Remote sensing image fusion assessment based on SVD
-
摘要: 研究遥感图像融合性能的客观评价问题,分析当前遥感图像融合效果评估方法特点的基础上,提出一种新的遥感图像融合效果评估方法——基于奇异值分解(SVD, Singular Value Decomposition)的方法.利用源图像与融合结果图像的奇异值差异,测量它们之间能量信息失真情况,从而进行融合算法的评估.仿真实验从两方面入手,当融合源图像中含有SAR(Synthetic Aperlure Radar)图像时,对比Piella和Xydeas评估方法较为有效;另一方面,对多类型传感器、不同方法的像素级融合结果进行评估,与主观评价结果对比具有较高的一致性.这种客观评估方法能够较好地反映多类遥感图像融合的质量,是一种实现简单、高效、较为通用的遥感图像融合效果评估方法.Abstract: Existent measures of fused image quality were analyzed to research the objective evaluation problem of remote sensing image fusion. A new approach based on the singular value decomposition (SVD) was proposed for the remote sensing image fusion assessment. This method measured the divergence of the singular value features between the source images and fused image, and calculated energy distortion of fused image from source images. By that means, effect of fusion algorithm was measured. Experiments were conducted from two aspects to confirm the idea. When the source images including SAR image, this method is more effective than the Piella-s evaluation methods and Xydeas-s evaluation methods. In the other hand, the experiments of different kinds of sensors and pixel-level fusion algorithms show that the objective evaluation appears highly consistent with the subjective evaluation. It is an effective and universal assessment method with high coherence to the subjective factors.
-
Key words:
- remote sensing /
- fuses /
- image quality /
- evaluation
-
[1] 夏明革,何友,唐小明,等.多传感器图像融合综述[J].电光与控制,2002, 9(4):1-7 Xia Mingge, He You, Tang Xiaoming, et al. A survey on multisensor image fusion[J]. Electronics Optics & Control, 2002, 9(4):1-7 [2] Wang Z, Bovik A. A universal image quality index [J]. IEEE Signal Processing Letters, 2002, 9 (3) :81-84 [3] Piella G, Heijmans H. A new quality metric for image fusion //H knutsson. IEEE International Conference on Image Processing. Barcelona,Spain:IEEE,2003: 173-176 [4] Xydeas C S, Petrovid V. Objective image fusion performance measure[J]. Electronics Letters, 2000,36(4):308-309 [5] 甘俊英,张有为.一种基于奇异值特征的神经网络人脸识别新途径[J].电子学报, 2004 ,32(1):170-173 Gan Junying, Zhang Youwei. A new approach for face recognition based on singular value feature s and neural networks[J]. ACTA Electronica Sinica, 2004 ,32(1):170-173 [6] Shnayderman A, Gusev A, Eskicioglu A M. A SVD based grayscale image quality measure for local and global assessment [J]. IEEE Transactions on Image Process, 2005, 14 (2):422-429
点击查看大图
计量
- 文章访问数: 3462
- HTML全文浏览量: 243
- PDF下载量: 1880
- 被引次数: 0