多源遥感影像融合的图像质量评价及变化区域识别

Quality evaluation of fusion images and identification of dynamic information area based on multi-resources remotely sensed imagery

遥感技术的发展以及遥感数据的广泛应用,使得影像信息及其变化信息的自动识别的需求日益迫切.论文利用TM(1998)、SPOT(2002)、ALOS(2006)3个时相的不同传感器接收的影像数据,进行小波变换与IHS变换相结合的影像融合处理,并进行融合影像的评价以及变化区域的识别研究,得出以下主要结论:小波变换与IHS变换相结合的影像融合方法既提高了影像的信息量,又较好地保持了影像的光谱特征;融合影像信息量的增加程度与原始多光谱影像的空间分辨率相关,原始多光谱分辨率越低,融合后信息量增加越多;从融合影像上还可以直观地看出,不同时相的影像融合时,有利于地物变化区域的自动显现,以此进一步地进行变化区域的自动提取与分类,促进遥感影像的应用.

With the development of remotely sensed technology and universal application throughout all fields, efficient extraction of land use/land cover information is urgent. In the paper,three frames of remotely sensed imagery, including TM(1998) , SPOT (2002) and ALOS (2006) ,are selected as the source data. A fusion method, which is the integration of wavelet transform with IHS, is used to improve image spatial detail and to detect the change information between multispectral bands and panchromatic band at diffe...