Limits...
Optical flow and driver's kinematics analysis for state of alert sensing.

Jiménez-Pinto J, Torres-Torriti M - Sensors (Basel) (2013)

Bottom Line: Our experiments show the effectiveness of the approach with a correct eyes detection rate of 99.41%, on average.The results obtained with the proposed approach in an experiment involving fifteen persons under different levels of sleep deprivation also confirm the discriminability of the fatigue levels.In addition to the measurement of fatigue and drowsiness, the pose tracking capability of the proposed approach has potential applications in distraction assessment and alerting of machine operators.

View Article: PubMed Central - PubMed

Affiliation: Department of Electrical Engineering, Pontificia Universidad Catòlica de Chile, Vicuña Mackenna 4860, Casilla 306-22, Santiago, Chile. jejimenp@puc.cl

ABSTRACT
Road accident statistics from different countries show that a significant number of accidents occur due to driver's fatigue and lack of awareness to traffic conditions. In particular, about 60% of the accidents in which long haul truck and bus drivers are involved are attributed to drowsiness and fatigue. It is thus fundamental to improve non-invasive systems for sensing a driver's state of alert. One of the main challenges to correctly resolve the state of alert is measuring the percentage of eyelid closure over time (PERCLOS), despite the driver's head and body movements. In this paper, we propose a technique that involves optical flow and driver's kinematics analysis to improve the robustness of the driver's alert state measurement under pose changes using a single camera with near-infrared illumination. The proposed approach infers and keeps track of the driver's pose in 3D space in order to ensure that eyes can be located correctly, even after periods of partial occlusion, for example, when the driver stares away from the camera. Our experiments show the effectiveness of the approach with a correct eyes detection rate of 99.41%, on average. The results obtained with the proposed approach in an experiment involving fifteen persons under different levels of sleep deprivation also confirm the discriminability of the fatigue levels. In addition to the measurement of fatigue and drowsiness, the pose tracking capability of the proposed approach has potential applications in distraction assessment and alerting of machine operators.

Show MeSH

Related in: MedlinePlus

Snapshot of the system running.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3673081&req=5

f5-sensors-13-04225: Snapshot of the system running.

Mentions: Visual tracking in 3D space of any object using a single-camera view is a challenging problem, because depth information is lost due to the perspective projection that maps scene points in 3D space onto image points in the 2D sensor plane of the camera. However, whenever some knowledge of the object's geometry and size is available a priori, it is often possible to recover 3D motion and pose information. The proposed driver tracking approach takes advantage of the fact that all salient points of the SPG belong to the driver's head, which for the purpose of the proposed approach, can be regarded as a rigid object of standard size located at a regular nominal distance from the camera. By rigid object it is meant that the skull as a structuring element is non-deformable. Hence, the eyes, the cheek bones and the nose preserve their relative distances with respect to each other. It is to be noted that locally around the eyes and mouth, the face is a deformable (non-rigid) object that changes when the driver talks or makes gestures. However, unlike our prior work [44], here, we are not considering gestures, such as eyebrow raisings or yawning, as the small deviations of SPG points around the mouth can be handled correctly by the Lucas-Kanade tracker. Therefore, for the purpose of the proposed approach, changes in gesture can be neglected, and the SPG can be treated as a set of salient points that can be consistently tracked and that preserve their relative distance in 3D space, as illustrated in Figure 3 and shown for a real driver in Figure 5. By rigid object we do not mean the head is motionless or rigidly fixed. On the other hand, while there do exist correlations between the morphological characteristics of bones and the population that tend to follow geographic boundaries often coinciding with climatic zones, the size of the head changes little across different populations and phenotypes for people 18 years or older (see, for example, [58]). The average male head is around 20 × 15 ± 2.2 × 0.7 cm, while the average female head is 18 × 13 ± 1.2 × 1.2 cm. This ±3 cm variation relative to the camera-head operating distance is negligible. Because of this reason, it is possible to consider the head as an object of standard size, that changes little from one driver to another, and due to its low variance, its size is very predictable. Furthermore, an important feature of the proposed approach is that it does not require the heads to be exactly equal, because the SPG is created on-line for each person.


