SYSTEM FOR DETECTING DRIVER IMPAIRMENT

Abstract

System (10) for detecting impairment of a vehicle driver (14), the system comprising: an alcohol sensor (22); a camera (20); a start control (30); and a controller (16) electrically connected to the alcohol sensor (22), the camera (20) and the start control (30), the controller (16) including a processor (34) and a memory (36), the memory (36) including instructions such that the processor is programmed to: Determining that alcohol has been detected using the alcohol sensor (22); Deactivating the start control (30) due to the detection of alcohol; Carrying out an assessment of gaze nystagmus of the driver (14) using the camera (20); Determining, based on the assessment of the gaze nystagmus, whether the driver (14) is impaired; and Activating the start control (30) if the driver (14) is not impaired, wherein the processor (34) is further programmed to: perform the assessment of the gaze nystagmus of the vehicle operator (14) by tracking an eye movement of the vehicle operator (14) and recording gaze samples over time during the eye movement, the gaze samples comprising horizontal positions of the eyes during the eye movement of the vehicle operator (14) at time intervals; characterized in that the processor (34) is further programmed to: Transforming the gaze samples by deriving the horizontal positions of the eyes twice to determine a velocity and an acceleration at each time interval, wherein the velocity and acceleration are approximated or smoothed; and Estimating a Gaussian mixture model using the derivatives of the horizontal positions of the eyes to determine whether there are micro-eye movements during the time intervals and to determine an amplitude and frequency of the micro-eye movements in the time intervals.

Description

INITIATIONThe present disclosure relates to a system and method for detecting impairment of a vehicle operator, and more particularly to a system and method for detecting impairment of a vehicle operator that detects whether a vehicle operator has physical signs of alcohol impairment. As prior art, reference is made to DE 10 2018 112 076 A1 and DE 10 2011 119 261 A1.The driver's impairment in driving a vehicle increases the risk of accidents. To reduce this risk, discovery solutions are proposed. One such solution is to employ an on-board chemical alcohol detector, such as the operator alcohol detection system for safety (operator alcohol detection system for safety). The on-board alcohol detector uses chemical sensors and algorithms that estimate the BAK (blood alcohol concentration) and BAC (blood alcohol content), respectively, from the breathing air of the driver. Moreover, the BAK can also be estimated by transmitting the driver's fingertip with infrared light and using useful absorption algorithms. While these alcohol detector systems can provide accurate BAK measurements, there is a need for a driver impairment detector that uses existing systems in the vehicle and is reliable and consistent. In particular, in cases where these earlier technologies are not present and / or have given erroneous measurements.SUMMARYThe present invention is defined by the features of the appended independent claim 1. Advantageous further developments are evident from the following description and from the dependent claims.A system for detecting impairment of a driver of a vehicle is provided. The system includes an alcohol sensor, a camera, a launch controller, and a controller electrically connected to the alcohol sensor, the camera, and the launch controller, the controller including a processor and a memory, the memory including instructions such that the processor is programmed to: determine alcohol is detected with the alcohol sensor, disable the launch controller based on the determination of alcohol, perform a rating of the gaze ystagmus of the driver with the camera, determine whether the driver is impaired based on the rating of the gaze ystagmus, and enable the launch controller if the driver is not impaired.In one aspect, the processor is further programmed to perform the assessment of the driver's gaze nystagmus using at least one of a display assessment using an in-vehicle display, an audio assessment using an in-vehicle audio system, and a passive assessment.According to the invention, the processor is further programmed to perform the assessment of the driver's gaze nystagmus by tracking an eye movement of the driver and recording gaze samples over time during the eye movement, the gaze samples comprising horizontal positions of the eyes during the driver's eye movement at time intervals.In another aspect, the processor is also programmed to determine whether the driver is impaired by passing the gaze samples through a gaze nystagmus model.According to the invention, the processor is further programmed to transform the gaze samples by deriving the horizontal positions of the eyes twice to determine a speed and an acceleration in each time interval, approximating or smoothing the speed and the acceleration, and to estimate a Gaussian mixture model using the derivatives of the horizontal positions of the eyes, to determine whether there are micro-eye movements during the time intervals, and to determine an amplitude and frequency of the micro-eye movements in the time intervals.In another aspect, the processor is further programmed to determine whether the driver is impaired by determining whether there are micro-eye movements during the time intervals during eye movement, determining whether there are micro-eye movements during the time intervals at a maximum horizontal position of the eyes, and determining whether there are micro-eye movements during the time intervals forty-five degrees from a center during eye movement.In another aspect, the processor is further programmed to determine whether the driver is impaired by comparing the amplitude and frequency of micro-eye movements during eye movement to a first amplitude threshold and a first frequency threshold, respectively, and determine that nystagmus is detected when the amplitude and frequency during eye movement are greater than the first amplitude threshold and the first frequency threshold, respectively.In another aspect, the processor is further programmed to determine whether