摘要
针对多小波理论在图像处理领域的应用,给出了多小波域中图像的方向对比度的定义,并提出了基于多小波变换的、使用方向对比度进行自适应加权平均融合和选择性融合的算法。该算法利用多小波变换对原图像进行分解,根据得到的分解系数计算相应的对比度,然后建立基于对比度的融合规则。使用这些算法对多谱图像进行了融合实验,并给出了评价融合质量的对比结果。结果表明,基于多小波变换的方向对比度的图像融合算法在保留图像细节信息、增加信息量方面都有明显的提高。
Aiming at the application of multiwavelet theory in image processing field, the definition of image directional contrast in multiwavelet domain is given. In addition, an adaptive weighted-average fusion algorithm and selective fusion one which are based on multiwavelet transform and directional contrast are proposed. This algorithm decomposes original images by multiwavelet transform, then corresponding directional contrast are calculated according to decomposing coefficients, and the fusion rule based on contrast is set up. The experi- ments using these algorithms to multi-spectral image are done. Moreover, the comparative results of evaluating fusion quality are listed. The results show that image fusion algorithms based on directional contrast defined by multiwavelet transform perform obviously better in reserving image details of the information and increasing the amount of information.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2008年第10期1860-1863,共4页
Systems Engineering and Electronics
基金
总装预研重点基金资助课题(9140A01060507BQ01)
