基于高分影像和ASIFT算法的滑坡位移场监测方法

针对利用高分遥感影像和SIFT算法进行大变形滑坡位移场监测时存在的不足,引入了一种具备完全仿射不变性的图像特征匹配算法(ASIFT),对不同时相高分遥感影像进行特征点提取与匹配,并在抚顺西露天矿特大型滑坡的位移场监测中进行了应用.结果表明,与SIFT算法相比,ASIFT算法弥补了特征提取算法在仿射不变性上的不足,增加了特征匹配点,提高了滑坡空间位移矢量场标定精度,且自动化程度高、成本低,非常适合于特大型滑坡的大变形位移场监测.

Monitoring Method of Landslide Displacement Field Based on High-Resolution Images and ASIFT Algorithm

In view of the shortcomings of displacement field of large deformation landslide with SIFT algorithm and high-resolution remote sensing images， a feature matching algorithm with fully affine invariability(ASIFT) was introduced to extract and match features on different phases of high-resolution remote sensing images. The Fushun West Open-Pit Mine was selected as a study area， and the displacement field was analyzed with the ASIFT algorithm. The results showed that compared with the SIFT algorithm， the deficiency of the feature extraction algorithm in affine invariance is compensated， the feature matching points increase and the accuracy of the displacement vector field of the landslide is improved by using ASIFT algorithm. The ASIFT algorithm is more suitable for monitoring the displacement field of large landslide with higher degree of automation and lower cost.