the driver is impaired by comparing the amplitude and frequency of micro-eye movements during the time intervals at the maximum horizontal position of the eyes to a second amplitude threshold and a second frequency threshold, respectively, and determine that a nystagmus is detected when the amplitude and frequency during the time intervals at the maximum horizontal position of the eyes are greater than the second amplitude threshold and the second frequency threshold, respectively.In another aspect, the processor is further programmed to determine whether the driver is impaired by comparing the amplitude and frequency of micro-eye movements during the time intervals before forty-five degrees from the center during eye movement to a third amplitude threshold and a third frequency threshold, respectively, and determine that a nystagmus is detected when the amplitude and frequency are greater than the third amplitude threshold and the third frequency threshold, respectively, during the time intervals before forty-five degrees from the center during eye movement.In another aspect, the processor is further programmed to assign, for each eye, a score based on whether a nystagmus was detected during the time intervals and compare the score to a threshold for impairment, and determine that the driver is impaired if the score exceeds the threshold for impairment.In another embodiment, a method for detecting impairment of a vehicle driver is provided. The method includes determining that alcohol is detected using an alcohol sensor disposed in the vehicle, disabling start control based on the detection of alcohol in the vehicle, performing evaluation of the gaze ystagmus of the driver using a camera in the vehicle, determining that the driver is not impaired based on the gaze ystagmus, and enabling start control if the driver is not impaired.In one aspect, the assessment of the gaze nystagmus of the driver includes using at least one of a display assessment using an in-vehicle display, an audio assessment using an in-vehicle audio system, and a passive assessment.In another aspect, performing the evaluation of the gaze nystagmus of the driver includes tracking eye movement of the driver and recording gaze samples over time during eye movement, the gaze samples including horizontal positions of the eyes during eye movement of the driver at time intervals.In another aspect, determining whether the driver is impaired includes transforming the gaze samples by deriving the horizontal positions of the eyes twice to determine a speed and an acceleration in each time interval, approximating or smoothing the speed and the acceleration, and estimating a Gaussian mixture model using the derivatives of the horizontal positions of the eyes to determine whether there are micro-eye movements during the time intervals and to determine an amplitude and frequency of the micro-eye movements in the time intervals.In another aspect, determining whether the driver is impaired includes determining whether there are micro-eye movements during the time intervals during the eye movement, determining whether there are micro-eye movements during the time intervals at a maximum horizontal position of the eyes, and determining whether there are micro-eye movements during the time intervals before forty-five degrees from a center during the eye movement.In another aspect, determining whether the driver is impaired includes comparing the amplitude and frequency of micro-eye movements during eye movement to a first amplitude threshold and a first frequency threshold, respectively, and determining that nystagmus is detected when the amplitude and frequency during eye movement are greater than the first amplitude threshold and the first frequency threshold, respectively.In another aspect, determining whether the driver is impaired includes comparing the amplitude and frequency of micro-eye movements during the time intervals at the maximum horizontal position of the eyes to a second amplitude threshold and a second frequency threshold, respectively, and determining that nystagmus is detected when the amplitude and frequency during the time intervals at the maximum horizontal position of the eyes are greater than the second amplitude threshold and the second frequency threshold, respectively.In another aspect, determining whether the driver is impaired includes comparing the amplitude and frequency of micro-eye movements during time intervals before forty-five degrees from the center during eye movement with a third amplitude threshold and a third frequency threshold, respectively, and determining that nystagmus is detected when the amplitude and frequency during time intervals before forty-five degrees from the center during eye movement are greater than the third amplitude threshold and the third frequency threshold, respectively.In another aspect, the method further comprises assigning a score to each eye based on whether a nystagmus was detected during the time intervals, and comparing the score to a threshold for impairment, and determining that the driver is impaired if the score exceeds the threshold for impairment.In another embodiment, a system is provided for detecting impairment of a vehicle's driver. The system includes an alcohol sensor, a camera, a display on the vehicle, a launch controller, a controller in electrical communication with the alcohol sensor, the camera, the display, and the launch controller, the controller including a processor and a memory, the memory including instructions such that the processor is programmed to: determine, using the alcohol sensor, that alcohol is detected, disable the launch controller based on the detection of alcohol, perform a assessment of the gaze stagmus of the driver, including displaying a graphic on the display, instructing the driver to follow the graphic on the display with his eyes, and tracking eye movement using the camera to record gaze samples over time intervals, transforming gaze samples by deriving horizontal positions of the eyes twice to determine a speed and an acceleration in each time interval, approximating or smoothing the speed and the acceleration, estimating a Gaussian mixture model using the derivatives of the horizontal positions of the eyes to determine whether micro eye movements