Detection and controlling of drivers' visual focus of attention

Partha Pratim Debnath, A. F. M. Rashidul Hasan,Dipankar Das

2017 International Conference on Electrical, Computer and Communication Engineering (ECCE)(2017)

引用 2|浏览1
暂无评分
摘要
The paper presents an approach to detect and control the focus of attention of the driver using his/her eye gaze and head movement direction. If the driver changes his/her eye gaze, the corresponding coordinate of the pupil also changes. To determine the eye gaze direction, we first divide the detected eye area into different regions corresponding to different target objects. Then the gaze direction is detected based on in which region the coordinate of pupil is located in the eye area. We also determine and classify the head movement direction into three areas: Central Field of View (CFV), Near Peripheral Field of View (NPFV), and Far Peripheral Field of View (FPFV). Finally, the combination of eye gazes and head movement areas are used to determine the driver's focus of attention. We detect both transient and sustained attention. If the driver changes his/her direction of attention abruptly, in such a situation we cannot always confirm that the driver has a new attention to another direction because it may be a transient attention. To detect a sustained attention we need to wait few moments (in experiment we set it to 8 seconds). We have tried to detect the transient focuses with different duration as well as the sustained focus of attention with optimal accuracy that arises while real driving. If the system detects the sustained focus of attention to other direction than driving, we generate a controlling signal to return his/her focus of attention to driving. We have implemented and tested the proposed system in both controlled and real environment. The experimental results reveal that the proposed system is able to detect and control the driver's focus of attention in real driving situation.
更多
查看译文
关键词
Visual focus of attention,eye center localization,gaze detection,sustained attention
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要