基于视觉的车道线检测识别算法研究.pdfVIP

基于视觉的车道线检测识别算法研究 摘要 随着我国经济社会的发展，人们的出行方式更多的选择了自主车辆驾驶， 随着 便捷快速出行的同时，交通事故的发生也给人们的生命财产带来巨大的损失。据统 计，在所有交通事故起因中，占据将近三分之一的因素为疲劳驾驶以及注意力不集 中。究其原因，我们发现是由此引发的车道线偏离，汽车在正常行驶过程中偏离正确 轨道。 为了减少由于车道线偏离所造成的交通事故，我们想实现车道线偏离预警，而最 关键的环节就是车道线检测识别算法的设计研究。该研究最主要的环节则是车道线的 提取与识别，即在图像背景下检测识别出车道线与行驶安全区域，以检测车辆的行驶 状态。当车辆发生偏离预定车道线的情况时，能够及时提醒驾驶员，改变行车状态， 从而避免不必要的交通事故发生。 本文通过车道线模型，完成了车道线的检测识别，主要研究内容包括：道路图像 的预处理、图像的滤波、图像的阈值分割、边缘检测、霍夫变换的车道检测算法、感 兴趣算法对车道线的跟踪、高斯混合模型预测车道检测准确性。 关键词 车道线检测，霍夫变换，图像处理，边缘检测 Research on detection and recognition of lane based on vision Abstract With the development of economy and society, the way people travel tend to more choices of autonomous driving, along with a fast and convenient travel at the same time, traffic accidents also caused huge losses to peoples lives and property . According to statistics, in the cause of all accidents, accounting for nearly one-third of the factors driving fatigue and inattention. The reason we found the consequent lane deviation during normal driving the car deviated from the right track. In order to reduce accidents due to the lane deviation caused, we want to achieve lane departure warning, and the most critical aspects of the design study is to detect lane recognition algorithm. The most important aspect of the study is to e tract and identify the lane line is detected to identify the lane and driving safety area in the background of the image in order to detect a vehicle traveling state. When it deviates from a predetermined lane of the vehicle occurs, timely to remind the driver to change the driving state, thereby avoiding unnecessary traffic accidents. By the lane line model, finished lane detection and identification, the main contents include: Threshold road image pre