are present during the time intervals, determining an amplitude and frequency of the micro eye movements in the time intervals when micro eye movements are present, comparing the amplitude and frequency of the micro eye movements during a full eye movement, at a maximum position of the eyes and before a center of the full eye movement with one or more threshold values, assigning a score for each eye, when the amplitude and frequency of the micro-eye movements in the complete eye movement, at the maximum position of the eyes and before the midpoint of the complete eye movement exceeds the one or more threshold values, compare the value with a value for the impairment and activate the starting control if the value is less than the value for the impairment and deactivate the starting control if the value is greater than or equal to the value for the impairment.Further areas of applicability will become apparent from the description given herein. It is to be understood that the description and specific examples are intended for purposes of illustration only and are not intended to limit the scope of the present disclosure.BRIEF DESCRIPTION OF THE DRAWINGSThe drawings described herein are for illustrative purposes only and are not intended to limit the scope of the present disclosure in any way. FIG. 1 is a schematic illustration of a system for detecting operator impairment according to an example embodiment; FIG. 2A is a schematic diagram of a display assessment used to assess gaze nystagmus using the operator impairment detection system; FIG. 2B is a schematic diagram of another example of the gaze nystagmus assessment using the operator impairment detection system; FIG. 2C is a schematic diagram of the audio assessment used to assess gaze nystagmus using the operator impairment detection system; FIG. 2D is a schematic illustration of a passive assessment used to assess gaze nystagmus using the operator impairment detection system; FIG. 3 is a flow diagram of a method for determining operator impairment; FIG. 4 is a flow chart of a subroutine for determining the presence and intensity of a gaze nystagmus; and FIG. 5 is a graph illustrating the gaze nystagmus.DETAILED DESCRIPTIONThe following description is merely illustrative in nature and is not intended to limit the present disclosure, its application, or uses.Referring now to FIG. 1, a system for detecting operator impairment is shown and generally designated by the reference numeral 10. The operator impairment detection system 10 is shown with an example vehicle 12. Although a passenger car is shown, the vehicle 12 may be any type of vehicle or equipment without departing from the scope of the present disclosure. The operator impairment detection system 10 is used to detect the gaze ystagmus of a driver 14 of the vehicle 12 to determine an impairment of the driver, as will be described in more detail below. The operator impairment detection system 10 generally includes a controller 16, a display 18, a camera 20, an alcohol sensor 22, an audio system 24, a driver monitoring system 26, a human machine interface (HMI) 28, and a launch controller 30.The controller 16 is operable to perform a method 100 for detecting impairment of the driver 14 of the vehicle 12, as described below. The controller 16 includes at least one processor 34 and a non-transitory computer readable storage device or medium 36. the processor 34 may be a custom or commercially available processor, a central processing unit (CPU), a graphics processing unit (GPU), an auxiliary processor among multiple processors connected to the controller 16, a semiconductor-based microprocessor (in the form of a microchip or chipset), a macroprocessor, a combination thereof, or generally an apparatus for executing instructions. The computer readable storage device or medium 36 may include volatile and non-volatile memory, such as read-only memory (ROM), random-access memory (RAM), and keep-alive memory (KAM). KAM is a persistent or non-volatile memory that can be used to store various operating variables while the processor 34 is off. The computer readable storage device or medium 36 may be implemented using a variety of storage devices, such as programmable read only memories (PROMs), electrically EPROMs (PROMs), electrically erasable PROMs (EEPROMs), flash memory, or other electrical, magnetic, optical, or combined storage devices capable of storing data, some of which represent executable instructions, used by the controller 16 in controlling various systems of the vehicle 12. The controller 16 may also be comprised of a plurality of controllers in electrical communication with one another. The controller 16 may be connected to additional systems and / or controllers of the vehicle 12 such that the controller 16 may access data such as speed, acceleration, braking, and steering angle of the vehicle 12.The controller 16 is in electrical communication with the display 18, the camera 20, the alcohol sensor 22, the audio system 24, the driver monitoring system 26, the human-machine interface (HMI) 28, and the launch controller 30.In an exemplary embodiment, the electrical communication is established, for example, via a CAN bus, a WiFi network, a cellular data network or the like. It should be appreciated that various additional wired and wireless techniques and communication protocols for communicating with the controller 16 are within the scope of the present disclosure.The display 18 serves to perform a guided assessment during the method for detecting impairment of the driver 14. The display 18 is a wide-screen display, such as a wide-screen display on a console of the vehicle 12, a full head-up display (HUD) on a windshield 38 of the vehicle 12, a reflective or diffractive wide-screen HUD with a pillar display on the windshield 38 of the vehicle 12, or another type of wide-screen display.The camera 20 is used to record images of the eyes, in particular pupils, of the driver 14. The camera 20 may be any vehicle camera that may include various types of lenses, e.g., wide angle lenses and / or tele-angle lenses, and / or infrared LED (light emitting diode) illumination to generate corneal reflection to improve eye detection.The alcohol sensor 22 is part of a DADDS system (operator alcohol detection system