小波多尺度分析不同融合准则的算法研究

Research on Fusion Rules Based on Wavelet Multiscale Analysis

在分析小波多尺度变换的基础上,对图像进行小波分解,在不同频率域,针对低频部分和高频部分的不同特点,采用不同的融合策略,保留其轮廓,增强其细节信息。在高频部分分别建立不同的融合算法(基于局部方差融合、基于局部能量融合、基于边缘算子融合),对不同聚焦的可见光图像以及红外图像进行融合实验,并且应用客观评价指标信息熵、交叉熵、标准差对融合结果进行分析比较。研究结果表明,基于边缘算子融合的效果优于基于局部方差融合和基于局部能量融合。

The wavelet decomposition was used to decompose original images to two different parts,namely,the low frequency part and the high frequency part based on the analysis of wavelet multi-scale transformation.To preserve the original image contours and boost up its details,different fusion strategies are applied in different frequency fields.To the fusion results of multi-focus visible images and infrared images by different fusion algorithms in high frequency,the comparison and analysis are given according to evaluation criteria put forward in the paper.