多源信息融合在驾驶疲劳检测中的应用

疲劳驾驶已成为引起交通事故的主要原因之一。目前众多驾驶疲劳检测方法都是通过单一的基于图像处理技术实现对驾驶疲劳的识别检测,而这种方法易受驾驶环境的影响,限制了其检测的准确性和可靠性。针对这一局限性,引入多源信息融合技术,选择基于图像的PERCLOS值以及基于非图像的方向变化与驾驶员反应不一致情况、方向盘动作状态和连续驾驶时间作为融合参数,并采用TS模糊神经网络(TSFNN)进行综合判断的方法,提高了驾驶疲劳检测的准确性和可靠性。经过实验表明该方法有一定的有效性。

MULTI-SOURCE INFORMATION FUSION APPLICATION TO DRIVING FATIGUE DETECTION

Driving fatigue has become one of the major causes of traffic accidents.At present most driving fatigue detection methods are realized simply based on the image processing technology to realize driving fatigue recognition and detection.However,due to its vulnerability to driving environments,the accuracy and reliability of the detection of the method is limited.In view of this limitation,the paper introduces a multi-source information fusion technology,chooses circumstances of unconformity between image-based PERCLOS values as well as non-image-based direction changes and driver reflections,steer wheel action status and continuous driving time as fusion parameters,and utilizes TSFNN for comprehensive judgement in order to improve the accuracy and reliability of driving fatigue detection.Experiments have been carried out to prove that the proposed method is considerably effective.