for safety). The alcohol sensor 22 is used to determine the blood alcohol level of the driver 14. Alternatively, the alcohol sensor 22 may be an infrared emitter and detector that measures blood alcohol content by radiating infrared light through the fingertip of the driver 14. Various other types of alcohol sensors 22 may be used without departing from the scope of the present disclosure.The audio system 24 is for issuing commands to the driver 14 in performing a guided assessment during the method of detecting adverse effects on the driver 14. the audio system 24 includes one or more speakers located in the vehicle 12 and may also include one or more microphones.The driver monitoring system 26 is for monitoring the driver 14 during operation of the vehicle 12. The driver monitoring system 26 may utilize the display 18 as well as the audio system 24 to provide feedback to the driver 14 based on the collected information. For example, the driver monitoring system 26 may track the attention of the driver 14 using the camera 20 and issue an audible alarm via the audio system 24 when the driver monitoring system 26 determines that the driver 14 is inattentive or slap. The driver monitoring system 26 may also check whether the driver 14, which is the subject of the operator impairment detection system 10, is actually the driver 14 of the vehicle 12 by face recognition.The human machine interface (HMI) 28 is used by the driver 14 to give commands to the vehicle 12. The HMI 28 may be a touch screen display of an infotainment system, a rotary knob, or a button, or may be used to activate a voice command system via the audio system 24. In addition, the HMI 28 may include sensors for reading hand gesture inputs.The start controller 30 is for starting and stopping the vehicle 12, and the start controller 30 may be used for starting and stopping an internal combustion engine, a hybrid electric motor, or an electric motor in the vehicle 12. In the present disclosure, the operator impairment detection system 10 overrides the launch controller 30 based on an assessment of the impairment of the driver 14, as described in more detail below.The operator impairment detection system 10 is used to detect and evaluate the driver's gaze ystagmus. Nystagmus is defined as repetitive uncontrolled movement of the eyes. These movements often result in reduced vision and depth perception and may impair balance and coordination. Nystagmus correlates strongly with driver impairment, for example, at a blood alcohol content of greater than 0.8 per mille. These involuntary eye movements may occur from side to side (horizontal nystagmus), up and down (vertical nystagmus) or in a circular pattern (rotating nystagmus). In the present example, the operator impairment detection system 10 evaluates the horizontal gaze nystagmus (HGN) by detecting left-to-right eye movements along a horizontal x-axis relative to the driver 14 over a period of time. However, it should be noted that the vertical gaze ystagmus or the rotating gaze ystagmus may also be evaluated without departing from the scope of the present disclosure. Moreover, the nystagmus may be pendulous or jerky. The jerky nystagmus may be a slow acceleration phase jerk, a slow deceleration phase jerk, or a slow linear velocity jerk.The operator impairment detection system 10 performs an assessment of the gaze ystagmus of the driver 14 to determine HGN using one of three methods: a display assessment shown in FIGS. 2A and 2B, an audio assessment shown in FIG. 2C, and a passive assessment shown in FIG. 2D.As shown in FIGS. 2A and 2B, in the display evaluation, the display 18 is used to evaluate the HGN of the driver. In FIG. 2A, the display 18 is an HUD ranging from column to column. Once the display evaluation is initiated, the display 18 is used to project a graphic 50 onto the windshield 38. In FIG. 2B, the display 18 is a wide screen on a console of the vehicle 12. The graphic 50 may be accompanied by voice commands issued via the audio system 24, or may include graphic or linguistic instructions displayed on the display 18. The assessment includes instructing the driver 14 to fix a first point 52 and direct his eyes to a second point 54, each of which is displayed on the display 18. The first point 52 is associated with a fixed position in front of the driver 14. The second point 54 is associated with a maximum position of the eyes to the right of the driver 14. In one example, the graphic 50 allows the driver 14 to follow a visual prompt (e.g., a moving point) so that the speed and angle of gaze can be steered to achieve diagnostic conditions that increase the likelihood of accurate HGN assessment. In cases where a visual prompt may continuously move across the display 18, the driver 14 is prompted to follow a fixation point while maintaining his head pose. The movement of the visual prompt continues until sufficient data on eye position has been collected to make an assessment. The eye position is defined as the position of the eyes and head in space. The direction of gaze is defined as the direction in which the gaze is directed. The camera 20 captures the x-axis direction of the eyes of the driver 14 over time as the eyes track the fixation point as it moves from the first point 52 to the second point 54. If the controller 16 determines that the driver 14 does not correctly follow the instructions during the evaluation (e.g., does not keep the head quiet or rotates away from the camera 20), the controller 16 may issue correction instructions. The acquired data on the direction of the eyes over time is referred to as a gaze pattern. In an alternative embodiment, the driver 14 may be instructed to point his gaze and turn his head away from that point using visual or audible prompts.As shown in FIG. 2B, audio system 24 is used in audio assessment to assess the driver's HGN. Once the assessment is initiated, the audio system 24 issues voice commands to the driver 14 to fix his gaze to a first position 56 in the vehicle 12 and then move his gaze to a second position 58 in the vehicle 12. In the present example, the first position 56 is on an infotainment system 6 60 and the second position 58 is on a side mirror 62, however, it should be appreciated that other locations are also possible. The camera 