摘要
图像融合技术已经成为计算机视觉、自动控制、机器人、目标识别跟踪等领域的关键技术之一。图像融合技术主要研究对象是配准后共视场(FOV)的多源图像系统,现实生活中存在大量不同特性的不同视场的图像系统,不同视场的多源图像融合技术能够在不增加硬件设备投入的情况下,提升图像采集系统的性能;针对不同视场的多源图像融合技术,提出一种仿射变换和线性插值相结合的配准方法;对融合前后的图像在时域、空域进行了分析比较,并对融合的结果进行了评价和对比,验证了此技术应用的可行性。
In recent years, the image fusion technology already has become one of key technologies in some areas, such as computer vision, automatic control, robot, target identification and track. Its main research object is the registered multiple source image having the same field of view（ FOV）. There exist a lot of image systems with different features in different fields of view in real life. The muhi-source image fusion technology for different fields of view can improve the performance of image acquisition system without investment increase in hardware. For the multi-image fusion technology of different FOV, in this paper, a registration method is proposed, which combines affine transformation and linear interpolation, the images before and after fusion are analyzed and compared in temporal and spatial domains and the re- sults are evaluated to verify the feasibility of the method.
出处
《兵工学报》
EI
CAS
CSCD
北大核心
2012年第1期26-31,共6页
Acta Armamentarii
基金
江苏省普通高校研究生科研创新计划项目(CX09B_097Z)
关键词
信息处理技术
多源图像
不同视场
配准
融合
information processing
multi-source image
different field of view
registration
fusion
