邵振峰, 刘军, 李德仁. 一种基于高斯影像立方体的空间投影融合方法[J]. 武汉大学学报 ( 信息科学版), 2010, 35(10): 1207-1211.
引用本文: 邵振峰, 刘军, 李德仁. 一种基于高斯影像立方体的空间投影融合方法[J]. 武汉大学学报 ( 信息科学版), 2010, 35(10): 1207-1211.
SHAO Zhenfeng, LIU Jun, LI Deren. A New Spatial Projection Image Fusion Method Based on Gaussian Image Cube[J]. Geomatics and Information Science of Wuhan University, 2010, 35(10): 1207-1211.
Citation: SHAO Zhenfeng, LIU Jun, LI Deren. A New Spatial Projection Image Fusion Method Based on Gaussian Image Cube[J]. Geomatics and Information Science of Wuhan University, 2010, 35(10): 1207-1211.

一种基于高斯影像立方体的空间投影融合方法

A New Spatial Projection Image Fusion Method Based on Gaussian Image Cube

  • 摘要: 为了使融合影像能够最大限度地保持原始多光谱影像的光谱特征,同时又能尽可能地提高融合影像的空间分辨率,引入了模拟人眼视觉机理的高斯尺度空间技术,提出了一种基于高斯影像立方体的空间投影融合方法。该算法在避免光谱扭曲的同时,又能提高融合影像的空间分辨率。采用QuickBird遥感影像进行的实验结果表明,本方法在光谱偏差指数、相关系数、UIQI、ERGAS、运行时间、主观视觉感受等指标上较传统方法有显著优势。

     

    Abstract: To make sure that the fused multispectral image keeps greatly the spectral characteristics of the original multispectral(MS) image and most likely to improve the spatial resolution,this paper introduces the technology of Gaussian scale space that can simulate the mechanism of human visual and proposes a spatial projection method based on Gaussian image cube.The proposed method can improve the spatial resolution of the fused MS image effectively,while it can make little spectrum distortion to the fused images.The experimental results of fusing QuickBird remote sensing images show that,comparing to traditional methods,the method proposed in this paper has a distinct advantage in subjective visual experience,spectral deviation,correlation coefficient,UIQI,ERGAS and Run-time indicators.

     

/

返回文章
返回