20 captures the direction of the x-axis of the eyes of the driver 14 over time as the eyes move from the first position 56 to the second position 58. If the controller 16 determines that the driver 14 does not correctly follow the instructions during the evaluation (e.g., does not keep the head quiet or rotates away from the camera 20), the controller 16 may issue correction instructions. In an alternative embodiment, the driver 14 may be instructed to point his gaze to a particular position and turn his head using audio prompts.As seen in FIG. 2C, in the passive evaluation, the eye movements of the driver 14 are detected during the operation of the vehicle. The eye movement is tracked as the driver 14 fixes objects he is viewing through the windshield 38 with the eyes. For example, the camera 20 captures the x-axis direction of the driver's 14 eyes as they move from a first object 64 on the roadway to a second object 66 adjacent the roadway. The head and gaze vectors monitored by the driver monitoring system 26 may aid in the assessment.Once the gaze samples have been collected using an HCN evaluation as described above, the controller 16 runs the gaze samples through an HGN detection model to determine whether an HGN is present. In general, the HGN recognition model comprises two major parts: (1) a transformation of the representation and (2) an estimation using a Gaussian mixture model (GMM).The eye samples are taken at a fixed interval during eye movements. The gaze sample taken at time t i may be referred to as: (x, y, t) i, where x, y represents the horizontal and vertical gaze directions, respectively, and t represents time. In the transformation of the representation stage, the horizontal direction x is derived twice. The first derivative is the velocity and is denoted dx:The second derivative is the horizontal acceleration of the viewing direction and is referred to as ddx:However, the estimates of the derivatives in equations (1) and (2) may be noisy, so a smoother estimate is required or an approximation is desired. The new representation (dx, ddx, t) i is location invariant and therefore the GMM is used to estimate the HGN from the gaze sample. For example, when analyzing data in which the driver 14 has fixed his gaze to a single position, the data (x, y, t) i concentrates on an arbitrary position, while the data (dx, ddx, t) i concentrates on about (0.0), which is defined as a fixation of the gaze. Therefore, a Gaussian (normal distribution) with an average at (0,0) models the fixation gaze, and its distribution has the following form:Here, g is the normal distribution N(μ,Σ), and the parameters of a single Gaussian model are μ,Σ over space (dx, ddx). Overall, the parameters in equation (3) are: μ,Σ.In another example, where the driver 14 has a fixed gaze at two different objects and a saccade (a fast movement of the eye between fixation points) therebetween, the representation (x, y, t) i would show the two fixations and a single behavior with a temporal pattern (e.g., saccade) between the two fixations. Meanwhile, the change of the plot (dx, ddx, t) i connects the two fixes to a location (0, 0), and the temporal pattern is represented as a set of unconcentrated dots that are not located at the location (0, 0). Therefore, this distribution is not a normal distribution, but a plurality of Gaussian distributions are assumed. The Gaussian distributions are weighted, one Gaussian distribution models fixation (number 1 in equation (4)), and a second Gaussian distribution models saccade (number 2 in equation (4)):In this case, g 1 and g 2 are two normal distributions N(μ,Σ) with the parameters μ 1, Σ 1 and μ 2, Σ 2. Moreover, each Gaussian distribution has its own weight, denoted w 1 and w 2. Overall, the parameters in equation (4) are: w 1, μ 1, Σ 1, w 2, μ 2, Σ 2.In another example, the gaze patterns are modeled with more than two Gaussian distributions. In this general case, with a weighted mixture of Gaussian distributions with G Gaussian distributions, the distribution has the following form:The parameters in Equation (5) are: w 1, μ 1, Σ 1,..., w i, μ i, Σ i, ··,w G, μ G, Σ G. Therefore, when the gaze samples are composed of multiple behavioral patterns, the components of the GMM are divided between the behavioral patterns.This assignment is not predefined and takes place within the framework of the training of the GMM. The HGN model using the GMM determines whether micro-eye movements, i.e., a certain amount of nystagmus, are present at the time intervals.FIG. 3 illustrates the method 100 for detecting impairment of the driver 14 of the vehicle 12. The method 100 begins at step 102 when the driver 14 is approaching the vehicle 12. In this step, the vehicle 12 is off. The method 100 then proceeds to step 104.In step 104, the controller 16 determines whether the alcohol sensor 22 detects alcohol. Depending on which type of alcohol sensor 22 is used in the vehicle 12, the alcohol sensor 22 may detect alcohol (ethanol vapors) in the environment of the vehicle 12 and convert this alcohol concentration into an estimated blood alcohol concentration (BAK). Alternatively, using the infrared scan of the driver's 14 finger, the image material of the driver 14 is analyzed by the controller 16 and the estimated BAK is determined. If no alcohol is detected, the method proceeds to step 106 and the start controller 30 is activated. When the start controller 30 is activated, the driver 14 may start the vehicle 12. If alcohol is detected, the method 100 continues to step 108.In step 108, the controller 16 compares the estimated BAK to a BAK threshold. In one example, the BAK threshold is 0.8 per millet. However, the threshold value may be set lower or higher depending on regulations of local authorities. If the estimated BAK is below the BAK threshold, the method continues with step 106 and the start controller 30 is activated. However, if the estimated BAK is equal to or greater than the BAK threshold, the method continues to step 110 and the start controller 30 is deactivated. When the start controller 30 is deactivated, the driver 14 may not start the vehicle 12. The method 100 then proceeds to step 112.In step 112, the controller 16 determines whether the assessment of the gaze nystagmus has been initiated. The rating