基于FAST特征的智能安全帽静态报警方法的研究

Study of Static Alarm Method of Smart Helmet Using FAST Feature

目前有关智能安全帽大多只是关注工作人员是否佩戴安全帽和他们的异常操作,而对作业人员本身的异常状态则基本忽略,针对上述现状,文章提出了一种基于FAST特征的智能安全帽报警方法,检测佩戴智能安全帽人员的非正常状态(晕厥、猝死等静止状态)并进行报警。利用智能安全帽上搭载的摄像头采集图像数据,使用FAST特征检测算法快速提取特征,然后比较2幅图像的特征是否匹配,以此检测当前安全帽是否处于静止状态。实验结果表明,这种基于FAST特征检测的方法能够有效判断佩戴智能安全帽的人员是否处于静止状态,从而进行报警,在工人单独作业的情况下有较高的实用价值。

At present,most of the smart helmets are concerned about whether or not the staff members wear the helmet and their abnormal operation,while the abnormal state of the operator itself is ignored.In view of the above-mentioned situation,this paper proposes a smart helmet alarm method using FAST feature,which detects and alarms the abnormal state(syncope,sudden death,etc.)of the personnel wearing the smart helmet.Using the camera on the smart helmet to collect the image data,using the FAST feature detection algorithm to quickly extract the features,and then the method compares the features of two adjacent images to detect whether the current helmet is still or not.Experimental results demonstrate that the proposed method can effectively detect and alarm the person who wears the smart helmet is under abnormal condition.The method has higher practical value in the case of workers working independently.