基于卷积神经网络的车道线语义分割算法

Laneline semantic segmentation algorithm based on convolutional neural network

针对车道线检测环境复杂,光照变化复杂等特点,提出了一种新型车道线检测方法。首先运用可变形卷积神经网络提取特征,然后通过对白天、夜晚、雨天等复杂光照条件下的KITTI道路数据集进行联合训练,端到端获取车道线上下文信息。建立结构化道路车道线网络模型,进而对车道线进行图像语义分割,并判断车道线类型。该模型预测车道线像素所属的场景语义类别,实现车道线实时检测。实验结果表明,该方法具有较好的准确性和实时性,在多场景结构化道路上的车道线识别率可达96.83%。

Aiming at the complexities of environment and light of laneline detection,a new method for laneline detection is proposed.Firstly,the deformable convolution neural network is used to extract the laneline feature. Secondly,joint training of KITTI road data sets under complex light conditions such as daytime,night and rainy day to get end-to-end context lane information. Finally,the laneline network model of structural road was established to distinguish the laneline type and semantically segment the laneline. This model can predict the scene semantic category of laneline pixels and detect the laneline in real time. Experimental results show the proposed method has good accuracy and real-time performance,with the laneline recognition rate of the proposed method being 96.83% in the multi scene of structured road.