基于边缘检测和小波变换的遥感图像融合算法

Algorithm for Remote Sensing Image Fusion Based on Edge Detection and Wavelet Transformation

针对多光谱图像和全色图像的特点,提出一种边缘检测和小波变换相结合的遥感图像融合方法。该方法在传统图像小波变换的基础上,选择Canny算子对图像进行边缘检测。在小波域中,在各个尺度层对高频子带采用边缘检测,将边缘点完整保留,低频子带利用加权法,再进行小波逆变换重构融合图像。实验结果显示,该方法在保证光谱信息的同时,能有效地突出边缘细节,更好地保持图像的空间分辨力。与传统小波变换法遥感图像融合相比,信息熵提高了6.63%,清晰度提高了32%,相关系数提高了0.36%。

According to the characteristic of multi-spectral image and panchromatic image, an algorithm for remote sensing image fusion based on edge detection and wavelet transformation is proposed. This approach determines the image edge positions with Canny operator based on traditional wavelet transform algorithm. At each level in wavelet domain, edge detection is used in high frequency subbands to preserve the edge points perfectly. Average-weighted mean is used for low frequency coefficient fusion and then the fusion image is recon-structed by wavelet inverse transform . Experimental results show that the method can not only preserve spectral information but also effectively enhance the edge details and improve the spatial resolution of the image . Compared with the traditional wavelet transform method of remote sensing image fusion ,entropy increases by 6. 63%,definition increases by 32% and correlation coefficient increases by 0. 36%.