基于双特征提取网络的复杂环境车道线精准检测

doi:10.13229/j.cnki.jdxbgxb.20221129

摘要/Abstract

摘要：

为解决现有方法在复杂环境中检测精度低的问题，提出了一种基于双特征提取网络的复杂环境车道线精准检测算法。首先，搭建双特征提取网络，获取不同尺度的特征图，提取更有效的特征，提高模型在复杂环境下的特征提取能力。然后，构建跨通道联合注意力模块，提高模型对车道线细节的关注度，抑制无用信息。最后，结合改进的空洞空间金字塔池化模块扩大图像感受野，提高模型对上下文信息的利用率，以强化算法的检测能力。经实验验证，本文算法在CULane数据集上的F₁-measure达到了72.43%，相比于基线模型提升了4.03%，在复杂的场景中对车道线进行检测时效果提升明显。

关键词: 计算机应用, 车道线检测, 双特征提取, 多尺度, 跨通道联合注意力

Abstract:

The existing lane detection methods have the problem of low detection accuracy due to fuzzy details in a complex environment. Therefore， this paper proposes an accurate lane detection algorithm based on a double feature extraction network in a complex environment. Firstly， a double feature extraction network is built to obtain feature maps of different scales， extract more effective features， and improve the feature extraction ability of the model in complex environments. Besides， a cross-channel joint attention module is constructed to improve the attention of the model to lane details and suppress useless information. Finally， combined with the improved void space pyramid pooling module， the receptive field is enlarged to improve the utilization of context information of the model， to strengthen the detection ability. The experimental results show that the F₁-measure of the proposed algorithm on CULane dataset reaches 72.43%， which is 4.03% higher than that of the mainstream UFSD algorithm. When detecting lane lines in complex scenes， the detection effect of the proposed method is significantly improved， which has been proven to be able to meet the needs of practical applications.

Key words: computer application, lane detection, double feature extraction, multi-scale, combined attention mechanism

中图分类号:

TP391.4

张云佐,郑宇鑫,武存宇,张天. 基于双特征提取网络的复杂环境车道线精准检测[J]. 吉林大学学报(工学版), 2024, 54(7): 1894-1902.

Yun-zuo ZHANG,Yu-xin ZHENG,Cun-yu WU,Tian ZHANG. Accurate lane detection of complex environment based on double feature extraction network[J]. Journal of Jilin University(Engineering and Technology Edition), 2024, 54(7): 1894-1902.

