摘要
二十一世纪,随着汽车技术的研究更加深入,汽车技术领域发展迅速,目前汽车技术已经到达智能化阶段,机器视觉技术在汽车辅助驾驶技术中占有重要地位,研发内容也不是传统的机电设备问题,目前已发展到先进的辅助传动系统(ADAS)研究。文章主要以对车载摄像头所拍摄图像的处理技术和车道线的提取、检测方法的研究。灰度化和平滑处理使原始图像中的其他线条干扰程度降低,降低噪点使得车道线轮廓更加明显。运用梯度计算边缘检测与非最大值抑制方法检测边缘,通过选择高阀值过滤边缘像素。得到车道线轮廓线后基于Hough的车道线检测原理,使用二元性的点和线,将原始图像给出的曲线转化为一个点在参数空间的曲线表达形式,解决曲线检测问题转化为参数空间中的最大值,且将整体特征的检测转换为本地检测的特点。
In the 21st century,with the more in-depth research of automobile technology,the field of automobile technology has developed rapidly.At present,automobile technology has reached the intelligent stage,machine vision technology plays an important role in automobile auxiliary driving technology,and the content of research and development is not the problem of traditional mechanical and electrical equipment,but has been developed to the research of advanced driver assistance system(ADAS).This paper mainly focuses on the processing technology of the image captured by the vehicle camera and the extraction and detection method of the lane line.Graying and smoothing reduce the interference of other lines in the original image,and reduce the noise to make the outline of the lane more obvious.The edge detection method of gradient calculation and non-maximum suppression method is used to detect the edge,and the edge pixels are filtered by selecting a high threshold.After the lane contour is obtained,based on the principle of lane detection based on Hough,the curve given by the original image is transformed into a curve expression of a point in the parameter space by using binary points and lines,the curve detection problem is transformed into the maximum value in the parameter space,and the overall feature detection is transformed into the characteristics of local detection.
出处
《科技创新与应用》
2020年第11期49-51,共3页
Technology Innovation and Application
基金
山东省教育厅高等学校科技计划(编号:J17KA068)
国家级大学生创新创业计划项目(编号:116882019044)。
关键词
图像处理
边缘检测
车道线检测
image processing
edge detection
lane detection
