基于Haar小波变换的图像融合方法  

Image Fusion Method of Wavelet Transform Based on Haar Kernel

作　　者：何宏[1] 林剑[1] 

机构地区：[1]杭州电子科技大学科技处,浙江杭州310018

出　　处：《杭州电子科技大学学报（自然科学版）》2012年第2期54-57,共4页Journal of Hangzhou Dianzi University：Natural Sciences

基　　金：浙江省自然科学基金资助项目(Y1100771);浙江省教育厅科研计划资助项目(Y201016676;Y201120623)

摘　　要：为了有效的提高多个传感器的图像融合精度,该文提出了基于Haar小波变换的图像融合方法,首先分析了小波变换中不同频率分量对图像融合精度的影响,然后详细探讨了高频分量系数的确定方法。选取信息熵作为图像融合算法性能的评价指标,通过仿真实验定量分析了高频分量系数对图像融合精度的影响,实验结果表明高频分量系数并非越大越好,应根据融合后的图像信息熵确定高频分量系数。In order to effectively improve the image fusion accuracy of multi-sensors,an image fusion method of wavelet transform based on Haar kernel was proposed in this paper.Firstly the image fusion accuracy affected by different frequencies of wavelet transform was analyses.Then the method of determining the high frequency coefficients was discussed in detail.The information entropy was selected to evaluate the performance of the image fusion method,an emulating experiments was implemented to quantitatively analysis the image fusion accuracy affected by high frequencies.The experiment results indicated that the high frequency coefficients should be not too large,it should be determined according to the information entropy of the fused image.

关 键 词：图像融合 小波变换 信息熵 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]