一种新的面向对象Landsat-7 SLC-off影像插值修复算法  

Landsat-7 SLC-off Image Restoration Based on a New Object-oriented Interpolation Method

作　　者：孙悦[1] 张华[1] SUN Yue;ZHANG Hua(School of Environmental Science and Spatial Informatics,China University of Mining&Technology,Xuzhou,Jiangsu 221116,China)

机构地区：[1]中国矿业大学环境与测绘学院,江苏徐州221116

出　　处：《遥感信息》2018年第5期27-34,共8页Remote Sensing Information

基　　金：国家自然科学基金重点项目(41331175)

摘　　要：针对影像修复中基于规则窗口选择的相似像元不一定满足实际地物分布的问题,提出一种新的面向对象的像元插值法对ETM+SLC-off数据进行修复。首先对辅助影像进行多尺度分割,获取分割对象,然后将所获得的分割对象与待修复影像进行空间叠加,当待修复像元所在的分割对象中存在有效像元时,将同时利用所在分割对象区域中的有效像元及辅助影像中相应对象区域的像元进行插值;反之,使用辅助影像中相应对象区域的像元计算初步的待修复像元值,获取2幅影像间的局部最小二乘回归参数,对初步的待修复像元值应用回归参数得到最终待修复像元值。最后将其与全局直方图匹配法进行对比,并对修复结果进行定性与定量评价。实验结果表明,该方法修复精度更高,能够修复更多的空间细节信息,具有一定的可用性。For the problem of ETM+SLC-off image restoration,the selected similar pixels based on the sliding window does not necessarily satisfy the actual feature distribution.A new method named object-oriented pixel interpolator was proposed to restore ETM+SLC-off image.Firstly,multiresolution segmentation was applied in the auxiliary image to acquire segmented objects.Then,these objects were spatially overlaid with the image to be restored.When there exist valid pixels in the object that missing pixel belongs to,these valid pixels in corresponding object between auxiliary image and the image to be restored were used to recover the missing pixel.If there was no valid pixel in this object,the valid pixels in auxiliary image were used to calculate the preliminary missing pixel value.After acquiring the local least-square regression parameters between the auxiliary image and the image to be restored,the parameters were applied to calculate the ultimate value of missing pixel.In the experiment,the global histogram matching method and the proposed method were used to recover missing pixels and the performance was evaluated qualitatively and quantitatively.The results show that the proposed method can generate more accurate results and restore more spatial details,appearing a certain availability.

关 键 词：ETM+ SLC-off 面向对象 像元插值 条带修复 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]