A Framelet-Based SFIM Method to Pan-Sharpen THEOS Imagery
- 13 Downloads
This paper proposes an improved framelet-based pan-sharpening algorithm for Thailand Earth Observation System (THEOS) imagery to decrease the effects of different acquisition times between panchromatic (Pan) and multispectral (MS) images, in which the smoothing filter-based intensity modulation (SFIM) is introduced into low-frequency information fusion instead of the conventional “mean” rule. Moreover, a two-layer procedure is presented to reduce the impacts of mixed pixels caused by the large difference of spatial resolutions between the Pan and MS images. The proposed method is tested on two THEOS datasets and compared with the Gram–Schmidt, SFIM and traditional framelet-based methods. The portability across contourlet transform is also examined. Both qualitative and quantitative evaluation results demonstrate that the proposed method is more independent of the illumination of the Pan image and can achieve better spectral fidelity while maintaining spatial sharpness.
KeywordsTHEOS Pan-sharpening Framelet SFIM Spectral fidelity
Funding was provided by the Fundamental Research Funds for the Central Universities (Grant No. 2015XKMS050), and the National Natural Science Foundation of China (Grant Nos. 41571330, 41601453).
- Choi, M. J., Lee, D. H., & Lim, H. S. (2008). Framelet-based multiresolution image fusion with an improved intensity-hue-saturation transform. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 37, 1273–1280.Google Scholar
- Mallika, K., Arathi, T., Krishna Rao, G. V., & Soman, K. P. (2009). Framelet based image fusion for the enhancement of cloud associated shadow areas in satellite images. In International conference on advances in computing, control, and telecommunication technologies (pp. 507–509). IEEE.Google Scholar
- Zheng, H., Zheng, D., Hu, Y., & Li, S. (2010). Study on the optimal parameters of image fusion based on wavelet transform. Journal of Computational Information Systems, 6(1), 131–137.Google Scholar