计算机科学 ›› 2023, Vol. 50 ›› Issue (6A): 220200066-5.doi: 10.11896/jsjkx.220200066
万海波, 蒋磊, 王晓
WAN Haibo, JIANG Lei, WANG Xiao
摘要： 摩托车驾驶相比于其他驾驶方式更加危险,但缺乏有效的辅助驾驶系统,例如车道辅助保持系统、障碍物检测、预碰撞系统等,而判定摩托车在行驶中是否发生了偏离往往需要参照车辆行驶时车道线的位置,因此车道线检测对于辅助驾驶系统来说至关重要,因此文中提出了基于深度学习的摩托车车道实时检测算法。在Lanenet架构的基础下做出了以下3点改进:将车道坐标的绝对位置作为输入特征、使用K-means算法代替Mean-Shift聚类算法以及去除Hnet结构。由于目前缺乏公开的摩托车车道数据集,因此将使用采集的摩托车车道数据对模型进行拟合,拟合的结果证明了该算法的有效性,检测速度可达47.6fps,交并比可达0.71560,相比文献[3]中的算法在精度上提高了15.5%,在速度上提高了53.3%。
中图分类号:
| [1]WANG R B,YU T H,GUO L,et al.Research Review of Lane Departure Warning System Based on Machine Vision[J].Automotive Engineering,2005,27(4):463-466. [2]TANG J G,LI S B,LIU P.A Review of Lane Detection Me-thods based on Deep Learning[J].Pattern Recognition,2021,111:107623. [3]NEVEN D,BRABANDERE B D,GEORGOULIS S,et al.Towards End-to-End Lane Detection:an Instance Segmentation Approach[C]//2018 IEEE Intelligent Vehicles Symposium(IV).Changshu:IEEE,2018:286-291. [4]HALBERSBERG D,LERNER B.Young driver fatal motorcycle accident analysis by jointly maximizing accuracy and information[J].Accident Analysis & Prevention,2019(129):350-361. [5]CHANG F,XU P,ZHOU H,et al.Huang Investigating injury severities ofmotorcycle riders:a two-step method integrating latent class cluster analysis and random parameters logit model[J].Accident Analysis & Prevention,2019(131):316-326. [6]CHEN S J,CHEN C Y,LIN M R.Risk factors for crash in-volvement in older motorcycle riders[J].Accident Analysis & Prevention,2018(111):109-114. [7]CHO H J,WU Y T.Modeling and simulation of motorcycletraffic flow[C]//2004 IEEE International Conference on Systems,Man and Cybernetics(IEEE Cat. No.04CH37583).2004:6262-6267. [8]HSU T P,WEN K L.Effect of novel divergence markings onconflict prevention regarding motorcycle-involved right turn accidents of mixed traffic flow[J].Journal of Safety Research,2019(69):167-176. [9]YOUSIF M T,SADULLAH A F M,KASSIM K A A.A review of behavioural issues contribution to motorcycle safety[J].IATSS Research,2020,44(2):142-154. [10]CHANG C Y,LIN C H.An Efficient Method for Lane-MarkExtraction in Complex Conditions[C]//2012 9th International Conference on Ubiquitous Intelligence & Computing and 9th International Conference on Autonomic & Trusted Computing(UIC/ATC).IEEE Computer Society,2012. [11]BENLIGIRAY B,TOPAL C,AKINLAR C.Video-Based Lane Detection Using a Fast Vanishing Point Estimation Method[C]//IEEE International Symposium on Multimedia.IEEE,2012:348-351. [12]LOOKINGBILL A,ROGERS J,LIEB D,et al.Reverse Optical Flow for Self-Supervised Adaptive Autonomous Robot Navigation[J].International Journal of Computer Vision,2007,74(3):287-302. [13]TAN H,ZHOU Y,ZHU Y,et al.A novel curve lane detection based on Improved River Flow and RANSA[C]//17th International IEEE Conference on Intelligent Transportation Systems.IEEE,2014:133-138. [14]LI H,FENG M,WANG X.Inverse perspective mapping based Urban road markings detection[C]//IEEE International Conference on Cloud Computing & Intelligent Systems.IEEE,2012. [15]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. [16]LIU B,LIU H Z Lane Line Detection Algorithm Based on Improved Enet Network[J].Computer Science,2020,47(4):142-149. [17]HE B,AI R,YAN Y,et al.Accurate and robust lane detection based on Dual-View Convolutional Neutral Network[C]//IEEE Intelligent Vehicles Symposium.IEEE,2016. [18]PASZKE A,CHAURASIA A,KIM S,et al.Enet:A deep neural network architecture for real-time semantic segmentation[J].arXiv:1606.02147, 2016. [19]DE BRABANDERE B,NEVEN D,VAN GOOL L.Semantic instance segmentation with a discriminative loss function[J]. ar-Xiv:1708.02551,2017. [20]ZHU X,CHENG D,ZHANG Z,et al.An empirical study of spatial attention mechanisms in deep networks[C]//Procee-dings of the IEEE/CVF International Conference on Computer Vision.2019:6688-6697. |
|
||
首页