摘要
无人机技术发展愈发成熟,具备高机动性及视角范围广等特点,在民用领域得到广泛应用,尤其是用于交通监控等。在续航里程约束条件下,提出以视频交通监控平台为基础运用无人机高像素监控高速公路实现智能化管理。利用以特征为基础的视频稳像方法及光流法相结合方法对运动目标车辆检测。采用视频稳像方法对参考帧及匹配帧间的特征点进行获取,经过特征点匹配过程找出参考帧和匹配帧间转换关系,将动态目标变化为静态目标检测问题;利用光流法对运动车辆实行检测,实现在不同背景及不同分辨率条件下有效识别出运动车辆的目的。基于视频交通监控平台,将图像质量测评引进无人机视频监控模式中,依据图像质量及缓冲区实际条件,实时调整图像编码帧率和分辨率,确保输出图像质量及视频流稳定性。实验证明,运用低空无人机监控平台可实现高速公路远程管理目标。
UAV technology is developing more and more mature, and has the characteristics of high mobility and wide viewing range. It has been widely applied in the civil field, especially for traffic monitoring. Under the condition of the endurance of the endurance, the intelligent management of the expressway is put forward on the basis of the video traffic monitoring platform. A feature based video stabilization and optical flow method is used to detect moving target vehicles. A video stabilization method of reference frame and matching feature points between frames are obtained through feature matching conversion process to find out the relationship between the reference frame and matching between frames, dynamic target change for static target detection;detection of moving vehicles using the optical flow method, implemented in different background and different resolution under the condition of effective identification the purpose of moving vehicles. Based on the video traffic monitoring platform, the image quality assessment is introduced into the UAV video surveillance mode. According to the image quality and the actual conditions of the buffer, the image coding frame rate and resolution are adjusted in real time to ensure the quality of the output image and the stability of the video stream. The experiment proves that the application of low altitude unmanned aerial vehicle monitoring platform can achieve the goal of Expressway remote management.
作者
郭晓峰
陈钊正
谭政宇
张薇
Guo Xiaofeng;Chen Zhaozheng;Tan Zhengyu;Zhang Wei(Jiangxi Expressway Networking Management Center,Nanchang 330036,China;Software College,East China Jiaotong University,Nanchang 330000,China)
出处
《科技通报》
2019年第2期89-93,共5页
Bulletin of Science and Technology
关键词
续航里程
低空无人机
高像素监控
高速公路
air mileage
low altitude unmanned aerial vehicle
high pixel monitoring
freeway
作者简介
郭晓峰(1980),男,广东普宁人,高级工程师,研究方向:通信技术。
