AdaBoost视觉算法的士兵疲劳预警系统研究  

Research on Soldier Fatigue Early Warning System Based on AdaBoost Visual Algorithm

作　　者：任琳 岳光 郭保全[2] REN Lin;YUE Guang;GUO Bao-quan(Department of Electronic Engineering,Taiyuan Institute Of Technology,Taiyuan 030008,China;Research Institute of Collaborative Innovation of Military-Civilian Integration,North University of China,Taiyuan 030051,China)

机构地区：[1]太原工业学院电子工程系,山西太原030008 [2]中北大学军民融合协同创新研究院,山西太原030051

出　　处：《机械工程与自动化》2021年第2期40-42,45,共4页Mechanical Engineering & Automation

基　　金：山西省重点研发计划(指南)项目(201603D121040-1)。

摘　　要：针对我军士兵在高强度的自行武器驾驶军事训练中疲劳导致的装备故障多发,提出了一种基于AdaBoost视觉算法的士兵面部疲劳检测系统。该研究通过机器视觉检测士兵的疲劳状态,采用嵌入式平台对视觉面部图像信息处理并定位人眼,通过AdaBoost算法及Haar-like特征对人眼特征进行疲劳度检测并作出预警,及时更换士兵以提高自行武器的精确驾驶。实验结果验证了算法的有效性和合理性,对提高我军自行武器驾驶人员的精确驾驶和人身安全具有一定的实用价值。Aiming at the frequent occurrence of equipment failures caused by fatigue of our soldiers during the high-intensity self-propelled weapon driving military training,a soldier facial fatigue detection system based on the AdaBoost vision algorithm is proposed.The research uses machine vision to detect the fatigue status of soldiers,and uses an embedded platform to process visual facial image information and locate human eyes.The AdaBoost algorithm and Haar-like feature are used to detect the fatigue of human eyes and provide early warning,and replace soldiers in time to improve the accurate driving of self-propelled weapons.The experimental results verify the effectiveness and rationality of the algorithm,and have certain practical value for improving the precise driving and personal safety of our army's self-propelled weapon drivers.

关 键 词：面部疲劳预警系统 AdaBoost视觉算法 HAAR-LIKE特征 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]