摘要
为提高车道线检测的准确性以增强无人驾驶车辆的安全驾驶性能,在传统车道线检测的边缘提取、霍夫变换、颜色空间阈值提取、透视变换等方法的基础上,利用深度学习技术,提出一种基于改进FCN的车道线检测网络模型。该模型能够准确提取出车道线的特征信息,并在车道线检测数据集上进行模型训练,以评估该车道线检测网络的性能。通过实验对比,结果表明改进FCN模型在检测精度上比传统FCN网络模型提高了1%,具有良好的分割有效性。
In order to improve the accuracy of lane line detection and enhance the safe driving performance of driverless vehicles, a lane line detection network model based on improved FCN is proposed by using deep learning technology on the basis of traditional lane detection methods such as edge extraction, Hough transform, color space threshold extraction and perspective transform. The model can accurately extract the feature information of the lane line, and train the model on the lane line detection data set to evaluate the performance of the lane line detection network. Experimental comparison shows that the detection accuracy of the improved FCN model is 1% higher than that of the traditional FCN network model, and it has good segmentation effectiveness.
作者
杨莹
何志琴
YANG Ying;HE Zhiqin(The Electrical Engineering College,Guizhou University,Guiyang 550025,China)
出处
《微处理机》
2022年第1期30-33,共4页
Microprocessors
基金
黔科合LH字[2017]7229
黔科合基础[2018]1029。
关键词
无人驾驶
深度学习
图像分割
FCN模型
Driverless driving
Deep learning
Image segmentation
FCN model
