基于实时图像处理算法的交通信号主动智能优化方法研究

Research on traffic signal active intelligent optimization method based on real⁃time image processing algorithm

为了提升城市交通的通勤效率,文中基于图像处理技术对交通信号灯的智能优化方案进行了研究。该方案引入了一种基于混合高斯模型的运动目标检测方法,可有效从道路背景中分离车辆的前景信息。并采用虚拟线圈法进行车辆统计,在识别出交叉路口的车道信息后按照不同车道分别统计车辆数量。同时还设计了一个以平均排队时长、排队长度和停车次数为优化目标的多目标约束规划模型,并通过粒子群算法完成了模型的求解。仿真结果表明,所提技术方案可适应不同交通流量交叉路口的车辆统计,且精度能够达到95.4%。相较于现有的Webster方案,文中信号灯优化方案的平均延误时间、平均停车次数与平均等待长度分别下降了2.76%、8.35%及12.44%。

In order to improve the commuting efficiency of urban traffic,this paper studies the intelligent optimization scheme of traffic lights based on image processing technology.In this scheme,a moving object detection method based on Gaussian mixture model is introduced,which can effectively separate the vehicle foreground information from the road background,and the virtual loop method is used for vehicle statistics.After identifying the lane information of the intersection,the number of vehicles is counted according to different lanes.This paper designs a multi⁃objective constrained programming model with the average queue length,queue length and parking times as the optimization objectives,and completes the solution of the model by particle swarm optimization algorithm.The simulation results show that the technical scheme proposed in this paper can be adapted to the intersection vehicle statistics of different traffic flows,and the accuracy can reach 95.4%.Compared with the existing Webster scheme,the signal optimization scheme reduces the average delay time by 2.76%,the average number of stops by 8.35%,and the average waiting length by 12.44%.