基于局部熵的小波变换图像融合算法

研究基于小波变换的图像融合算法。首先对待融合图像进行小波多尺度分解,得到小波的分解系数,对高频分量和低频分量系数采取不同的融合规则。低频部分系数融合采用简单加权平均法,高频部分系数融合采用基于邻域熵值的方法。对融合后系数进行逆小波变换得到重构图像,融合后图像清晰、对比度好。与最大绝对值法、领域方差法作比较。最后通过熵、标准差、平均梯度值等参数对融合后的图像进行定量分析。实验数据表明提出的方法是切实可行的。

Wavelet Transform of Image Fusion Algorithm Based on Local Entropy

The paper researched the image fusion algorithm based on wavelet transform.Firstly,a wavelet multi-scale decomposition was performed on source images and wavelet coefficients could be obtained.Different fusion rules were applied to high-frequency and low-frequency coefficients.The low-frequency coefficient is based on simple plus mean method,and the high-frequency part is based on coefficient's neighborhood entropy.The fused image was reconstructed by performing inverse wavelet transform,the fused image had...