基于多特征模糊融合的疲劳状态判决

在基于视觉的疲劳驾驶识别过程中,使用单个特征进行疲劳驾驶判断,常常会受到非普遍适用性、噪声等因素的影响,从而导致识别率降低。为了解决眼部特征参数或者是嘴巴特征参数单个特征识别率低甚至特使环境无法识别的问题,提出一种基于多特征融合的判决方法,利用了各种特征之间的优势互补,可以降低噪声和类内类间差异的影响,从而提高系统的性能,并且能增强适用性。最后通过实验结果证明,使用眼部和嘴巴特征融合的方法比单一的判决方法准确率更高。

Decision of Fatigue State Based on Characteristics' Fuzzy Fusion

The fatigue recognition process based on visual,using a single feature to fatigue driving, often by the universal applicability,the influence of factors such as noise,leading to the recognition rate is reduced. Characteristic parameters in order to solve the eye or mouth feature parameters of the individual character recognition rate is low and even envoy environment cannot be any other questions, this paper proposes a decision method based on feature fusion,using a variety of characteristics between the complementary advantages,can reduce the effects of difference between noise and class in the class,so as to improve the performance of the system,and can enhance the applicability. Finally, the experimental results show that using the eye and mouth features fusion method is higher accuracy than a single sentence.