论文部分内容阅读
由于成像条件与环境的差异,多景待镶嵌遥感影像之间往往会出现色彩差异,针对此问题,提出一种基于支持向量回归(SVR)的色彩一致性处理方法。采用NDVI(归一化植被指数)阈值分割并结合光谱角匹配(SAM)的方法在影像重叠区域自动选取具有“不变特征”的像素作为样本;通过SVR建立输入影像到参考影像的灰度值变换方程,并对输入影像进行处理,使得待镶嵌影像具有与参考影像相同或者相似的亮度与对比度。采用TM、SPOT、无人机(UAV)影像等多源数据进行了实验,结果表明,该方法能够有效消除由系统因素引起的色差,与线性回归方法相比,该算法在方差、辐射分辨率等方面具有优势。
Due to the differences between the imaging conditions and the environment, there are often color differences between the mosaic views and remote sensing images. To solve this problem, a color consistency method based on Support Vector Regression (SVR) is proposed. NDVI (Normalized Vegetation Index) threshold segmentation combined with spectral angle matching (SAM) was used to automatically select the pixels with “invariant feature” in the image overlap region as samples; the input image to the reference image gray Degree conversion equations, and the input image processing, making the image to be mosaic with the same or similar reference image brightness and contrast. The experimental results show that this method can effectively eliminate the chromatic aberration caused by the system factors. Compared with the linear regression method, the proposed method has better performance in terms of variance, radiation resolution Etc. have advantages.