RESEARCH ON VEHICLE-BASED DRIVER STATUS/PERFORMANCE MONITORING;DEVELOPMENT, VALIDATION,AND REFINEMENT OF ALGORITHMS FOR DETECTION OF DRIVER DROWSINESS

摘要

This report summarizes the results of a 3-year research project to develop reliable algorithms for the detection of motor vehicle driver impairment due to drowsiness. These algorithms are based on driving performance measures that can potentially be computed on-board a vehicle during highway driving, such as measures of steering wheel movements and lane tracking. A principal objective of such algorithms is that they correlate highly with, and thus are indicative of, psychophysiological measures of driver alertness/fatigue. Additional objectives are that developed algorithms produce low false alarm rates, that there should be minimal encumbering of (interference with) the driver, and that the algorithms should be suitable for later field testing. This report describes driving simulation and other studies performed to develop, validate, and refine such algorithms.