摘要
为了及时准确地识别城市主干路交通状态,使用无人机悬停于观测路段上方进行交通数据采集,对采集的交通视频数据进行分帧、裁剪,建立交通状态图像数据集并进行预处理。针对PCA算法易受光照影响这一问题,融合LBP算法提取图像特征,并通过粒子群算法优化SVM算法,搭建交通状态识别模型。利用已建立的交通状态图像数据集训练LBP-PCA-SVM和PCA-SVM模型,对所得结果进行对比分析。结果表明,LBP-PCA-SVM模型相比PCA-SVM模型识别性能较高,并且能够满足实时性要求。
In order to identify the traffic state of urban main roads timely and accurately,the UAV is used to hover over the observation section to collect the traffic data,and the collected traffic video data is framed and cropped to obtain the traffic state image dataset prior to preprocessing.To solve the problem that the PCA algorithm is easily affected by illumination,the LBP algorithm is fused to extract image features,and the SVM algorithm is optimized by the particle swarm optimization algorithm to build the traffic state recognition model.The LBP-PCA-SVM and PCA-SVM models are trained using the established traffic state dataset,and the results are compared and analyzed.The experimental results show that the LBP-PCA-SVM model has higher recognition performance than PCA-SVM model and can meet the real-time requirements.
作者
崔淑艳
宇仁德
朱燕华
CUI Shuyan;YU Rende;ZHU Yanhua(School of Transportation and Vehicle Engineering,Shandong University of Technology,Zibo 255049,China)
出处
《山东理工大学学报（自然科学版）》
CAS
2024年第5期7-11,60,共6页
Journal of Shandong University of Technology:Natural Science Edition
关键词
交通状态识别
支持向量机
无人机数据采集
traffic state recognition
support vector machine
UAV data acquisition
