THEO DÕI TỶ LỆ MẮT THEO THỜI GIAN THỰC ĐỂ PHÁT HIỆN NGỦ GẬT SỬ DỤNG CÔNG NGHỆ THỊ GIÁC MÁY DỰA TRÊN ĐIỂM MỐC KHUÔN MẶT

Tóm tắt

The article proposes a real-time algorithm to detect the drowsiness
phenomenon based on a standard camera. Face landmarks are identified,
especially around the subject's eyes. On that basis, a scale function from
extracting scalar quantities of face landmarks is calculated representing realtime eye-opening and closing times. In addition, the article also investigated the
reliability of the solution used in three cases: changing light intensity, subject
using glasses, face tilt angle changes relative to the scanning direction of the
camera. The results of the article can be applied to vehicles to warn drivers of
falling asleep to help drive safer vehicle.