局部相位特征描述的多源遥感影像自动匹配

发布时间：2018-04-07 18:43

本文选题：遥感影像　切入点：影像匹配　出处：《武汉大学学报(信息科学版)》2017年09期

【摘要】：由于影像间显著的几何和辐射差异,多源遥感影像自动匹配一直是目前研究的难点问题。首先引入具有光照和对比度不变性的相位一致性模型,并对其进行扩展,构建相位一致性的特征方向信息,然后借助于梯度方向直方图的模板结构,利用其特征值和特征方向,建立一种局部特征描述符——局部相位一致性方向直方图(local histogram of orientated phase congruency,LHOPC),最后利用欧氏距离作为匹配测度进行同点名识别。对四组多源遥感影像进行试验,其结果表明,相比于尺度不变特征转换和加速鲁棒性特征算法,LHOPC能更为有效的抵抗影像间的辐射差异,提高了匹配性能。
[Abstract]:Due to the significant geometric and radiative differences between images, automatic matching of multi-source remote sensing images has been a difficult problem.Firstly, the phase consistency model with illumination and contrast invariance is introduced and extended to construct the feature direction information of phase consistency. Then, with the help of the template structure of gradient direction histogram, the characteristic value and characteristic direction are used.A local histogram of orientated phase congestion LHOPC-local phase consistent direction histogram is established. Finally, Euclidean distance is used as a matching measure to identify the same name.Four groups of multi-source remote sensing images are tested. The results show that LHOPC is more effective than scaling invariant feature conversion and accelerated robust feature algorithm to resist the radiative difference between images and improve the matching performance.
【作者单位】：西南交通大学高速铁路运营安全空间信息技术国家地方联合工程实验室;
【基金】：国家重点基础研究发展计划(2012CB719001) 国家自然科学基金(41401369,41401374) 测绘遥感信息工程国家重点实验室资助项目(14E01) 地理国情监测国家测绘地理信息局重点实验室开放基金(2014NGCM13) 四川省科技计划项目(2015SZ0046)~~
【分类号】：TP751

【相似文献】