摘要
针对同一场景红外图像与可见光图像的融合问题,提出了一种基于二代curvelet与wavelet变换的自适应图像融合算法。首先对源图像进行快速离散curvelet变换,得到不同尺度与方向下的粗尺度系数和细尺度系数;根据红外图像与可见光图像的不同物理特性以及人类视觉系统特性,对不同尺度与方向下的粗尺度系数和细尺度系数采用基于离散小波变换的图像融合方法,在小波域中,对低频系数采用基于红外图像与可见光图像的不同物理特性的自适应融合规则,对高频系数采用基于邻域方向对比度与局部区域匹配度相结合的自适应融合规则,然后进行小波逆变换得到融合的curvelet系数;最后,进行快速离散curvelet逆变换得到融合图像。实验结果表明,该方法能够更加有效、准确地提取图像中的特征,是一种有效可行的图像融合算法。
For the fusion problem of infrared and visible light images with the same scene,an adaptive image fusion algorithm based on the second generation curvelet and wavelet transform is proposed.Firstly,source images are decomposed by the fast discrete curvelet transform,thus the coarse scale and fine scale coefficients are obtained at different scales and in various directions.Secondly,according to the different physical features of infrared and visible light images and human visual system features,the coarse scale and fine scale coefficients are fused using image fusion method based on discrete wavelet transform.In wavelet domain,for the low frequency coefficients,we present an adaptive fusion rule based on the physical features of infrared and visible light images;while for the high frequency coefficients,we present an adaptive fusion rule based on the neighborhood directional contrast combined with the local area matching.Fused curvelet coefficients are obtained through the inverse wavelet transform.Finally,the fusion image is obtained through the inverse fast discrete curvelet transform.The experimental results illustrate that the proposed algorithm is effective for extracting the characteristics of the original images.
出处
《激光与红外》
CAS
CSCD
北大核心
2010年第9期1010-1016,共7页
Laser & Infrared
基金
江苏省自然科学基金项目(No.BK20080544)资助
