摘要
阐述了一种基于块同化的空间频率多聚焦图像融合算法,对现有的多聚焦图像在空间频率上的融合算法进行了改进。首先对多聚焦图像在对应位置上计算出每个像素的空间频率,由空间频率来计算出它对融合图像的权值,进而决定选取哪个源图的像素作为融合后的对应位置的像素,再对融合后的图像采用窗口同化的方法,达到滤波作用,减少偶然误差。通过仿真验证了该算法的有效性,结果表明该方法得到的融合图像优于传统的多聚焦图像融合方法。
This paper elaborates the assimilation of a block-based multi-focus image fusion spatial frequency algorithm,to improve the existing multi-focus image fusion algorithm on the spatial frequency.Firstly,calculate the spatial frequency in the corresponding position of each pixel of multi-focus image,and calculate the integration of the image rights of its value by the spatial frequency,and then decided to select a source image pixel location as a fusion of the corresponding pixels by the value.And then,the fused image uses the method of assimilation window,to filter,reducing the accidental errors.Through simulation the effectiveness of the program,the results show that the method to be superior to the traditional image fusion of multi-focus image fusion method.
出处
《科学技术与工程》
北大核心
2012年第1期64-67,共4页
Science Technology and Engineering
基金
国家自然科学基金项目(60973095)资助
关键词
图像融合
空间频率
多聚焦图像
块同化
image fusion spatial frequency multi-focus image blocks assimilation
