Research On Target Recognitionand Target Trackingtechnology For Laiwu'sintelligent Transportationsystem

With the rapid development of the national economy,the growth of passenger cars in China has maintained a high speed in recent years.However,with the increase of the number of cars,many problems are also faced by traffic management.There are a large number of automobile traffic congestion problems in the complex traffic network,so the technology of target identification and tracking is particularly important,especially in the crime tracking and target identification.Taking the intelligent traffic system of Lai Wu city as an example,this paper mainly studies the omni-directional target identification and tracking technology based on intelligent traffic network.The current situation of intelligent traffic network and target identification and tracking at home and abroad is investigated,and the design of intelligent traffic system in Lai Wu city is completed by combining the development of intelligent traffic system at home and abroad.In this paper,the image preprocessing and post-processing techniques in traffic monitoring network are introduced in detail,the image feature extraction method is introduced,and the image feature extraction algorithm based on adaptive heredity is proposed.This method overcomes the disadvantage that the traditional method is easy to fall into local optimum and the effect of the feature extraction is verified by the symbolic regression experiment.The target detection algorithm in intelligent traffic monitoring is studied,and a moving target detection algorithm based on self-coding neural network and softmax is proposed.After trial,the cascade deep neural network algorithm can recognize moving objects quickly and effectively.The effect is very significant after the city of the traffic camera shooting algorithm screen verification.Finally,the target tracking algorithm based on the intelligent traffic monitoring network are studied and the particle filter based on EKF predictive sampling and the feature matching tracking algorithm are proposed based on kalman filter,combined with the two algorithms to carry out experimental verification,which achieves satisfactory results.