trigger timing of the alarm is appropriately delayed, so that the driver has enough reaction time (alert drivers require less reaction time than distracted drivers), and at the same time, false alarms can be reduced; when it is detected that the driver is distracted or the attention is not in the direction where the dangerous working conditions occur, the trigger timing of the alarm / control intervention is maintained or appropriately advanced, so that more time can be left for the driver to regain attention (distracted drivers require longer reaction time than alert drivers), ensuring sufficient reaction time.

[0123] As Figure 6 shown, a structural schematic diagram of an embodiment of a driving assistance adaptive alarm trigger system provided by the present invention is shown. In combination with Figures 7 to 9 shown, in this embodiment, the system includes:

[0124] A driver original state acquisition unit 10, configured to obtain the driver's line-of-sight direction and driver attention level information through a driver monitoring camera 1;

[0125] A risk direction monitoring unit 20, configured to monitor the direction where the current risk exists through an ADAS system 2;

[0126] A matching degree coefficient acquisition unit 21, configured to, after the ADAS system 2 receives the driver's line-of-sight direction information output by the driver monitoring camera, match it with the direction where the current risk exists to obtain a driver line-of-sight and risk matching degree coefficient;

[0127] An attention gain coefficient acquisition unit 22, configured to obtain a driver attention gain coefficient according to the driver line-of-sight and risk matching degree coefficient and the driver attention level information;

[0128] An alarm trigger threshold adjustment and processing unit 23, configured to adjust the alarm trigger threshold of the ADAS system using the driver attention gain coefficient and perform alarm trigger judgment processing.

[0129] Wherein, the driver original state acquisition unit 10 further includes:

[0130] The driver's line-of-sight orientation information acquisition unit 100 is configured to obtain an image of the driver through a driver monitoring camera, analyze and obtain the coordinates or area information corresponding to the driver's line-of-sight orientation, determine it as the driver's line-of-sight orientation information, and transmit it to the ADAS system;

[0131] The driver's attention level information acquisition unit 101 is configured to obtain the driver's face image through a driver monitoring camera, judge the closing degree of the driver's eyelids and combine the images in the past period of time, determine the driver's attention level information and transmit it to the ADAS system; wherein, the driver's attention level information is a real number between 0 and 1, or a fatigue level.

[0132] Wherein, the matching degree coefficient acquisition unit 21 further includes:

[0133] A calculation unit, configured to calculate according to the driver's line-of-sight orientation and the current risk existence orientation through the following formula to obtain the driver's line-of-sight and risk matching degree coefficient:

[0134] Driver's line-of-sight and risk matching degree coefficient =

[0135] A_θ*[1 - tanh(abs(θ_h - θ_a) / 30°)]+(1 - A_θ)*[1 - tanh(abs(D_h - D_a) / 100m)]

[0136] Wherein, A_θ is the weight of the azimuth angle in the driver's line-of-sight and risk matching degree, which is a predetermined calibration;

[0137] 1 - A_θ is the weight of the distance in the driver's line-of-sight and risk matching degree;

[0138] θ_h is the angular value of the polar coordinates of the risk existence position;

[0139] θ_a is the angular value of the polar coordinates of the driver's line-of-sight position;

[0140] D_h is the distance value of the polar coordinates of the risk existence position;

[0141] D_a is the distance value of the polar coordinates of the driver's line-of-sight position.

[0142] Wherein, the attention gain coefficient acquisition unit 22 further includes:

[0143] An intermediate value calculation unit 220, configured to calculate and obtain the intermediate value G of the ADAS alarm driver attention gain coefficient through the following formula DALtemp ):

[0144] G DALtemp = Driver's attention level × Driver's line-of-sight and risk matching degree coefficient

[0145] A final value obtaining unit 221 is configured to pre-calibrate an upper limit value G of the driver attention gain coefficient DALMax and a lower limit value G DALMin ) and perform conversion by using a conversion function: f_coefficient conversion() to obtain a final value G of the ADAS alarm driver attention gain coefficient DAL :

[0146] G DAL = f_coefficient conversion(G DALMiddle ) = G DALMax *(1 - G DALtemp ) + G DALMin *G DALtemp

[0147] wherein, the upper limit value G of the driver attention gain coefficient DALMax is a real number greater than or equal to 1, which represents the degree to which the system is expected to be made more sensitive; the lower limit value G of the driver attention gain coefficient DALMin is a real number greater than 0 and less than or equal to 1, which represents the degree to which the system is expected to be made less sensitive.