图/表 9

图1

图2

图3

图4

表1

表2

图5

表3

表4

参考文献 27

1	Kumar S, Jailia M, Varshney S. A comparative study of deep learning based lane detection methods[C]∥2022 9th International Conference on Computing for Sustainable Global Development (INDIACom), New Delhi, India, 2022: 579-584.
2	时小虎, 吴佳琦, 吴春国, 等. 基于残差网络的弯道增强车道线检测方法[J]. 吉林大学学报: 工学版, 2022(2): 584-592.
	Shi Xiao-hu, Wu Jia-qi, Wu Chun-guo, et al. Residual network based curve enhanced lane detection method[J]. Journal of Jilin University (Engineering and Technology Edition), 2022(2): 584-592.
3	Tang J, Li S, Liu P. A review of lane detection methods based on deep learning[J]. Pattern Recognition, 2021, 111: No.107623.
4	洪伟, 王吉通, 刘宇. 基于 DBSCAN 的复杂环境下车道线鲁棒检测及跟踪[J]. 吉林大学学报 :工学版, 2020, 50(6): 2122-2130.
	Hong Wei, Wang Ji-tong, Liu Yu. Robust lane detection and tracking in complex environment based on DBSCAN[J]. Journal of Jilin University (Engineering and Technology Edition), 2020, 50(6): 2122-2130.
5	Dang H S, Guo C J. Structure lane detection based on saliency feature of color and direction[C]∥Applied Mechanics and Materials, Trans Tech Publications Ltd, Xi´an, China,2014, 513: 2876-2879.
6	都雪静, 张美欧. 基于光流法与背景建模法融合的车道线识别算法研究[J]. 重庆理工大学学报: 自然科学, 2021, 35(3): 29-35.
	Du Xue-jing, Zhang Mei-ou. Based on fusing optical flow method and background modeling for lane line recomgnition[J]. Journal of Chongqing University of Technology (Natural Science), 2021, 35(3): 29-35.
7	Zhang Y, Lu Z, Zhang X, et al. Deep learning in lane marking detection: a survey[J]. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(7): 5976-5992.
8	Yu F, Koltun V. Multi-scale context aggregation by dilated convolutions[DB/OL]. [2022-05-17]. .
9	Poudel R P K, Liwicki S, Cipolla R. Fast-scnn: fast semantic segmentation network[DB/OL]. [2022-05-23]. .
10	Borji A. Vanishing point detection with convolutional neural networks[DB/OL]. 2022-05-26]. .
11	Lee S, Kim J, Shin Yoon J, et al. Vpgnet: vanishing point guided network for lane and road marking detection and recognition[C]∥Proceedings of the IEEE international conference on computer vision, Venice, Italy, 2017: 1947-1955.
12	崔文靓, 王玉静, 康守强, 等. 基于改进 YOLOv3 算法的公路车道线检测方法[J]. 自动化学报, 2022, 45: 1-9.
	Cui Wen-liang, Wang Yu-jing, Kang Shou-qiang, et al. Road lane line detection method based on improved YOLOv3 algorithm[J]. Journal of Automatica Sinica,2022, 45: 1-9.
13	高明华, 杨璨. 基于改进卷积神经网络的交通目标检测方法[J]. 吉林大学学报 :工学版, 2022, 52(6): 1353-1361.
	Gao Ming-hua, Yang Can. Traffic target detection method based on improved convolution neural network[J]. Journal of Jilin University (Engineering and Technology Edition), 2022, 52(6): 1353-1361.
14	Tabelini L, Berriel R, Paixao T M, et al. Polylanenet: lane estimation via deep polynomial regression[C]∥2020 25th International Conference on Pattern Recognition (ICPR), Milan, Italy, 2021: 6150-6156.
15	Sun Y, Wang L, Chen Y, et al. Accurate lane detection with atrous convolution and spatial pyramid pooling for autonomous driving[C]∥2019 IEEE International Conference on Robotics and Biomimetics (ROBIO), Dali, China, 2019: 642-647.
16	Neven D, De Brabandere B, Georgoulis S, et al. Towards end-to-end lane detection: an instance segmentation approach[C]∥IEEE intelligent vehicles symposium (IV),Changshu, China, 2018: 286-291.
17	Tabelini L, Berriel R, Paixao T M, et al. Keep your eyes on the lane: real-time attention-guided lane detection[C]∥Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, Nashville,USA,2021: 294-302.
18	汪鹏飞, 沈庆宏, 张维利, 等. 基于多尺度特征图像分割的车道线提取方法[J]. 南京大学学报 :自然科学版, 2022, 58(2): 336-344.
	Wang Peng-fei, Shen Qing-hong, Zhang Wei-li, et al. Lane extraction method based on multi-scale feature image segmentation[J]. Journal of Nanjing University (Natural Science), 2022, 58(2): 336-344.
19	Fan R, Wang X, Hou Q, et al. SpinNet: spinning convolutional network for lane boundary detection[J]. Computational Visual Media, 2019, 5(4): 417-428.
20	Qin Z, Wang H, Li X. Ultra fast structure-aware deep lane detection[C]∥European Conference on Computer Vision. Cham: Springer, 2020: 276-291.
21	Woo S, Park J, Lee J Y, et al. Cbam: convolutional block attention module [C]∥European Conference on Computer Vision. Berlin: Springer, 2018: 3-19.
22	Wang Q, Wu B, Zhu P, et al. ECA-Net: efficient channel attention for deep convolutional neural networks[C]∥2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle,USA,2020: 11531-11539.
23	Philion J. Fastdraw: addressing the long tail of lane detection by adapting a sequential prediction network[C]∥Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Long Beach, USA,2019: 11582-11591.
24	Pan X, Shi J, Luo P, et al. Spatial as deep: spatial cnn for traffic scene understanding[C]∥Proceedings of the AAAI Conference on Artificial Intelligence, New Orleans, USA, 2018, 32(1): 7276-7283.
25	Hou Y, Ma Z, Liu C, et al. Learning lightweight lane detection cnns by self-attention distillation[C]∥Proce-edings of the IEEE/CVF international conference on computer vision,Seoul, Korea (South),2019: 1013-1021.
26	Xu H, Wang S, Cai X, et al. Curvelane-nas: unifying lane-sensitive architecture search and adaptive point blending[C]∥European Conference on Computer Vision. Cham:Springer, 2020: 689-704.
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.

