摘要
传统车道线图像识别与现实路况差异较大,提出了一种车道线实时视频图像识别方法。方法对车道线图片进行逆透视转变,采用改进的Sobel算子对最近车道线边缘进行检测。在此基础上利用自适应双阈值对检测到的图像做二值化处理,增强图像信息。使用改进Hough变换算法减少拍摄图像与实际路况的差异,完成车道线实时视频图像识别。仿真结果证明,研究方法可以在不同道路环境中实现车道线视频图像准确识别,运行时间短、效率高。
The traditional lane line image recognition is quite different from the actual road condition. In this paper, a method recognizing the real-time image of lane line was put forward. This method transformed the image of the lane line by inverse perspective transformation, and then the improved Sobel operator was used to detect the edge of the nearest lane line. On this basis, an adaptive double threshold was used to binarize the detected image and thus to enhance the image information. The improved Hough transform algorithm was used to reduce the difference between the captured image and the actual road condition, and thus to complete the real-time video image recognition of lane line. Simulation results show that the proposed method can realize accurate recognition of video images of lane lines in different road environments. In addition, this method has a short running time and high efficiency.
作者
王敬
WANG Jing(Minnan Science and Technology College,Fujian Normal University,Quanzhou Fujian 362332,China)
出处
《计算机仿真》
北大核心
2021年第5期116-120,共5页
Computer Simulation
基金
福建省教育厅中青年教师教育科学项目(科技类)(2018年度)(JT180832)。
关键词
车道偏离预警
自适应双阈值
图像去噪增强
车道线检测
Lane departure warning
Adaptive double threshold
Image denoising enhancement
Lane detection
作者简介
王敬(1986-),女(汉族),河北沧州人,硕士,讲师,研究方向:无线通信系统研究,图像处理。
