“资源三号”卫星图像影像特征匹配方法研究

随着中国遥感卫星的迅速发展,要求影像几何质量评价方法可以在待评图像和参考图像间提取出精确且分布均匀的控制点信息。文章提出一种基于多源、高精度遥感图像的特征点匹配方法。该方法首先用加速稳健特征（Speeded Up Robust Features,Surf）算法对＂资源三号＂卫星图像和参考图像进行粗匹配以建立两幅图像间的整体几何关系,通过对相同区域进行wallis滤波增强图像纹理信息,然后用Surf算法进行特征点的提取和匹配,最后利用对极几何约束剔除误匹配点。试验结果表明,该方法可以全自动、快速和精确的提取影像控制点。

Research on Feature Matching Algorithm for ZY-3 Image

With the rapid requested to extract precise and development of remote sensing satellite,image quality evaluation method is even-distributed control points. This paper proposes a feature matching method for the multi-source, high-precision remote sensing images. Firstly,we use Surf algorithm to get geometric relationship between the reference image and the ZY-3 image;secondly,Wallis filter is used to enhance image texture in the same area;then Surf algorithm is used to extract and match the control points;finally we use the geometric constraints to remove the mismatched points. points automatically with high accuracy and computation Experiment shows that our method can extracts control efficiency.