摘要
为提高无人车的主动安全性能,对其环境感知技术进行了研究,提出了一种在结构化道路上基于单目视觉的汽车前方可行区域内的障碍物检测方法;在确定路面感兴趣区域时,提出了基于模糊聚类的边界跟踪检测算法,实现对车道线的识别;针对序列图像帧间差分法的不足,提出了一种基于帧间差分法的二次改进算法,在突出障碍物特征信息后,采用了静态单帧图像特征处理算法,确定出障碍物位置。实验结果表明,该方法能够有效地识别出车道标志线以及感兴趣区域内的障碍物信息。
To improve active safety performance ,the environmental perception technology of UV was studied ,an obstacle detection method based on monocular vision for travelable area ahead of car on the structured road was put forward .In determining the surface region of interest ,the boundary tracking detection algorithm based on fuzzy clustering was proposed to achieve lane recognition .In light of the shortcomings of inter-frame difference method ,a secondary algorithm based on inter-frame difference method was proposed ,which adopted a static single-frame image feature processing algorithm to determine the position of the obstacle after protruding obstacle feature information .The experimental results show that this method can effectively identify the lane marking and obstacle information in the region of interest .
出处
《机械工程与自动化》
2014年第2期57-58,共2页
Mechanical Engineering & Automation
关键词
单目视觉
感兴趣区域
模糊聚类
线性拟合
帧间差分法
环境感知
无人车
monocular vision
region of interest
fuzzy clustering
linear fitting
inter-frame difference method
environment perception
unmanned vehicle