of the gaze nystagmus is initiated by the driver 14 via the HMI 28. Alternatively, the gaze nystagmus rating may be selected as a default setting or made dependent on the estimated BAK (e.g., an estimated BAK between 0.6 and 1.0 triggers the gaze nystagmus rating). If the gaze nystagmus assessment has not been initiated, the method 100 returns to step 110 and the launch controller 30 remains deactivated for a specified period of time or until the controller 16 detects entry of another driver 14 into the vehicle 12. If the gaze nystagmus assessment was initiated, the method 100 proceeds to step 114.In step 114, the assessment of the gaze nystagmus is performed. For the evaluation of the gaze nystagmus, either the display evaluation shown in FIG. 2A, the audio evaluation shown in FIG. 2B, or the passive evaluation shown in FIG. 2C is used. The method 100 then proceeds to step 116.In step 116, the driver monitoring system 26 tracks the eye positions of the driver 14 at time intervals using the camera 20 during the assessment of the gaze ystagmus. As already mentioned, the tracked gaze directions are formed from gaze samples that include the x-axis direction of the eyes of the driver 14 at certain time intervals. The method 100 then proceeds to step 118.In step 118, the controller 16 uses the driver monitoring system 26 to determine whether the driver 14 is performing the assessment of the gaze nystagmus correctly. The assessment of gaze nystagmus is performed incorrectly if a factor affects the accurate acquisition of gaze samples. For example, if the driver 14 moves his head during the evaluation, closes his eyes during the evaluation, or the eyes are not visible due to glasses or sunglasses, the gaze samples may be unreliable and the evaluation is not performed correctly. If the assessment is not performed correctly, the method returns to step 114 and the assessment of the gaze nystagmus is restarted. Optionally, the controller 16 may provide correction instructions via the display 18 and / or the audio system 24. If the assessment is performed correctly, the method 100 proceeds to subroutine 120.In the subroutine 120, the controller 16 determines whether a gaze nystagmus is present based on the gaze samples collected during the evaluation of the gaze nystagmus, and determines an intensity of the gaze nystagmus when present. As will be described in more detail below, the presence and intensity of the gaze ystagmus for each eye of the driver 14 is determined using the gaze ystagmus model. As mentioned above, the gaze nystagmus model transforms the gaze samples using two derivatives and estimates a GMM using the derivatives to determine whether micro-eye movements (i.e., a form of nystagmus) are present. The amplitude and frequency (i.e., intensity) of the micro-eye movements are determined from the gaze samples. In one embodiment, the controller 16 provides a numerical score based on the presence and intensity of the nystagmus for each eye. The method 100 then proceeds to step 122.In step 122, the controller 16 determines whether the presence and intensity of the nystagmus determined in subroutine 120 indicate that the driver 14 is impaired. For example, if a numerical score is assigned in subroutine 120, controller 16 compares the score to a threshold for impairment. The threshold value for the impairment is four in this example, but can be set lower or higher depending on the local laws. If the score is greater than or equal to the impairment threshold, the driver 14 is impaired and the method returns to step 110 and the start controller 30 is deactivated. If the score is below the impairment threshold, the driver 14 is not impaired and the method 100 continues to step 106 and the launch controller 30 is activated despite the detection of alcohol in step 108.Referring now to FIG. 4, the subroutine 120 for determining whether a gaze nystagmus is present based on the gaze samples collected during the assessment of the gaze nystagmus and for determining the intensity of the gaze nystagmus if present is described in more detail. The subroutine 120 begins with step 202 in which eye tracking data, including gaze samples, is imported from the driver monitoring system 26 into the controller 16. The subroutine 120 then proceeds to step 204.In step 204, the controller 16 determines whether micro-eye movements occur in the right or left eye during the time intervals associated with the complete eye movement during the assessment of the gaze nystagmus. The time intervals used in this step are associated, for example, with the complete movement of the eyes from the first point 52 to the second point 54 or from the first position 56 to the second position 58 during the gaze assessment. The controller 16 determines the presence of micro eye movements based on the gaze nystagmus model. If micro-eye movements are not detected for either eye, the subroutine 120 proceeds to step 206. If micro-eye movements are detected for one of the two eyes, the sub-program 120 proceeds to step 208.In step 208, the controller 16 determines an amplitude and frequency of the detected micro-eye movements for each eye using the gaze samples. The amplitude is defined as the difference between the x-axis positions of the eye during a slow phase of detected micro-eye motion. The frequency is defined as the number of detected micro-eye movements within a time interval. An example of a graph showing the HGN jerk with the slow phase of the linear velocity is shown in FIG. 5 and is labeled with reference numeral 300. The graph 300 shows the amplitude of micro-eye motion between times (s) and (f) during the slow phase. The amplitude may be an average of the amplitudes of the slow phases over eye movement. Returning to FIG. 4, the subroutine 120 then proceeds to step 210.In step 210, the controller 16 compares the amplitude and frequency detected during the full eye movement to a first amplitude threshold and a first frequency threshold, respectively. The first amplitude threshold and the first frequency threshold may be determined by local laws or governmental regulations. If the amplitude or frequency during the full eye movement is less than the first amplitude threshold and the first frequency threshold, respectively, the subroutine 120 proceeds to step 