基于视觉注意机制的高速公路路侧景观对累积驾驶疲劳的影响研究

项目名称： 基于视觉注意机制的高速公路路侧景观对累积驾驶疲劳的影响研究

项目编号： No.51308251

项目类型： 青年科学基金项目

立项/批准年度： 2014

项目学科： 建筑科学

项目作者： 王琳虹

作者单位： 吉林大学

项目金额： 25万元

中文摘要： 在高速公路行车过程中，机动车驾驶员长时间被动地受路侧景观所施加的累积影响，容易导致注意力分散、反应速度变慢等疲劳特性的变化，这些变化均可以体现在驾驶员的视觉特征参数中，进而影响其对车辆操纵行为，引发或避免交通事故。合理的道路景观可以提高驾驶员注意力，有效缓解驾驶疲劳、提高交通安全水平，反之将有可能导致交通安全水平的下降。因此，本项目基于驾驶员视觉注意机制，研究高速公路路侧景观对驾驶员疲劳特性的影响具有重要意义。首先基于图像处理技术、模式识别算法及聚类分析等进行高速公路路侧景观的提取及量化；之后利用驾驶员的注意特性、执行能力、反应能力统一量化累积驾驶疲劳；最后基于视觉搜索模式分析、视觉流线分析等方法研究驾驶员的视觉注意机制，以驾驶员视觉注意机制为桥梁，建立量化后的景观与累积驾驶疲劳的关系模型，并确定驾驶疲劳阈值。研究成果将丰富交通安全理论体系，为基于驾驶疲劳的高速公路景观设计提供理论依据。

中文关键词： 高速公路；路侧景观；疲劳判别；驾驶行为；影响规律

英文摘要： During the rapid movements on freeway, drivers of motor vehicles are affected by the roadside landscape passively and accordingly their attention and reaction features are changed. These changes can be reflected in drivers' visual parameters, and would pay impact on the manipulation behaviors of drivers and cause or prevent traffic accidents. Rational roadside landscape can alleviate driver's fatigue effectively and improve the driver's attention and traffic safety. Otherwise, the traffic safety may be decreased. Thus, this project studies the impact of landscape color on driver's fatigue based on drivers' visual attention mechanism. First, the features of freeway roadside landscape are extracted and quantified based on image processing technology, pattern recognition algorithms and clustering analysis. Then the attention characteristics, executive capacity and reaction ability are selected and they are used in the quantification of cumulative driving fatigue.At last, the drivers' visual attention mechanism based on the analysis of visual search mode and visual flow lines is analyzed. Selecting visual attention mechanism as the bridge to build the model between landscape and cumulative driving fatigue, and identify the threshold of cumulative driving fatigue. The results of this study will rich the traffic safet

英文关键词： freeway；roadside landscape；fatigue identification；driver behavior；affect law