[0148] Wherein, the alarm trigger threshold adjustment processing unit 23 further includes:

[0149] An adjustment unit 230 is configured to multiply a parameter threshold for ADAS alarm trigger judgment by the ADAS alarm driver attention gain coefficient to obtain an adjusted alarm trigger threshold;

[0150] An alarm trigger judgment unit 231 is configured to perform alarm trigger judgment processing by using the adjusted alarm trigger threshold.

[0151] Implementing the embodiments of the present invention has the following beneficial effects:

[0152] The present invention provides a driving assistance adaptive alarm trigger method and system. It can adaptively adjust the alarm timing in combination with the driver state monitoring system, so as to not only reduce false triggers, but also leave enough reaction time for the driver and improve driving safety.

[0153] In the embodiments of the present invention, by introducing the combination of a driver state monitoring system and an active safety driving assistance system, the triggering timing and suppression conditions of alarm / control intervention are adaptively adjusted according to different attention states of the driver: when it is monitored that the driver is in a concentrated attention state and the perspective can better notice dangerous working conditions, the triggering timing of the alarm is appropriately postponed, so that the driver has enough reaction time and at the same time false alarms can be reduced; when it is monitored that the driver is distracted or the attention is not in the direction where the dangerous working condition occurs, the triggering timing of the alarm / control intervention is maintained or appropriately advanced, so that more time can be reserved for the driver to regain attention and ensure sufficient reaction time. Thereby, the driving safety is improved.

[0154] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, an apparatus, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0155] The present invention is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, and the combination of processes and / or blocks in the flowcharts and / or block diagrams, can be realized by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for realizing the functions specified in one Figure 1 one process or multiple processes and / or blocks Figure 1 a block or multiple blocks.

[0156] The above-disclosed is only a preferred embodiment of the present invention, and of course, it cannot be used to limit the scope of the rights of the present invention. Therefore, equivalent changes made according to the claims of the present invention still fall within the scope covered by the present invention.

Claims

1. A driving assistance adaptive alarm triggering method, characterized in that, It includes the following steps: Step S10: Obtain the driver's line-of-sight direction and driver attention level information through a driver monitoring camera; Step S11: After the ADAS system receives the driver's line-of-sight direction information output by the driver monitoring camera, match it with the current risk presence direction to obtain the driver's line-of-sight and risk matching degree coefficient; Step S12: Obtain the driver attention gain coefficient according to the driver's line-of-sight and risk matching degree coefficient and the driver attention level information; Step S13: Use the driver attention gain coefficient to adjust the alarm trigger threshold of the ADAS system and perform alarm trigger judgment processing; Among them, step S11 further includes: Calculate according to the driver's line-of-sight direction and the current risk presence direction through the following formula to obtain the driver's line-of-sight and risk matching degree coefficient: Driver's line-of-sight and risk matching degree coefficient = A_θ*[1 - tanh(abs(θ_h - θ_a) / 30°)]+(1 - A_θ)*[1 - tanh(abs(D_h - D_a) / 100m)] Where, A_θ is the weight of the azimuth angle in the driver's line-of-sight and risk matching degree, which is a predetermined calibration; 1 - A_θ is the weight of the distance in the driver's line-of-sight and risk matching degree; θ_h is the angular value of the polar coordinates of the risk presence position; θ_a is the angular value of the polar coordinates of the driver's line-of-sight position; D_h is the distance value of the polar coordinates of the risk presence position; D_a is the distance value of the polar coordinates of the driver's line-of-sight position.

2. The method according to claim 1, wherein Step S10 further includes: Obtain the driver's image through the driver monitoring camera, analyze and obtain the coordinate or area information corresponding to the driver's line-of-sight direction, determine it as the driver's line-of-sight direction information, and transmit it to the ADAS system; Obtain the driver's face image through the driver monitoring camera, judge the closing degree of the driver's eyelids and combine the images in the past period of time, determine the driver attention level information and transmit it to the ADAS system; among them, the driver attention level information is a real number between 0 and 1.

3. The method according to claim 2, characterized in that, Step S12 further includes: Step S120, calculate and obtain the intermediate value G of the ADAS alarm driver attention gain coefficient through the following formula DALtemp ) G DALtemp = Driver attention level × Coefficient of matching between driver's line of sight and risk Step S121, pre-determine the upper limit value G of the driver attention gain coefficient DALMax and the lower limit value G DALMin ), and obtain the final value G of the ADAS alarm driver attention gain coefficient by using the following formula DAL : G DAL = G DALMax *(1 - G DALtemp ) + G DALMin * G DALtemp Among them, the upper limit value G of the driver attention gain coefficient DALMax is a real number greater than or equal to 1, which characterizes the degree of sensitivity of the system expected to be adjusted; the lower limit value G of the driver attention gain coefficient DALMin is a real number greater than 0 and less than or equal to 1, which characterizes the degree of insensitivity of the system expected to be adjusted.

