Perclose Image Processing with HRV Signal for Hypnosis Highway Detection

Abstract

The driver's hypnosis detection system will be processed by image processing. The background and purpose of this research is the detection of street hypnosis to increase street awareness. The road hypnosis detection method used is based on eyelid closure and the detection of physiological parameters of the driver's HRV signal to improve the accuracy of detecting the occurrence of hypnosis. This study was to detect a decrease in HRV signals related to the condition of the driver being asleep or awake using the facial landmark method. The initial process begins with streaming the camera with a webcam mounted on the vehicle's dashboard. Detection of the facial area uses an eye aspect ratio algorithm that is processed in the process of processing images captured by a webcam, then an eye ratio algorithm is used to report sleepy eyes in the form of output. The prediction for the first test data is very close to the actual value (0.5567 vs 0.58), which suggests that the model can perform the prediction quite well.