基于提升小波变换的多传感器图像融合算法研究

Multi-sensor Image Fusion Algorithm based on the Lifting Wavelet Transform

为了更好的进行图像融合,基于提升小波变换,采用了一种基于区域方差和方向对比度的融合规则相结合的图像融合新算法.该算法结合提升小波的优势,将图像进行多分辨率分解;针对变换后的低频分量和高频分量的不同特点,采用了不同的融合规则进行融合;最后通过提升小波逆变换得到融合图像.实验结果表明,该算法具有增强图像空间细节的能力,使得融合后的图像内容清晰,相比于传统小波变换法,具有更好的融合效果.

The paper proposes an image fusion algorithm based on the regional variance and directional contrast which can fuse image better based on the lifting wavelet transform.This algorithm combines the advantages of lifting wavelet transform to acquire the multi-resolution decomposition of the image;then according to the different characteristics of the low frequency components and high frequency components,the paper fuses the image using the different fusion rules;finally we can acquire the fused image through inverse transformation of lifting wavelet.Experiment results show that the images fused by the proposed algorithm are of better quality than that produced through the traditional algorithms with wavelet transform,since the algorithm can enhance the details of the image space and make the fused image clearer.