TM多光谱与SPOT 全色遥感图像融合算法对比

融合技术是遥感数据处理中一种重要的方法。而TM多光谱与SPOT全色图像是遥感融合最为普遍的选择。为了对比分析不同方法在融合TM多光谱与SPOT全色图像上的效果,提出基于色彩空间的HSV变换、基于算数技术的Brovey变换和Gram—Schmidt波谱锐化3种融合方法相结合,实现了对同一传感器的全色和多光谱数据融合。试验表明:就空间信息量而言,经过HSV变换的图像具有最大的空间信息,但其光谱保真能力最差;Brovey变换最大限度保持了原始图像的光谱信息,而空间信息的详细程度较差;Gram-Schmidt波谱锐化后的影像不仅保持了多光谱影像的光谱信息,同时又保持了高光谱全色影像的空间细节信息,是一种较好的图像融合方法。

Comparison of Multi-Spectral and Panchromatic Remote Sensing Image Fusion Algorithm of TM&SPOT

The fusion technology is remote sensing data processing an important method.The TM multi-spectral and panchromatic SPOT images are the most popular choice of remote sensing.Comparative analysis of different methods for integration of TM multi-spectral and panchromatic SPOT image with the effect of the proposed change based on HSV color space,technology Brovey arithmetic transform and Gram-Schmidt spectral sharpening three fusion combined to achieve the same panchromatic and multi-spectral sensor data fusion.The assessment of experimental results indicates that： the amount of information in terms of space,after HSV transform the image with the largest spatial information,but its spectral fidelity have the least capacity;Brovey transform the original image to maintain the maximum spectral information,while the lower level of detail of spatial information;Gram-Schmidt spectral sharpening the image after the multi-spectral image not only maintain the spectral information,while maintaining a high spectral panchromatic image spatial detail information,is a better image fusion method.