复杂路况下的车道线高精度检测方法

doi:10.3778/j.issn.1002-8331.2404-0322

摘要/Abstract

摘要： 复杂路况下的大雾、强光或无视觉线索使得车道线检测的可用信息十分有限，导致车道线的检测精度不高。针对该问题，在一种基于分割+分类的检测模型上提出一种新的车道线检测方法DrfLane（DCNV3+RT-DETR+FocalDiceLoss Lane）。在残差网络的特征提取模块中引入动态卷积，使网络能够根据不同特征动态调整卷积核的形状，增强模型在不同场景下的适应能力。在此基础上对残差网络的输出特征进行多尺度特征融合，使得模型能够同时捕捉到更多局部细节和全局结构，以提升模型对不同尺度车道线的感知能力，从而提高预测精度。此外，针对传统方法在检测过程中未能考虑到车道线的细长结构进而导致正负样本不均衡，通过建立FocalDiceLoss解决样本类别不平衡和难易样本问题。将该方法与目前先进的车道线检测算法在大型复杂路况数据集CULane上进行了实验对比，结果表明该方法在检测精度方面比其他优秀方法均有提升，验证了所提方法对车道线检测的准确性和有效性。

关键词: 车道线检测, 动态卷积, 多尺度融合, 结构损失, 锚点

Abstract: Fog, glare or no visual cues in complex road conditions make the available information for lane line detection very limited, resulting in poor lane line detection accuracy. To address this problem, this paper proposes a new lane line detection method DrfLane (DCNV3+RT-DETR+FocalDiceLoss Lane) on a segmentation+classification based detection model. Dynamic convolution is introduced into the feature extraction module of the residual network, so that the network can dynamically adjust the shape of the convolution kernel according to different features, and enhance the model’s adaptability in different scenes. On this basis, multi-scale feature fusion is performed on the output features of the residual network, which enables the model to capture more local details and global structures at the same time, in order to enhance the model’s ability to perceive lane lines at different scales, and thus improve the prediction accuracy. In addition, to address the problem that traditional methods fail to consider the elongated structure of lane lines during the detection process and thus lead to the imbalance of positive and negative samples, the problem of sample category imbalance and difficult and easy samples is solved by establishing the FocalDiceLoss. The experimental comparison between the pproposed method and the current advanced lane line detection algorithms is carried out on a large and complex road condition dataset CULane, and the results show that the proposed method has improved the detection accuracy than other excellent methods, which verifies the accuracy and effectiveness of the proposed method for lane line detection.

Key words: lane line detection, dynamic convolution, multiscale fusion, structural loss, anchor points

李超, 唐雅琪, 彭琴. 复杂路况下的车道线高精度检测方法[J]. 计算机工程与应用, 2025, 61(15): 269-277.

LI Chao, TANG Yaqi, PENG Qin. High-Precision Detection of Lane Lines in Complex Road Conditions[J]. Computer Engineering and Applications, 2025, 61(15): 269-277.