206. If the amplitude and frequency during the full eye movement are equal to or greater than the first amplitude threshold and the first frequency threshold, respectively, the subroutine 120 proceeds to step 212.In step 212, the controller 16 passes a score for each eye at which a nystagmus is detected and the intensity of which is above the thresholds. The subroutine 120 then proceeds to step 206.In step 206, the controller 16 determines whether micro-eye movements occur in the right or left eye during the time intervals associated with the maximum eye movement during the assessment of the gaze nystagmus. The time intervals used in this step relate, for example, to the eyes which are fixed to the first point 52 or the first position 56 during the gaze detection. The controller 16 determines the presence of micro eye movements based on the gaze nystagmus model. If micro-eye movements are not detected for either eye, the subroutine 120 proceeds to step 214. If micro-eye movements are detected for one of the two eyes, the sub-program 120 proceeds to step 216.In step 216, the controller 16 determines an amplitude and frequency of the detected micro-eye movements for each eye using the gaze samples at the maximum eye position. The amplitude is defined as the difference between the x-axis positions of the eye during a slow phase of detected micro-eye motion. The frequency is defined as the number of detected micro-eye movements within a time interval. The amplitude may be an average of the amplitudes of the slow phases over eye movement. The subroutine 120 then proceeds to step 218.In step 218, the controller 16 compares the amplitude and frequency at the maximum horizontal position of the eyes to a second amplitude threshold and a second frequency threshold, respectively. The second amplitude threshold and the second frequency threshold may be determined by local laws or governmental regulations. If the amplitude or frequency at the maximum eye position is less than the second amplitude threshold and the second frequency threshold, respectively, the subroutine 120 proceeds to step 214. If the amplitude and the frequency at the maximum eye position are equal to or greater than the second amplitude threshold and the second frequency threshold, respectively, the subroutine 120 proceeds to step 220.In step 220, the controller 16 passes a score for each eye at which a nystagmus is detected and the intensity of which is above the thresholds. The subroutine 120 then proceeds to step 214.In step 214, the controller 16 determines whether there are micro eye movements in the right or left eye before the eyes deviate forty-five degrees from the starting or mid position during the assessment of the gaze ystagmus. The forty-five degrees from the initial position of the eye correspond to a middle position between the initial position of the eye and the maximum eye position during the evaluation of the gaze ystagmus. The time intervals used in this step are associated with, for example, the movement of the eyes between the first point 52 and a midpoint between the first point 52 and the second point 54 or with the movement of the eyes between the first position 56 and a midpoint between the first position 56 and the second position 58 during the gaze assessment. The controller 16 determines the presence of micro eye movements based on the gaze nystagmus model. If micro-eye movements are not detected for either eye, the subroutine 120 proceeds to step 222. If micro-eye movements are detected for one of the two eyes, the sub-program 120 proceeds to step 224.In step 224, the controller 16 determines an amplitude and frequency of the detected micro-eye movements for each eye from the eye samples at 45 degrees. The amplitude is defined as the difference between the x-axis positions of the eye during a slow phase of detected micro-eye motion. The frequency is defined as the number of detected micro-eye movements within a time interval. The amplitude may be an average of the amplitudes of the slow phases over eye movement. The subroutine 120 then proceeds to step 226.In step 226, the controller 16 compares the amplitude and frequency before the 45-degree eye position to a third amplitude threshold and a third frequency threshold, respectively. The third amplitude threshold and the third frequency threshold may be determined by local laws or governmental regulations. If the amplitude or frequency before the 45-degree eye position is less than the third amplitude threshold and the third frequency threshold, respectively, the subroutine 120 proceeds to step 222. If the amplitude and frequency before the 45-degree eye position are equal to or greater than the third amplitude threshold and the third frequency threshold, respectively, the subroutine 120 proceeds to step 228.In step 228, the controller 16 provides a score for each eye at which a nystagmus is detected and the intensity of which is above the thresholds. The subroutine 120 then proceeds to step 222.In step 222, the controller 16 summarizes (i.e., adds) the individual results for each eye for each test to determine the overall result. The subroutine 120 then ends and proceeds to step 122 in FIG. 3.Steps 204, 206 and 214 may also be performed in a different order without departing from the scope of the present disclosure. Moreover, the first, second, and third amplitude thresholds may be the same, and the first, second, and third frequency thresholds may be the same, without departing from the scope of the present disclosure.The system 10 and method 100 of the present disclosure provide several advantages. First, the system 10 and method 100 enable false alarms due to alcohol sensor 22 malfunction or detection of chemicals in the air to be overcome. Second, the system 10 and method 100 may detect actual impairment due to nystagmus even when alcohol is not detected. Finally, the system 10 and method 100 provide repeatability, consistency, and increased accuracy in determining nystagmus.The description of the present disclosure is merely exemplary, and variations that do not depart from the gist of the present disclosure are intended to fall within the scope of the present disclosure. Such variations are not to be regarded as a departure from the spirit and scope of the present disclosure.

