Search Results

Vehicle safety systems may use occupant physiological information, e.g., occupant heights and weights to further enhance occupant safety. Determining occupant physiological information in a vehicle, however, is a challenging problem due to variations in pose, lighting conditions and background complexity. In this paper, a novel occupant height estimation approach is presented. Depth information from a depth camera, e.g., Microsoft Kinect is used. In this 3D approach, first, human body and frontal face views (restricted by the Pitch and Roll values in the pose estimation) based on RGB and depth information are detected. Next, the eye location (2D coordinates) is detected from frontal facial views by Haar-cascade detectors. The eye-location co-ordinates are then transferred into vehicle co-ordinates, and seated occupant eye height is estimated according to similar triangles and fields of view of Kinect.

Driver state sensing technologies start to be widely used in vehicular systems developed from different manufacturers. To optimize the cost and minimize the intrusiveness towards driving, majority of these systems rely on in-cabin camera(s) and other optical sensors. With their great capabilities of detecting and intervening driver distraction and inattention, these technologies might become key components in future vehicle safety and control systems. However, currently there are no common standards available to compare the performance of these technologies, thus it is necessary to develop one standardized process for the evaluation purpose.

A vehicle integrated sensing and analysis system has been designed, implemented, and demonstrated to nonintrusively monitor several biological signals of the driver. The biological driver signals measured by the system are the heart electrical signals or pseudo Lead-I electrocardiography (pLI-ECG), the galvanic skin response (GSR) or electrical conductance measured from the driver's fingers to palm, the palm skin temperature, the face skin temperature, and the respiration rate. The pLI-ECG and GSR measurements are made through direct contact of the driver hands with stainless steel electrodes integrated in the steering wheel rim. The temperature measurements are made with non-contacting infrared temperature sensors, also located on the steering wheel. The respiration rate was measured using a flexible thin film piezoelectric sensor affixed to the seatbelt.

An assessment tool was developed for rollover CAE signals evaluation to assess primarily the qualities of CAE generated sensor waveforms. This is a key tool to be used to assess CAE results as to whether they can be used for algorithm calibration and identify areas for further improvement of sensor. Currently, the method is developed using error estimates on mean, peak and standard deviation. More metrics, if necessary, can be added to the assessment tool in the future. This method has been applied to various simulated signals for laboratory-based rollover test modes with rigid-body-based MADYDO models.

This paper presents an image analysis of a laboratory-based rollover crash test using camera-matching photogrammetry. The procedures pertaining to setup, analysis and data process used in this method are outlined. Vehicle roll angle and rate calculated using the method are presented and compared to the measured values obtained using a vehicle mounted angular rate sensor. Areas for improvement, accuracy determination, and vehicle kinematics analysis are discussed. This paper concludes that the photogrammetric method presented is a useful tool to extract vehicle roll angle data from test video. However, development of a robust post-processing tool for general application to crash safety analysis requires further exploration.

A number of studies, using data from NASS-CDS, have shown a large percentage of rollover crashes can be classified as tripped events. In many cases, the requirements for a tripped rollover detection algorithm are driven by the timely activation of an occupant containment device. To meet these requirements rollover detection algorithms have been developed by utilizing vehicle roll rate, lateral and vertical accelerations data collected primarily from laboratory tests. This study identifies and examines several challenges associated with developing a rollover detection algorithm with enhanced capabilities. Enhancement of the detection algorithm is explored by considering additional vehicle responses: forward velocity and sideslip angle. With the additional signals, discrimination of rollover crashes from other crash modes is discussed. Potential field/laboratory test modes are proposed to generate the additional vehicle signals.

State of the art electronic restraint systems rely on the acceleration measured during a vehicle crash for deployment decisions. The acceleration signal is analyzed with different criteria, among which the velocity change is a dominant criterion in almost any existing crash detection algorithm. Sensors in the front crush zone have recently been added to help develop restraint systems that comply with the new FMVSS208 and EuroNCAP regulations. Front crash sensors are usually evaluated for their velocity change during a crash and typically play a key role in the deployment decision. CAE based FEA analysis has recently been used to generate signals at the sensor module locations in crash simulations to provide supplemental information for crash sensing algorithm development and calibration. This paper presents an initial effort in developing a velocity-based crash detection algorithm, that allows broad use of CAE generated velocity time histories for system calibration.