一种新的全色图像与光谱图像融合方法研究  

A New Spectral and Panchromatic Images Fusion Method

作　　者：赵永强[1] 潘泉[1] 张洪才[1] 

机构地区：[1]西北工业大学自动化学院,西安710072

出　　处：《光子学报》2007年第1期180-183,共4页Acta Photonica Sinica

基　　金：国家自然科学基金(60602056);国家遥感重点实验室开放基金(SK050013);航空科学基金(03D53032);武器装备预研基金(51401040204HK0359);西北工业大学英才计划资助

摘　　要：给出了一种新的综合伪彩色映射技术和小波变换理论的图像融合方法,并将其应用于全色图像和光谱图像的融合中.通过提取两个不同谱段光谱图像的共有信息和特别信息,并进行伪彩色映射融合,来增强目标与背景的对比度.同时将伪彩色映射融合后的图像进一步用IHS变换提取空间信息,在小波变换框架下将其与全色图像进行融合以提高目标的边缘细节信息,使所获得的融合结果不仅包含丰富的光谱信息的同时还具有较高的空间分辨率.仿真图和评价指标表明,该算法在增强目标与背景的对比度以及保留目标信息等方面具有较强的优势.To detect the object efficiently, the difference information presented by spectral images and detail information contained in panchromatic images are combined. A new false color mapping and wavelet integrated approach is presented, and is used to fuse the panchromatic image and spectral images. The contrast between object and background is enhanced through fusing common and distinctive information by false color mapping. Then false color image and panchromatic image are fused by IHS and wavelet transform to retain more detail information. The fusion results and evaluation index show that this algorithm can enhance the visibility of certain details and preserve the contrast of different materials. It also implies that the investigated approach can eventually be applied in object detection and identification.

关 键 词：图像处理 图像融合 伪彩色映射 小波变换 光谱图像 

分 类 号：TN911.73[电子电信—通信与信息系统]