Claims

A system (10) for detecting impairment of a vehicle operator (14), the system comprising: an alcohol sensor (22); a camera (20); a launch controller (30); and a controller (16) electrically connected to the alcohol sensor (22), the camera (20), and the launch controller (30), the controller (16) including a processor (34) and a memory (36), the memory (36) including instructions such that the processor is programmed to: determine that alcohol has been detected using the alcohol sensor (22); disable the launch controller (30) based on the detection of alcohol; perform a gaze ytagmus assessment of the vehicle operator (14) using the camera (20); determining whether the vehicle operator (14) is impaired based on the assessment of the gaze ystagmus; and activating the launch controller (30) when the vehicle operator (14) is not impaired, wherein the processor (34) is further programmed to: perform the assessment of the gaze ystagmus of the vehicle operator (14) by tracking eye movement of the vehicle operator (14) and recording gaze samples over time during the eye movement, wherein the gaze samples include horizontal positions of the eyes during the eye movement of the vehicle operator (14) at time intervals; characterized in that the processor (34) is further programmed to: transform the gaze samples by deriving the horizontal positions of the eyes twice to determine a speed and an acceleration in each time interval, wherein the speed and the acceleration are approximated or smoothed; and estimating a Gaussian mixture model using the derivatives of the horizontal positions of the eyes to determine whether micro-eye movements exist during the time intervals and to determine an amplitude and frequency of the micro-eye movements in the time intervals.The system (10) of claim 1, wherein the processor (34) is further programmed to: perform the assessment of the gaze nystagmus of the vehicle operator (14) using at least one of the following methods: a display assessment using a display (18) in the vehicle (12), an audio assessment using an audio system (24) in the vehicle (12), and a passive assessment.The system (10) of claim 1, wherein the processor (34) is further programmed to: determine whether the driver (14) is impaired by passing the gaze samples through a gaze nystagmus model.The system (10) of claim 1, wherein the processor (34) is further programmed to: determine whether the driver (14) is impaired by determining whether micro-eye movements are present during the time intervals during the eye movement, determine whether micro-eye movements are present during the time intervals at a maximum horizontal position of the eyes, and determine whether micro-eye movements are present during the time intervals before forty-five degrees from a center during the eye movement.The system (10) of claim 4, wherein the processor (34) is further programmed to: determine whether the driver (14) is impaired by comparing the amplitude and frequency of micro-eye movements during eye movement to a first amplitude threshold and a first frequency threshold, respectively, and determine that a nystagmus is detected when the amplitude and frequency during eye movement are greater than the first amplitude threshold and the first frequency threshold, respectively.The system (10) of claim 5, wherein the processor (34) is further programmed to: determine whether the driver (14) is impaired by comparing the amplitude and frequency of micro-eye movements during the time intervals at the maximum horizontal position of the eyes to a second amplitude threshold and a second frequency threshold, respectively, and determine that a nystagmus is detected if the amplitude and frequency during the time intervals at the maximum horizontal position of the eyes are greater than the second amplitude threshold and the second frequency threshold, respectively.The system (10) of claim 6, wherein the processor (34) is further programmed to: determine whether the driver (14) is impaired by comparing the amplitude and frequency of micro-eye movements during the time intervals before forty-five degrees from the center during eye movement with a third amplitude threshold and a third frequency threshold, respectively, and determine that a nystagmus is detected when the amplitude and frequency during the time intervals before forty-five degrees from the center during eye movement are greater than the third amplitude threshold and the third frequency threshold, respectively.The system (10) of claim 7, wherein the processor (34) is further programmed to: assign a score for each eye based on whether a nystagmus was detected during the time intervals, and compare the score to a threshold for impairment; and determine that the driver (14) is impaired if the score exceeds the threshold for impairment.

Images