首页|基于夜间环境下的车道线检测

基于夜间环境下的车道线检测

扫码查看
针对夜间道路光线不足、图像亮度较暗、不容易检测车道线的问题,提出一种在夜间环境下的车道线检测方法。第一部分进行夜间图像增强:首先对灰度化的图像进行双重滤波降噪处理,通过比较不同区域的图像灰度值,使用局部直方图均衡化算法增强部分区域对比度,再采用拉普拉斯算子对图像进行增强,图像的纹理和细节得到有效的提升,然后最后通过伽马变换矫正整体图像。第二部分进行车道线边缘检测:对处理后的图像进行Canny算子边缘检测,拟合车道直线部分采用概率霍夫变换。结果表明,该方法具有较强的夜间车道线检测能力,为智能驾驶提供了安全保障。
Lane-line detection based on Night-time Environment
As the image is not very clear making it difficult to detect lane lines,a lane line detection algorithm in the night environment is proposed for the problem that the road is not well lit at night.The first part performs nighttime image enhancement.First,the detected image is double-filtered and noise-reduced,and the contrast of some areas is enhanced by comparing the values of different areas using a local histogram equalisation algorithm.Second,the image is enhanced by applying the Laplacian operator,the texture and details of the image are effectively enhanced.Finally,the overall image is corrected by using Gamma transformation.In the second part,the edge detection of the lane lines is performed:the processed image is edge detected by applying the Canny operator and the lane line part is fitted with the Progressive Probabilistic Hough Transform.The results show that the proposed algorithm has a strong ability to detect lane lines at night,which provides safety for intelligent driving.

image enhancementdouble filteringlane line detectionedge detection

曾丽娟、刘嘉炜、敖邦乾

展开 >

遵义师范学院物理与电子科学学院,贵州遵义 563006

遵义师范学院工学院,贵州遵义 563006

图像增强 双重滤波 车道线检测 边缘检测

遵义市科技计划项目遵义市科技计划项目遵义师范学院服务地方产业革命项目遵义师范学院基金贵州省大学生创新训练项目

遵市科合2022129号遵市科合HZ字2023163号遵师CXY[2023]8号遵师BS[2023]2号202310664057

2024

遵义师范学院学报
遵义师范学院

遵义师范学院学报

影响因子:0.165
ISSN:1009-3583
年,卷(期):2024.26(5)