摘要
针对背景像素的移动,提出了SURF特征稳像和光流法向量相结合的新方法来解决低空视频的道路车辆检测。首先,检测两帧图像的SURF特征;再用最近邻匹配得到两幅图像的匹配点对;随后结合RANSAC和最小二乘法计算全局运动参数向量,获得稳定的帧;最后,根据稳定的帧计算光流法向量,并检测出运动车辆。实验结果表明,基于SURF算子的图像稳像算法在不损失稳像精度的前提下,能够提高图像稳像算法的速度,所提方法能够有效地检测出运动车辆。
Detecting moving ground vehicles from airborne video is a difficult problem because all pixels in the image are moving due to the self motion of the camera.We present a technique based on the normal component of the residual flow algorithm applied on using SURF operator stabilized frames to detect moving vehicles.Experiments show using SURF operator can improve the image stabilizing algorithm speed without loss of accuracy and this approach can effectively detect the moving vehicles.This method and airborne platform can be used as a new attempt to traffic information collection when the routine method can not be used.
出处
《武汉大学学报（信息科学版）》
EI
CSCD
北大核心
2011年第3期316-320,共5页
Geomatics and Information Science of Wuhan University
基金
国家创新团队资助项目(40721001)
国家博士点专项基金资助项目(20070486001)
中央高校基本科研业务费专项资金资助项目(10CX05007A)
关键词
航拍视频
图像稳像
车辆检测
车辆速度估计
airborne video
image stabilization
vehicle detection
vehicle speed estimation
