基于循环特征融合的弯道增强车道线检测算法

doi:10.13474/j.cnki.11-2246.2023.0354

测绘通报 ›› 2023, Vol. 0 ›› Issue (12): 25-30.doi: 10.13474/j.cnki.11-2246.2023.0354

基于循环特征融合的弯道增强车道线检测算法

鲁维佳¹, 刘泽帅², 潘玉恒¹, 李国燕¹, 李慧洁¹, 丛佳¹

1. 天津城建大学计算机信息与工程学院, 天津 300384;
2. 中国铁塔股份有限公司天津市分公司, 天津 300011

收稿日期:2023-06-28 发布日期:2024-01-08
通讯作者: 潘玉恒。E-mail:panyuheng@tju.edu.cn
作者简介:鲁维佳(1980-),男,硕士,讲师,研究方向为嵌入式系统设计、自动驾驶。E-mail:13920156816@139.com
基金资助:
国家自然科学基金(62204168);天津市科技计划(20YDTPJC00160;21YDTPJC00780);天津市教委科研计划(2019KJ101)

Enhanced lane line detection algorithm for curves based on Resa-CC

LU Weijia¹, LIU Zeshuai², PAN Yuheng¹, LI Guoyan¹, LI Huijie¹, CONG Jia¹

1. Tianjin Chengjian University, School of Computer Information and Engineering, Tianjin 300384, China;
2. Tianjin Branch Company of China Tower Corperation Limited, Tianjin 300011, China

Received:2023-06-28 Published:2024-01-08

摘要/Abstract

摘要： 针对道路转弯处曲率过大导致弯道识别的精度下降的问题,本文提出了一种基于循环特征融合Resa-CC的弯道增强车道线检测算法。该算法利用车道线的形状先验性,捕获图像像素中行与列的空间关系,融合信息生成特征图;以残差网络为主体框架,加入编码器、解码器和注意力机制模块,在损失函数中引入弯道结构约束来提高车道线弯道的识别精度。加入循环特征融合模块和自注意力机制模块后准确率分别提升3.41%和1.1%,证明了两模块的有效性;Resa-CC算法准确率可达96.83%,FPS为35.68,误检率FP和漏检率FN分别为0.031 5和0.028 2,表明本文算法具有较高的检测性能,在车辆行驶弯道路段中能更准确地推断出车道线的位置。

关键词: 交通工程, 车道线检测, 循环特征聚合, 深度学习, 自动驾驶, ResNet

Abstract: A curve enhanced lane detection algorithm based on cyclic feature fusion Resa-CC is proposed to address the issue of reduced accuracy in curve recognition caused by excessive curvature at road turns. This algorithm utilizes the shape priors of lane lines to capture the spatial relationships between rows and columns in image pixels, and fuse information to generate feature maps. The residual network is used as the main framework, and the encoder, decoder and attention mechanism modules are added. The Loss function adds curve structure constraints to improve the recognition accuracy of lane curves. The addition of the cyclic feature fusion module and the self attention mechanism module improved the accuracy by 3.41% and 1.1%, respectively, proving the effectiveness of the two modules. The accuracy of the Resa-CC algorithm can reach 96.83%, with an FPS of 35.68. The false detection rate FP and missed detection rate FN are 0.0315 and 0.0282, respectively. This indicates that the algorithm has high detection performance and can more accurately infer the position of the lane line in the curve when vehicles are driving.

Key words: traffic engineering, lane line detection, recurrent feature aggregation, deep learning, automatic driving, ResNet

中图分类号:

P237

鲁维佳, 刘泽帅, 潘玉恒, 李国燕, 李慧洁, 丛佳. 基于循环特征融合的弯道增强车道线检测算法[J]. 测绘通报, 2023, 0(12): 25-30.

LU Weijia, LIU Zeshuai, PAN Yuheng, LI Guoyan, LI Huijie, CONG Jia. Enhanced lane line detection algorithm for curves based on Resa-CC[J]. Bulletin of Surveying and Mapping, 2023, 0(12): 25-30.

