基于最佳小波包变换的遥感影像融合方法

Remote Sensing Image Fusion Based on Optimal Wavelet Packet Transform

从遥感影像的频率特性出发,提出了一种基于最佳小波包变换的影像融合方法。根据全色与多光谱影像的频率关系以及二进制小波包变换的特点,确定影像的最佳小波包分解形式;针对影像分解后的区域频率范围与特点进行融合,在融合中采用了基于投票表决法的多特征联合的融合策略,最后经小波包逆变换得到融合结果影像。该方法与传统融合方法进行了主观分析与客观定量比较,结果表明该方法具有良好的融合效果。

Remote sensing image fusion had become one of the hotspots in the researches and applications of remote sensing image processing in recent years. A new image fusion method based on optimal wavelet packet transform was proposed. The method utilized the frequency relationship between multispectral and panchromatic images and the frequency characteristics of wavelet packet transform to find the optimal form of wavelet packet transform. And weighted voting was used as fusion decision criteria in the process of image fusion. Compared with traditional image fusion methods, the proposed method has good fusion effects in the fusion experiment.