画像処理によるドライバの運転状態の監視 : 個人差に配慮したドライバの意識状態の検知

概要

論文の詳細を見る
We have developed a car driver monitoring system for measuring drivers consciousness, with which we aim to reduce car accidents caused by drowsiness of drivers. The system consists of the following three subsystems: an image capturing system with a pulsed infrared CCD camera, a system for detecting blinking waveform by the images using a neural network with which we can extract images of face and eye areas, and a system for measuring drivers consciousness analyzing the waveform with a fuzzy inference technique and others. The third subsystem extracts three factors from the waveform first, and analyzed them with a statistical method, while our previous system used only one factor. Our experiments showed that the three-factor method we used this time was more effective to measure drivers consciousness than the one-factor method we described in the previous paper. Moreover, the method is more suitable for fitting parameters of the system to each individual driver.
社団法人電気学会の論文
2006-02-01