无人机核线影像的稀疏匹配与稠密匹配

发布时间：2018-05-14 15:58

  本文选题：SIFT算子 + 稀疏匹配　； 参考：《测绘通报》2017年05期

【摘要】：无人机影像转化为水平核线影像后,能够有效地减少同名点的搜索空间。在此基础上,本文使用SIFT算子进行了稀疏匹配,并用BP算法进行了稠密匹配。结果表明:(1)SIFT算子获取的同名点比较少,但是计算方法简单,同名点空间坐标精确,适用于大范围获取简要的空间三维信息;(2)BP算法计算复杂度高,可以获取地物大量的同名点,适用于小范围的地物三维重建。总体而言,两者各有优缺点,在实际的应用中可互补。
[Abstract]:After the UAV image is converted to the horizontal nuclear line image, it can effectively reduce the search space of the same name points. On this basis, this paper uses the SIFT operator to carry out a sparse match, and uses the BP algorithm to carry out a dense matching. The results show that: (1) the SIFT operator obtains less homonymous points, but the calculation method is simple, the space coordinates of the same name point are accurate and suitable. It is used to obtain a wide range of spatial information in a wide range; (2) the BP algorithm has high computational complexity and can obtain a large number of homonyms of ground objects. It is suitable for a small area of three-dimensional reconstruction of ground objects. In general, both have advantages and disadvantages, and can complement each other in practical applications.

【作者单位】： 北京信息职业技术学院;长江水利委员会长江科学院;中国地质大学(北京);
【基金】：国家自然科学基金(41601298)
【分类号】：P237;TP391.41
，

本文编号：1888539