摘要
针对利用小波变换进行医学图像融合存在的问题,提出一种基于小波变换的加权局部对比度的医学图像融合方法,该方法在低频子带采用区域标准方差的融合规则.由于人类视觉系统对局部对比度比较敏感,高频子带根据小波变换的方向特性采用加权局部对比度的融合规则,结合人眼视觉特性给出了加权局部对比度的系数选择方案,该规则更加符合生理视觉的特性.实验结果表明:该方法与传统融合方法相比,得到具有更好视觉效果和更优量化指标的融合图像.
This paper proposes a weighted local contrast medical image fusion method based on wavelet trans- form against the existing problems in the medical image fusion methods by wavelet transform. A fusion rule based on region standard variance is adopted in low frequency sub-band. A weighted local contrast fusion rule is adopted in high frequency sub-bands according to directional characteristics of wavelet. The human visual system is more sensitive to local contrast, so a selection scheme based on weighted local contrast is presented according to human visual properties, the rule is more accordance with the human physiological visual char- acteristics. The experiment results show that the proposed method has better fusion effects and quantifying indicators over conventional fusion methods.
出处
《兰州大学学报（自然科学版）》
CAS
CSCD
北大核心
2013年第1期122-125,130,共5页
Journal of Lanzhou University(Natural Sciences)
基金
国家自然科学基金项目(60962004,61162016)
甘肃省自然科学基金项目(0803RJZA015)
甘肃省科技支撑计划项目(0708GKCA047)
兰州市科技计划项目(2010-01-06)
关键词
小波变换
加权局部对比度
医学图像
图像融合
wavelet transform
weighted local contrast
medical image
image fusion
