基于递归网络的高速公路图像超分辨率算法研究

京台智慧高速利用5G、北斗、大数据、云计算等信息技术,通过无人机巡查、毫米波雷达采集、图像监控等手段,提升交通系统管理效能,实现对智慧高速的精细化管控。智能监控系统作为京台智慧高速的关键一环,对监督车辆规范行驶具有重要意义。由于天气变化及车辆运动,导致智能监控系统抓拍的图像存在模糊或低分辨率现象。交通模糊图像复原已成为交通违规取证中亟需解决的难点,传统方法无法满足复杂交通图像的复原问题,本文以深度学习为基础,进行交通模糊图像复原算法的研究。

Using 5G,Beidou,big data,cloud computing and other information technologies,Beijing-Taipei Smart Expressway improves the management efficiency of the traffic system by means of UAV patrols,millimeter-wave radar acquisition,image monitoring and other means to achieve the fine control of the smart expressway.As a key link of Beijing-Taipei Smart Expressway,the intelligent monitoring system is of great significance in supervising the standardized driving of motor vehicles.Due to the weather changes and vehicle movements,the images captured by the intelligent monitoring system are fuzzy or low-resolution.The restoration of blurred traffic images has become an urgent problem to be solved in the collection of traffic violation evidence.Traditional methods can’t meet the requirements for the restoration of complex traffic images.This paper studies the restoration algorithm for blurred traffic images based on deep learning.