参考文献

[1] TANG J,LI S,LIU P. A review of lane detection methods based on deep learning[J]. Pattern Recognition,2020(1) 1-15.
[2] LIANG Dun,GUO Yuanchen,ZHANG Shaokui,et al. Lane detection: a survey with new results[J]. Journal of Computer Science & Technology,2020,35(3): 493-505.
[3] CHEN Weiwei,WANG Weixing,WANG K,et al. Lane departure warning systems and lane line detection methods based on image processing and semantic segmentation: a review[J]. Journal of Traffic and Transportation Engineering (English Edition),2020,7(6): 748-774.
[4] DING L,ZHANG H Y,XIAO J S,et al. A lane detection method based on semantic segmentation[J]. CMES-Computer Modeling in Engineering & Sciences,2020,122(3): 1039-1053.
[5] SIMONYAN K,ZISSERMAN A. Very deep convolutional networks for large-scale image recognition[EB/OL]. 2014-03-12[2023-03-10].https://arxiv.org/abs/1409.1556.pdf.
[6] XIONG H,YU D M,LIU J X,et al. Fast and robust approaches for lane detection using multi-camera fusion in complex scenes [J]. IET Intelligent Transport Systems,2020,14(12): 1582-1593.
[7] BADRINARAYANAN V,KENDALL A,CIPOLLA R. SegNet: a deep convolutional encoder-decoder architecture for image segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2017,39(12): 2481-2495.
[8] PIZZATI F,ALLODI M,BARRERA A,et al. Lane detection and classification using cascaded CNNs[EB/OL]. 2019-03-11[2023-03-10].https://arxiv.org/abs/1907.01294.pdf.
[9] ZOU Qin,JIANG Hanwen,DAI Qiyu,et al. Robust lane detection from continuous driving scenes using deep neural networks[J].IEEE Transactions on Vehicular Technology,2020,69(1):41-54.
[10] NEVEN D,BRABANDERE B D,GEORGOULIS S,et al. Towards end-to-end lane detection: an instance segmentation approach [C]//Proceedings of 2018 IEEE Intelligent Vehicles Symposium (IV). New York: IEEE,2018: 286-291.
[11] DE BRABANDERE B,NEVEN D,VAN GOOL L. Semantic instance segmentation with a discriminative loss function[EB/OL]. 2017-03-05[2023-03-20]. https://arxiv.org/abs/1708.02551.pdf.
[12] KO Y,LEE Y,AZAM S,et al. Key points estimation and point instance segmentation approach for lane detection [J]. IEEE Transactions on Intelligent Transportation Systems,2021,23(7): 8949-8958.
[13] XU Hang,WANG Shaoju,CAI Xinyue,et al. CurveLane-NAS: unifying lane-sensitive architecture search and adaptive point blending[EB/OL]. 2020-02-07[2023-03-30]. https://arxiv.org/abs/2007.12147.pdf.
[14] 周经美,王钰,宁航,等. 面向多元场景结合GLNet的车道线检测算法[J]. 中国公路学报,2021,34(7): 118-127.
[15] HUANG B,ZHENG J,GIANNAROU S,et al. H-Net: unsupervised attention-based stereo depth estimation leveraging epipolar geometry[C]//Proceedings of 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW).Orleans:IEEE,2020.
[16] TABELINI L,BERRIEL R,PAIXIAO T M,et al. Keep your eyes on the lane: Real-time attention-guided lane detection[C]//Proceedings of 2021 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Nashville:IEEE,2021.
[17] LI X,ZHANG Y,SHEN C,et al. RESA: recurrent feature-Shift aggregator for lane detection[C]// Proceedings of 2019 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).[S.l.]:IEEE,2019.
[18] HE K M,ZHANG X Y,REN S Q,et al. Deep residual learning for image recognition[C]// Proceedings of 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). [S.l.]:IEEE,2016.
[19] WOO S,PARK J,LEE J Y,et al. CBAM: convolutional block attention module[EB/OL]. 2018-03-05[2023-03-31].https://arxiv.org/abs/1807.06521.pdf.
[20] QIN Z,WANG H,LI X. Ultra fast structure-aware deep lane detection[C]// Proceedings of 2020 European Conference on Computer Vision (ECCV).[S.l.]:Springer,2020.

基于循环特征融合的弯道增强车道线检测算法

Enhanced lane line detection algorithm for curves based on Resa-CC

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

[1]	魏德宏, 禤键豪, 杨嘉伟, 张兴福, 余旭. 基于深度BP/ELMAN神经网络的山区GNSS高程转换精度分析[J]. 测绘通报, 2023, 0(9): 113-116,143.
[2]	张宝安, 高小龙, 金仔燕, 马兰花. 甘肃省地表要素遥感解译样本库建设与应用[J]. 测绘通报, 2023, 0(9): 144-149.
[3]	胡荣明, 李鑫, 竞霞, 武建强, 魏青博. YOLOv7在探地雷达B-Scan图像解译中的应用[J]. 测绘通报, 2023, 0(8): 29-33.
[4]	周宝定, 谢沛瑶, 郭文浩, 毛庆洲. 基于激光点云灰度图像的隧道渗水病害检测[J]. 测绘通报, 2023, 0(8): 34-39,90.
[5]	张戬, 高雅. 深度学习遥感影像解译技术在耕地保护中的应用[J]. 测绘通报, 2023, 0(8): 142-145.
[6]	杨盈, 邱芹军, 谢忠, 田苗, 郑诗语, 郑帅. 人在回路学习增强的地理命名实体识别[J]. 测绘通报, 2023, 0(8): 155-160,177.
[7]	钟静, 孙杰, 赖祖龙, 谌一夫. ICESat-2与Sentinel-2数据融合的深度学习浅滩水深测量[J]. 测绘通报, 2023, 0(7): 39-43.
[8]	费雯凯. 面向高精地图的局部相对位置动态生成方法及其应用[J]. 测绘通报, 2023, 0(6): 1-5.
[9]	赵明姝. 车载导航到自动驾驶电子地图发展综述[J]. 测绘通报, 2023, 0(6): 6-10,92.
[10]	王丽妍, 周勋, 胡伟, 刘立歆, 李飞雪. 面向自动驾驶的高精地图数据引擎模型[J]. 测绘通报, 2023, 0(6): 11-14,19.
[11]	沈鑫甦, 嵇灵. 面向遥感地类变化检测的U型深度学习神经网络改进方法[J]. 测绘通报, 2023, 0(6): 93-97,103.
[12]	陈桥驿, 闫宇飞, 黄永芳. 融合面向对象和深度置信网络的农村建筑物信息提取[J]. 测绘通报, 2023, 0(6): 134-137.
[13]	王丽梅, 王延正. 基于高分辨率遥感影像的建筑物提取[J]. 测绘通报, 2023, 0(6): 180-183.
[14]	李心媛, 贺智, 楼桉君, 肖曼. 双线性图卷积网络的环南海地区湿地遥感分类[J]. 测绘通报, 2023, 0(5): 44-50.
[15]	蔡军, 温日强, 江伟, 严娇, 卢丽娟. 传统村落航拍图像中民族建筑识别的DeepLabV3+改进算法[J]. 测绘通报, 2023, 0(4): 49-53.