参考文献

[1] ZOU Q, JIANG H W, DAI Q Y, et al. Robust lane detection from continuous driving scenes using deep neural networks[J]. IEEE Transactions on Vehicular Technology, 2020, 69(1): 41-54.
[2] HUR J, KANG S N, SEO S W. Multi-lane detection in urban driving environments using conditional random fields[C]//Proceedings of the 2013 IEEE Intelligent Vehicles Symposium. Piscataway: IEEE, 2013: 1297-1302.
[3] SUN H, LUO Z M, REN D, et al. Partial Siamese with multiscale bi-codec networks for remote sensing image haze removal[J]. IEEE Transactions on Geoscience and Remote Sensing, 2023, 61: 4106516.
[4] SUN H, LI B H, DAN Z P, et al. Multi-level feature intera-ction and efficient non-local information enhanced channel attention for image dehazing[J]. Neural Networks, 2023, 163: 10-27.
[5] ALY M. Real time detection of lane markers in urban streets[C]//Proceedings of the 2008 IEEE Intelligent Vehicles Symposium. Piscataway: IEEE, 2008: 7-12.
[6] QIN Z Q, WANG H Y, LI X. Ultra fast structure-aware deep lane detection[C]//Proceedings of the 16th European Conference on Computer Vision. Cham: Springer International Publishing, 2020: 276-291.
[7] 王学东, 黄宏成. 复杂变道场景下的轻量化车道线检测算法研究[J]. 传动技术, 2023, 37(3): 3-15.
WANG X D, HUANG H C. Research on lightweight lane detection algorithm in complex lane changing scenes[J]. Drive System Technique, 2023, 37(3): 3-15.
[8] CHOI H C, OH S Y. Illumination invariant lane color recognition by using road color reference & neural networks[C]//Proceedings of the 2010 International Joint Conference on Neural Networks. Piscataway: IEEE, 2010: 1-5.
[9] BORKAR A, HAYES M, SMITH M T. Polar randomized Hough transform for lane detection using loose constraints of parallel lines[C]//Proceedings of the 2011 IEEE International Conference on Acoustics, Speech and Signal Processing. Piscataway: IEEE, 2011: 1037-1040.
[10] BORKAR A, HAYES M, SMITH M T. Robust lane detection and tracking with ransac and Kalman filter[C]//Proceedings of the 2009 16th IEEE International Conference on Image Processing. Piscataway: IEEE, 2009: 3261-3264.
[11] PAN X G, SHI J P, LUO P, et al. Spatial as deep: spatial CNN for traffic scene understanding[C]//Proceedings of the AAAI Conference on Artificial Intelligence, 2018.
[12] KIM J, PARK C. End-to-end ego lane estimation based on sequential transfer learning for self-driving cars[C]//Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops. Piscataway: IEEE, 2017: 1194-1202.
[13] NEVEN D, DE BRABANDERE B, GEORGOULIS S, et al. Towards end-to-end lane detection: an instance segmentation approach[C]//Proceedings of the 2018 IEEE Intelligent Vehicles Symposium. Piscataway: IEEE, 2018: 286-291.
[14] ZHENG T, HUANG Y F, LIU Y, et al. CLRNet: cross layer refinement network for lane detection[C]//Proceedings of the 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2022: 888-897.
[15] KIM J, H KIM J, JANG G J, et al. Fast learning method for convolutional neural networks using extreme learning mach-ine and its application to lane detection[J]. Neural Networks, 2017, 87: 109-121.
[16] LI J, MEI X, PROKHOROV D, et al. Deep neural network for structural prediction and lane detection in traffic scene[J]. IEEE Transactions on Neural Networks and Learning Systems, 2017, 28(3): 690-703.
[17] GOODFELLOW I, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial networks[J]. Communications of the ACM, 2020, 63(11): 139-144.
[18] GHAFOORIAN M, NUGTEREN C, BAKA N, et al. EL-GAN: embedding loss driven generative adversarial networks for lane detection[C]//Proceedings of the European Conference on Computer Vision. Cham: Springer, 2019: 256-272.
[19] 袁志祥, 高永奇. InternDiffuseDet: 结合可变形卷积和扩散模型的目标检测方法[J]. 计算机工程与应用, 2024, 60(12): 203-215.
YUAN Z X, GAO Y Q. InternDiffuseDet: object detection method combining deformable convolution and diffusion model[J]. Computer Engineering and Applications, 2024, 60(12): 203-215.
[20] DAI J F, QI H Z, XIONG Y W, et al. Deformable convolutional networks[C]//Proceedings of the 2017 IEEE International Conference on Computer Vision. Piscataway: IEEE, 2017: 764-773.
[21] ZHU X Z, HU H, LIN S, et al. Deformable ConvNets V2: more deformable, better results[C]//Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2019: 9300-9308.
[22] 蔡腾, 陈慈发, 董方敏. 结合Transformer和动态特征融合的低照度目标检测[J]. 计算机工程与应用, 2024, 60(9): 135-141.
CAI T, CHEN C F, DONG F M. Low-light object detection combining Transformer and dynamic feature fusion[J]. Computer Engineering and Applications, 2024, 60(9): 135-141.
[23] 徐守坤, 顾佳楠, 庄丽华, 等. 基于两阶段计算Transformer的小目标检测[J]. 计算机科学与探索, 2023, 17(12): 2967-2983.
XU S K, GU J N, ZHUANG L H, et al. Small object detection based on two-stage calculation Transformer[J]. Journal of Frontiers of Computer Science and Technology, 2023, 17(12): 2967-2983.
[24] LIN T Y, GOYAL P, GIRSHICK R, et al. Focal loss for dense object detection[C]//Proceedings of the 2017 IEEE International Conference on Computer Vision. Piscataway: IEEE, 2017: 2999-3007.
[25] LI X Y, SUN X F, MENG Y X, et al. Dice loss for data-imbalanced NLP tasks[C]//Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics. Stroudsburg: ACL, 2020: 465-476.
[26] LENG Z, TAN M, LIU C, et al. Polyloss: a polynomial expansion perspective of classification loss functions[J]. arXiv:2204.12511, 2022.
[27] CHEN L C, PAPANDREOU G, KOKKINOS I, et al. DeepLab: semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40(4): 834-848.
[28] PHILION J. FastDraw: addressing the long tail of lane dete-ction by adapting a sequential prediction network[C]//Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2019: 11574-11583.
[29] HOU Y N, MA Z, LIU C X, et al. Learning lightweight lane detection CNNs by self attention distillation[C]//Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision. Piscataway: IEEE, 2019: 1013-1021.
[30] XU H, WANG S J, CAI X Y, et al. CurveLane-NAS: unifying lane-sensitive architecture search and adaptive point blending[C]//Proceedings of the 16th European Conference on Computer Vision. Cham: Springer International Publishing, 2020: 689-704.