Non-Invasive Driver Drowsiness Detection System

Drowsiness when in command of a vehicle leads to decline cognitive performance that affects driver behavior, potentially causing accidents. Drowsiness-related road accidents lead severe trauma, economic consequences, impact on others, physical injury and/or even death. Real-time and accurate drowsiness detection warnings systems are necessary schemes reduce tiredness-related driving accident rates. The research presented here aims at the classification drowsy non-drowsy states based respiration rate by non-invasive, non-touch, impulsive radio ultra-wideband (IR-UWB) radar. Chest movements 40 subjects were acquired for 5 m using lab-placed IR-UWB radar system, per minute was extracted from resulting signals. A structured dataset obtained comprising minute, age label (drowsy/non-drowsy). Different machine learning models, namely, Support Vector Machine, Decision Tree, Logistic regression, Gradient Boosting Extra Tree Classifier Multilayer Perceptron trained dataset, amongst which Machine shows best accuracy 87%. This provides ground truth verification assessment UWB be used effectively respiration.