Lane line detection algorithm based on improved UNet network

In order to solve the problems of loss of image information and insufficient acquisition of global context information in lane line detection, this paper proposes a lane line detection algorithm based on improved UNet network. On the basis of UNet, a CBAM module is introduced, which combines channel attention and spatial attention to enhance the perceptual and expressive abilities in the feature extraction process. By adaptively learning the importance weights of feature channels and spatial locations, the CBAM module is able to better capture the important features of lane lines and improve the attention to the region of interest. In addition, this paper optimizes the loss function by using the Dice loss function and the Focal loss function to solve the problems of sample imbalance and insufficient sensitivity to lane line boundary details. Experimental results show that the method proposed in this paper achieves superior performance on the Tusimple dataset.