相关文章 15

[1]	孙铭会,薛浩,金玉波,曲卫东,秦贵和. 联合时空注意力的视频显著性预测[J]. 吉林大学学报(工学版), 2024, 54(6): 1767-1776.
[2]	李延风,刘名扬,胡嘉明,孙华栋,孟婕妤,王奥颖,张涵玥,杨华民,韩开旭. 基于梯度转移和自编码器的红外与可见光图像融合[J]. 吉林大学学报(工学版), 2024, 54(6): 1777-1787.
[3]	张丽平,刘斌毓,李松,郝忠孝. 基于稀疏多头自注意力的轨迹kNN查询方法[J]. 吉林大学学报(工学版), 2024, 54(6): 1756-1766.
[4]	梁礼明,周珑颂,尹江,盛校棋. 融合多尺度Transformer的皮肤病变分割算法[J]. 吉林大学学报(工学版), 2024, 54(4): 1086-1098.
[5]	张云佐,郭威,李文博. 遥感图像密集小目标全方位精准检测算法[J]. 吉林大学学报(工学版), 2024, 54(4): 1105-1113.
[6]	李雄飞,宋紫萱,朱芮,张小利. 基于多尺度融合的遥感图像变化检测模型[J]. 吉林大学学报(工学版), 2024, 54(2): 516-523.
[7]	霍光,林大为,刘元宁,朱晓冬,袁梦,盖迪. 基于多尺度特征和注意力机制的轻量级虹膜分割模型[J]. 吉林大学学报(工学版), 2023, 53(9): 2591-2600.
[8]	何颖,王卓然,周旭,刘衍珩. 融合社交地理信息加权矩阵分解的兴趣点推荐算法[J]. 吉林大学学报(工学版), 2023, 53(9): 2632-2639.
[9]	张云佐,董旭,蔡昭权. 拟合下肢几何特征的多视角步态周期检测[J]. 吉林大学学报(工学版), 2023, 53(9): 2611-2619.
[10]	肖明尧,李雄飞,朱芮. 基于NSST域像素相关分析的医学图像融合[J]. 吉林大学学报(工学版), 2023, 53(9): 2640-2648.
[11]	赵亚慧,李飞雨,崔荣一,金国哲,张振国,李德,金小峰. 基于跨语言预训练模型的朝汉翻译质量评估[J]. 吉林大学学报(工学版), 2023, 53(8): 2371-2379.
[12]	郭晓新,李佳慧,张宝亮. 基于高分辨率网络的视杯和视盘的联合分割[J]. 吉林大学学报(工学版), 2023, 53(8): 2350-2357.
[13]	车翔玖,徐欢,潘明阳,刘全乐. 生物医学命名实体识别的两阶段学习算法[J]. 吉林大学学报(工学版), 2023, 53(8): 2380-2387.
[14]	王连明,吴鑫. 基于姿态估计的物体3D运动参数测量方法[J]. 吉林大学学报(工学版), 2023, 53(7): 2099-2108.
[15]	张则强,梁巍,谢梦柯,郑红斌. 混流双边拆卸线平衡问题的精英差分进化算法[J]. 吉林大学学报(工学版), 2023, 53(5): 1297-1304.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

算法模型	FastDraw^［24］	SCNN^［23］	SAD^［25］	UFSD^［20］	CurveLanes^［26］	本文
综合	-	71.60	70.80	68.40	71.40	72.43
正常	85.90	90.60	90.10	87.70	88.30	90.72
拥挤	63.6	69.70	68.80	66.00	68.60	72.28
强光	57.0	58.50	60.20	58.40	63.20	64.56
阴影	59.9	66.90	65.90	62.80	68.00	68.89
无车道线	40.6	43.40	41.60	40.20	47.90	43.38
箭头	79.4	84.10	84.00	81.00	82.50	86.46
弯道	65.2	64.40	65.70	57.90	66.00	64.12
交叉路口	7013	1990	1998	1743	2817	1804
夜晚	57.8	66.10	66.00	62.10	66.20	66.53

基础模型^［20］	DFE-Net	CBAM^［21］	AUAM	ASPP^［27］	ASPP*	F₁-measure
√						68.40
√	√					69.97
√	√	√				70.03
√	√		√			70.56
√	√		√	√		71.92
√	√		√		√	72.43

基础模型	DFE-Net	AUAM	ASPP*	F₁-measure
√				68.40
√	√			69.97
√	√	√		70.56
√	√	√	√	72.43