4. The method according to claim 3, wherein Step S13 further includes: Multiply the parameter threshold of the ADAS alarm trigger judgment by the ADAS alarm driver attention gain coefficient to obtain the adjusted alarm trigger threshold; Use the adjusted alarm trigger threshold to perform alarm trigger judgment processing.

5. A driving assistance adaptive alarm trigger system, characterized in that, It includes: Driver original state acquisition unit, used to obtain the driver's line-of-sight direction and driver attention level information through a driver monitoring camera; Matching degree coefficient acquisition unit, used to match with the current risk presence direction after the ADAS system receives the driver's line-of-sight direction information output by the driver monitoring camera to obtain the driver's line-of-sight and risk matching degree coefficient; Attention gain coefficient acquisition unit, used to obtain the driver attention gain coefficient according to the driver's line-of-sight and risk matching degree coefficient and the driver attention level information; Alarm trigger threshold adjustment processing unit, used to adjust the alarm trigger threshold of the ADAS system with the driver attention gain coefficient and perform alarm trigger judgment processing; Among them, the matching degree coefficient obtaining unit further includes: A calculation unit, configured to calculate according to the driver's line-of-sight orientation and the current risk existence orientation through the following formula to obtain the matching degree coefficient between the driver's line of sight and the risk: Matching degree coefficient between driver's line of sight and risk = A_θ * [1 - tanh(abs(θ_h - θ_a) / 30°)] + (1 - A_θ) * [1 - tanh(abs(D_h - D_a) / 100m)] where A_θ is the weight of the azimuth angle in the matching degree between the driver's line of sight and the risk, which is a predetermined calibration; 1 - A_θ is the weight of the distance in the matching degree between the driver's line of sight and the risk; θ_h is the angular value of the polar coordinates of the risk existence position; θ_a is the angular value of the polar coordinates of the driver's line of sight position; D_h is the distance value of the polar coordinates of the risk existence position; D_a is the distance value of the polar coordinates of the driver's line of sight position.

6. The system according to claim 5, characterized in that The driver's original state obtaining unit further includes: A driver's line-of-sight orientation information obtaining unit, configured to obtain an image of the driver through a driver monitoring camera, analyze and obtain the coordinate or area information corresponding to the driver's line-of-sight orientation, determine it as the driver's line-of-sight orientation information, and transmit it to the ADAS system; A driver's attention level information obtaining unit, configured to obtain the driver's face image through a driver monitoring camera, judge the closing degree of the driver's eyelids and combine the images in the past period of time to determine the driver's attention level information and transmit it to the ADAS system; wherein, the driver's attention level information is a real number between 0 and 1.

7. The system according to claim 6, wherein The attention gain coefficient obtaining unit further includes: An intermediate value calculation unit for calculating and obtaining an intermediate value G of the ADAS alarm driver attention gain coefficient through the following formula DALtemp ) G DALtemp = Driver attention level × Coefficient of matching between driver's line of sight and risk A final value obtaining unit, configured to pre-calibrate an upper limit value G of a driver attention gain coefficient DALMax and a lower limit value G DALMin ), and obtain a final value G of the ADAS warning driver attention gain coefficient by using the following formula DAL : G DAL = G DALMax *(1 - G DALtemp ) + G DALMin *G DALtemp Among them, the upper limit value \(G\) of the driver attention gain coefficient DALMax is a real number greater than or equal to 1, which characterizes the degree to which the system to be adjusted is expected to become more sensitive; the lower limit value \(G\) of the driver attention gain coefficient DALMin is a real number greater than 0 and less than or equal to 1, which characterizes the degree to which the system to be adjusted is expected to become less sensitive.

8. The system according to claim 7, wherein The alarm trigger threshold adjustment processing unit further includes: An adjustment unit, configured to multiply the parameter threshold of the ADAS alarm trigger judgment by the ADAS alarm driver attention gain coefficient to obtain an adjusted alarm trigger threshold; An alarm trigger judgment unit, configured to perform alarm trigger judgment processing using the adjusted alarm trigger threshold.

Images