基于ARM11的单目视觉车距监测系统  

Monocular Vision Car Distance Monitoring System Based on ARM11

作　　者：郗瑶颖[1] 张建阳[1] 赵祥模[1] 徐志刚[1] 

机构地区：[1]长安大学信息工程学院,西安710064

出　　处：《计算机系统应用》2012年第12期33-37,共5页Computer Systems & Applications

基　　金：国家高技术研究发展计划(863)(2009AA11Z203)

摘　　要：驾驶员行车时易出现视觉疲劳和失速现象,此时难以判断安全车间距,易导致车辆追尾事故.针对这一问题,以ARM11器件SAMSUNG S3C6410为核心搭建了安全车距监测系统;结合车辆的形态和纹理等特征,采用多特征融合技术完成复杂环境下的车辆识别,并根据单目视觉测距原理完成摄像机多角度下的车距检测.实验结果表明,该系统能有效监测前车车距,测量准确、可靠性高且满足实时性要求.Driver while driving prone to visual fatigue and stall phenomena, at this time it i~ difficult to judge the safety car distance, easily lead to rear-end collision accident. In order to solve this problem, uses the ARM11 core SAMSUNG S3C6410 devices build a car safety distance monitoring system. Combine of vehicle features such as shape and texture, system uses multi-feature fusion technology to complete the vehicle identification under complex environment, uses the monocular vision principle complete the car distance detection with the camera under multi-angle. The experiments show that this system can effective monitoring the front cars distance, accurate measurement, high reliability and meet the requirements of real-time.

关 键 词：安全车距监测 单目视觉 多特征融合 SAMSUNG S3C6410 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置] TP391.41[自动化与计算机技术—控制科学与工程]