一种基于子矩阵灰阶回代的亚像元分解与增强方法  

Sub Pixel Decomposition and Enhancement Based on Back Substitution of Grey Scale Values in the Sub-Matrix

作　　者：王艳霞[1] 龙晓敏[2] 丁琨[3] 周汝良[4] 

机构地区：[1]西南林业大学云南生物多样性研究院,云南昆明650224 [2]云南农业大学水利学院,云南昆明650224 [3]西南林业大学理学院,云南昆明650224 [4]西南林业大学生态旅游学院,云南昆明650224

出　　处：《地理与地理信息科学》2016年第6期12-17,共6页Geography and Geo-Information Science

基　　金：国家林业公益性行业科研专项项目(201404402);云南省科技创新人才计划项目(2014HC014);云南省教育厅资助项目(2015Z135);国家自然科学基金项目(41261085)

摘　　要：从任意显示图像的可视化需求出发,针对普通插值法存在的马赛克效应、信息丢失及图像模糊的问题,该文提出了一种基于子矩阵灰阶回代的亚像元分解与增强方法。该方法以分解前图像为回代的子矩阵,将普通内插与子矩阵灰阶回代协同处理,通过对数字图像进行一次亚像元分解后,同步回代分解前图像的中心像元灰阶(子矩阵灰阶回代),再经多次插值分解、回代、平滑和拉普拉斯增强处理,得到目标像元大小的亚像元分解与增强图像。以DEM、Landsat TM影像、Google Earth发布的影像等为实例,分解和增强后的可视化结果表明,该方法在数字图像放大时,不仅能解决马赛克效应和图像模糊问题,还可提高图像的可视化水平和图像的可读性。From the perspective of meeting the need of the random image visualization,this paper proposed a method of sub-pixel decomposition and enhancement based on back substitution of grey scale values in the sub-matrix.The method was for solving some problems of ordinary interpolation methods when displaying images in a large scale,such as mosaic effect,information loss and image blurring.The study considered a pre-decomposition image as one sub matrix.After decomposing by one of ordinary interpolation methods,the central pixels of the decomposed image were substituted by the grey scale values in the sub-matrix.Specifically,once digital images were done decomposition,the pixels of the sub-matrix were substituted back.Thus,interpolation and decomposition,substitution back were looped for many times until the image with the targeting sub-pixel resolution was attained with cooperative processing of image smoothing and enhancement by Laplace Transform.With the application examples of DEM(Digital Elevation Model),images of Landsat TM,Google Earth,etc.,the results showed that the method solved the problems of ordinary interpolations,effectively improved the level of image visualization and increased the readability of remote sensing images.

关 键 词：子矩阵 灰阶回代 亚像元分解 图像增强 

分 类 号：TP751.1[自动化与计算机技术—检测技术与自动化装置]