directionlet域的多波段遥感图像融合算法研究 - 安徽建筑大学学报.docVIP

Directionlet域的多波段遥感图像融合算法研究 李彦1, 张德祥2 (1. 安徽职业技术学院，安徽 合肥 230011) (2. 安徽大学电气工程与自动化学院，安徽 合肥 230601) 摘要: 提出一种有效的基于Directionlet变换的多波段遥感图像融合算法。Directionlet变换是一种新的基于格子的歪斜多尺度多方向各向异性小波变换工具，具有更高的逼近精度和更好的稀疏表达能力的图像分解变换，与小波变换相比，抑制了小波变换在图像边缘方向表示的固有局限性。首先采用具有多尺度、多方向特点的Directionlet变换对多波段遥感图像进行分解。对于低频系数采用平均融合算法，方向高频系数采用区域边缘检测实现多波段遥感图像的融合处理。实验结果表明，提出的算法与传统融合算法相比不仅原始图像的边缘和纹理信息可保留，而且可获得更好的融合视觉效果。 关键词: Directionlet变换; 多波段遥感图像; 区域边缘检测; 图像融合 中图法分类号: TN957.54　　 文献标识码: A Research on Fusion Algorithm of Multi-band Remote Sensing Image in Directionlet Domain LI Yan 1，ZHANG De-xiang 2 (1. Anhui Vocational and Technical College, Hefei, 230011, China； 2. The school of Electrical Engineering and Automation, Anhui University, Hefei, 230601, China) Abstract: An efficient fusion algorithm for multi-band remote sensing images based on directionlet transform is proposed. Directionlets transform is a new lattice based skewed multi-scale and multi-directional anisotropic wavelet transform tool with better approximation precision and sparse description compare with wavelet transform, and it has overcome some inherent limitations of wavelet in representing directions of edges in image. Firstly, the several of muliband remote sensing images are decomposed using directionlet transform, which have multi-scale, multi-direction characteristics. Then, for the low-pass coefficients, an averaging fusion rule is used. And the directional high-frequency coefficients are used to select the better coefficients by region edge detection algorithm for fusion. Experimental results show that compared with traditional algorithm, the proposed algorithm can get better visual effect and the significant information of original image like textures and contour details is well maintained. Key words: Directionlet transform; muliband remote sensing image; region edge detection; image fusion 0 引言随着现代科技的发展，不同类型、不同分辨率以及不同功能的传感器不断地出现，使得信息的获取通道更多，地物的特征信息在不同传感器下可以表现出不同侧面的信息表征。合成孔径雷达SAR（Synthetic Aperture Radar）遥感图像能提供多尺度多频段