基于混合核函数支持向量机的人眼疲劳检测
首发时间:2015-04-21
摘要:人眼疲劳检测是预防驾驶员疲劳驾驶的有效途径。为了提高人眼疲劳检测的正确率,提出了一种基于混合核函数支持向量机的人眼疲劳检测方法。Adaboost算法进行人脸检测,引入灰度投影法准确定位人眼。使用多项式核函数和径向基函数混合构造的核函数的支持向量机,确定最优混合系数,进行人眼睁开闭合程度量化。计算Perclos的值和眨眼频率,完成对人眼疲劳状态的检测。实验表明,该疲劳检测方法能够实时准确地检测人眼疲劳程度,具有一定实用性。
关键词: 疲劳检测 混合核函数 支持向量机 Perclos算法
For information in English, please click here
Eye Fatigue Detection Based On Mixed Kemel Function Of SVM
Abstract: Eye fatigue detection is an effective way to prevent driver fatigue driving. In order to improve the accuracy of detection of eye fatigue, it propose the method based on mixed unclear function of SVM for eye fatigue detection. Use Adaboost algorithm to face detection , and method of gray projection to accurately locate the human eye. Use polynomial kernel function and radial basis kemel function construction of mixed kemel function to support vector machine, and determine the optimal mix coefficient, to quantify the extent of eye open or closed. With Perclos calculate values and frequency of blink, completion of the human eye fatigue detection is finished. Experiments show that the fatigue detection method can accurately detect real-time eye fatigue. It is practical.
Keywords: Fatigue detection;Mixed Kemel Function;SVM;Perclos algorithm
基金:
论文图表:
引用
No.4637359104714914****
同行评议
共计0人参与
勘误表
基于混合核函数支持向量机的人眼疲劳检测
评论
全部评论0/1000