论文部分内容阅读
滑坡灾害是世界上破坏力极大的自然灾害之一,滑坡监测是预防滑坡灾害的重要基础。针对SIFT算法在滑坡遥感监测应用中存在特征点数量少、分布不均的问题,提出了一种将ASIFT算法和CSIFT算法相结合进行滑坡位移场标定的方法,即利用ASIFT处理全色影像,利用CSIFT处理多光谱融合影像,将两种方法提取的互不相同的特征点叠加综合得到更多的特征匹配点。该方法在露天矿滑坡的应用结果表明:与SIFT算法相比,新改进的方法大幅度增加了特征点数量,提高了滑坡位移场标定精度,且成本低、效率高,非常适合于大型滑坡后期大变形场监测。
Landslide disaster is one of the most destructive natural disasters in the world. Landslide monitoring is an important foundation for preventing landslide disasters. Aiming at the problems of small quantity and uneven distribution of feature points in landslide remote sensing monitoring applications, a new method of combining SIFT and CSIFT algorithm for landslide displacement field calibration is proposed, which uses ASIFT to process panchromatic images, CSIFT processes multispectral fusion images, and superposes and integrates the feature points extracted by the two methods to obtain more feature matching points. The application results of the method in the open pit landslide show that compared with the SIFT algorithm, the new improved method greatly increases the number of feature points and improves the calibration accuracy of the landslide displacement field with low cost and high efficiency, which is very suitable for large-scale landslide Large deformation field monitoring.