利用YCbCr和SWT实现彩色图像融合  

Color Image Fusion Utilizing YCbCr Transform and SWT

作　　者：李婵飞[1] 邓奕[1] LI Chanfei;DENG Yi(HanKou University,Wuhan 430212,China)

机构地区：[1]汉口学院,湖北武汉430212

出　　处：《红外技术》2018年第7期654-659,共6页Infrared Technology

基　　金：湖北省教育厅科学研究计划指导性项目(B2017311)

摘　　要：针对彩色红外图像和可见光图像融合问题,提出了利用YCbCr和平稳小波变换(Stationary Wavelet Transform,SWT)实现彩色图像融合的算法。该算法首先将源彩色红外图像转换为灰度图像,将彩色可见光图像从RGB空间转换到YCbCr空间获取Y、Cb、Cr三分量,然后将红外灰度图像和Y分量利用SWT融合。在融合过程中应用了一种绝对值阈值和区域能量相结合的融合准则,并利用彩色可见光图像Y分量对融合结果进行区域线性变换获取彩色融合图像Y分量。最后将Cb、Cr和融合Y分量从YCbCr空间转换到RGB空间获得彩色融合图像。实验结果表明,该算法可以综合YCbCr空间、SWT和区域线性变换优点,较好地保持源图像场景信息、空间色彩信息和目标人物特性,突出源图像的纹理变化特性和边缘细节信息。A method is proposed for the fusion of color-infrared and visible-light image based on the YCbCr space and SWT.The source color-infrared image is first converted into a grayscale image.To obtain the components of Y,Cb,and Cr,the visible-light image is converted from RGB space to YCbCr space.The infrared grayscale image and Y component are then fused by SWT to obtain a new Y component.The finalY component is obtained by local linear transformation of this Y component according to the region feature of the visible-light image’s Y component.Finally,the fused color image is obtained by the inverse YCbCr transform of Cb,Cr,and Y components.The simulation results demonstrate that the method can well preserve the scene information,color information,and target character.Moreover,the texture characteristics and edge information of the source image are maintained.

关 键 词：YCBCR SWT 线性变换 彩色图像融合 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]