摘要
文章针对使用预处理对图像进行二值化的传统道路识别方法时,丢失大量有效的信息和识别效果差,介绍了一种新型车辆偏离车道预警系统。该系统利用灰度图像提取道路边界的灰度和梯度信息,并构造基于小块统计的目标函数,去除复杂环境下的各种噪声并且识别出车道线,判断车辆是否偏离车道从而提供报警信息。该系统结构简单、实时性高,能显著提高车辆行驶的主动安全性,减少事故的发生。
The traditional method of road identification uses pretreatment to make the image binarized, which leads to great loss of effective information and poor identification effect. In this paper, an objective function based on grid counting is constructed by usir;g the grayscale and gradient of road boundary in monochrome image. All kinds of irregular texture noises in complex condition can he eliminated effectively and the lane boundary can be identified through the method, which can offer the alarm information through judging whether the vehicle deviates from the lane. The presented system has high real-timeness, and it can improve the active safety of the vehicle obviously and reduce the accidents.
出处
《合肥工业大学学报（自然科学版）》
CAS
CSCD
北大核心
2012年第7期874-877,共4页
Journal of Hefei University of Technology：Natural Science
基金
国家自然科学基金资助项目(50676012)
关键词
灰度图像
边界识别
偏离
报警
monochrome image
boundary identification
deviation
warning
