灰色关联分析与模糊推理边缘检测图像融合法

为解决融合图像不同程度的光谱失真问题,提出了一种结合灰色关联分析、模糊推理和IHS变换的图像融合算法。首先通过灰色关联分析和模糊推理算出全色图像的边缘点和非边缘点,得到丰富的边缘信息,然后对多光谱图像进行IHS变换,以亮度分量为依据对全色图像进行直方图匹配,再基于边缘信息对亮度分量和直方图匹配后的全色图像进行线性加权,最后通过IHS逆变换得到融合图像。为验证本文方法的有效性,与5种常用方法比较,从视觉和定量两方面进行评价,且采用降尺度评价和全分辨率评价。结果表明,该方法得到的融合图像比其他5种方法更优越。本文方法不仅提高了遥感图像的空间分辨率,也较好保留了多光谱图像的光谱信息。

Image Fusion Based Edge Detection of Grey Relational Analysis and Fuzzy Inference

In order to solve the problem that the fused image produces different degree of spectral distortion,a fusion algorithm is proposed based on the gray relational analysis,fuzzy inference and IHS pansharpening method.First,the edge image of the panchromatic image(PAN)is obtained by the edge detection algorithm based on the gray relational analysis and fuzzy inference.Then,the multi-spectral image(MS)is performed by the IHS transform.The histogram matching is done on the PAN image referencing to the intensity of MS image.Based on edge information,the PAN image after histogram matching and the intensity of MS image is combined by using linear weighted.Finally,the fused image is obtained by using the inverse IHS transform.To verify the validity of this method,compared with other five methods,the visual and the quantitative evaluation are used to evaluate fusion results,including the reduced-resolution assessment and the full-resolution validation.The results show that the fusion image of the proposed method is better than that of other five traditional methods.The method not only improves the spatial resolution,but also keeps the spectral information of MS.