多特征结合的疲劳驾驶检测方案设计与FPGA验证OA
Design and FPGA Verification of Fatigue Driving Detection Scheme Based on Multi-feature Combination
随着社会经济的高速发展,人们对出行的需求和车辆的依赖正在不断增长,道路交通的出行体量也随之逐渐增大,产生了各种交通问题,安全驾驶在当下显得愈发重要.基于此,面向物联网中的车联网疲劳驾驶检测场景,使用计算机视觉技术,依据面部检测得到的面部特征,以及头部姿态估计提供的头部特征,研究设计多疲劳特征结合的软硬件协同疲劳驾驶检测方案,有效降低了误报率.将所设计的方案在现场可编程门阵列(Field Programmble Gate Array,FPGA)上验证,检测准确率约为 96%,硬件实测结果验证了设计方案的有效性和可行性.
With the rapid development of the social economy,people's demand for travel and reliance on vehicles continue to grow,leading to an increasing volume of road traffic.This has given rise to various traffic problems,making driving safety increasingly critical.In view of this,targeting the scenario of fatigue driving detection for internet of vehicles within internet of things,computer vision technology is used,and based on facial features obtained from facial detection and head features provided by head pose estimation,a software and hardware collaborative fatigue driving detection scheme that combines multiple fatigue features is investigated and designed,which effectively reduces the false alarm rate.The proposed scheme is ultimately verified on Field Programmable Gate Array(FPGA),and the detection accuracy reaches approximately 96%,with hardware test results confirming its effectiveness and feasibility.
林粤伟;宋丹;张涛
青岛科技大学信息科学技术学院,山东 青岛 266061海信视像科技股份有限公司,山东 青岛 266500青岛科技大学信息科学技术学院,山东 青岛 266061
信息技术与安全科学
现场可编程门阵列疲劳驾驶检测计算机视觉软硬件协同物联网
FPGAfatigue driving detectioncomputer visionsoftware and hardware collaborationinternet of things
《无线电工程》 2026 (2)
242-252,11
评论