基于Matlab的车辆图像的配准与融合  

Matlab-Based Vehicle Image Registration and Fusion

作　　者：曹洁[1] 李浩茹[1] 陈继开[2] 

机构地区：[1]兰州理工大学电气工程与信息工程学院,甘肃兰州730050 [2]兰州理工大学机电工程学院,甘肃兰州730050

出　　处：《计算机工程与科学》2009年第4期43-45,共3页Computer Engineering & Science

摘　　要：采用模板匹配法对同源传感器视频采集到的车辆图像进行匹配,对所配准的图像采用加权平均的融合算法进行融合,可以获得对同一场景、同一目标更准确、全面、可靠的图像描述,从而实现对车辆图像特征的精确提取。通过Mat-lab仿真实验表明,该方法能更有效地提取车辆图像特征,达到对车辆外型的准确描述。By adopting the method of template correlation matching to match the vehicle images from a video camera and fuse the image fusion algorithms of weighted averaging, we can achieve a more accurate, more complete and more reliable image data from the same location. In this way,the vehicle feature image extraction is realized. The experimental results show that the proposed approach outperforms the conventional vehicle feature image extraction in describing vehicle exterior precisely.

关 键 词：模板匹配算法 加权平均融合 MATLAB 车辆图像 车辆外型 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]