Optical flow and driver's kinematics analysis for state of alert sensing.

Jiménez-Pinto J, Torres-Torriti M - Sensors (Basel) (2013)

Snapshot of the system running.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3673081&req=5

f5-sensors-13-04225: Snapshot of the system running.
Mentions: Visual tracking in 3D space of any object using a single-camera view is a challenging problem, because depth information is lost due to the perspective projection that maps scene points in 3D space onto image points in the 2D sensor plane of the camera. However, whenever some knowledge of the object's geometry and size is available a priori, it is often possible to recover 3D motion and pose information. The proposed driver tracking approach takes advantage of the fact that all salient points of the SPG belong to the driver's head, which for the purpose of the proposed approach, can be regarded as a rigid object of standard size located at a regular nominal distance from the camera. By rigid object it is meant that the skull as a structuring element is non-deformable. Hence, the eyes, the cheek bones and the nose preserve their relative distances with respect to each other. It is to be noted that locally around the eyes and mouth, the face is a deformable (non-rigid) object that changes when the driver talks or makes gestures. However, unlike our prior work [44], here, we are not considering gestures, such as eyebrow raisings or yawning, as the small deviations of SPG points around the mouth can be handled correctly by the Lucas-Kanade tracker. Therefore, for the purpose of the proposed approach, changes in gesture can be neglected, and the SPG can be treated as a set of salient points that can be consistently tracked and that preserve their relative distance in 3D space, as illustrated in Figure 3 and shown for a real driver in Figure 5. By rigid object we do not mean the head is motionless or rigidly fixed. On the other hand, while there do exist correlations between the morphological characteristics of bones and the population that tend to follow geographic boundaries often coinciding with climatic zones, the size of the head changes little across different populations and phenotypes for people 18 years or older (see, for example, [58]). The average male head is around 20 × 15 ± 2.2 × 0.7 cm, while the average female head is 18 × 13 ± 1.2 × 1.2 cm. This ±3 cm variation relative to the camera-head operating distance is negligible. Because of this reason, it is possible to consider the head as an object of standard size, that changes little from one driver to another, and due to its low variance, its size is very predictable. Furthermore, an important feature of the proposed approach is that it does not require the heads to be exactly equal, because the SPG is created on-line for each person.

Bottom Line: Our experiments show the effectiveness of the approach with a correct eyes detection rate of 99.41%, on average.The results obtained with the proposed approach in an experiment involving fifteen persons under different levels of sleep deprivation also confirm the discriminability of the fatigue levels.In addition to the measurement of fatigue and drowsiness, the pose tracking capability of the proposed approach has potential applications in distraction assessment and alerting of machine operators.

View Article: PubMed Central - PubMed

Affiliation: Department of Electrical Engineering, Pontificia Universidad Catòlica de Chile, Vicuña Mackenna 4860, Casilla 306-22, Santiago, Chile. jejimenp@puc.cl

ABSTRACT
Road accident statistics from different countries show that a significant number of accidents occur due to driver's fatigue and lack of awareness to traffic conditions. In particular, about 60% of the accidents in which long haul truck and bus drivers are involved are attributed to drowsiness and fatigue. It is thus fundamental to improve non-invasive systems for sensing a driver's state of alert. One of the main challenges to correctly resolve the state of alert is measuring the percentage of eyelid closure over time (PERCLOS), despite the driver's head and body movements. In this paper, we propose a technique that involves optical flow and driver's kinematics analysis to improve the robustness of the driver's alert state measurement under pose changes using a single camera with near-infrared illumination. The proposed approach infers and keeps track of the driver's pose in 3D space in order to ensure that eyes can be located correctly, even after periods of partial occlusion, for example, when the driver stares away from the camera. Our experiments show the effectiveness of the approach with a correct eyes detection rate of 99.41%, on average. The results obtained with the proposed approach in an experiment involving fifteen persons under different levels of sleep deprivation also confirm the discriminability of the fatigue levels. In addition to the measurement of fatigue and drowsiness, the pose tracking capability of the proposed approach has potential applications in distraction assessment and alerting of machine operators.

Show MeSH
Related in: MedlinePlus