论文部分内容阅读
利用SIFT算法对无人机影像进行匹配,采用距离阈值的方法剔除匹配后距离较近的特征点。为了降低误差累积对影像拼接效果的影响,提出了最佳基准影像的选择方法,同时将Dijkstra算法引入到拼接路径的搜索中,利用影像间的投影转换误差构建搜索权阵。结果表明:最优路径的拼接方法能有效避免投影误差较大的影像对后续拼接影像的影响,减少了影像投影转换次数,改善了影像的拼接效果。
SIFT algorithm is used to match the UAV image, and the distance threshold is used to remove the feature points which are close to each other. In order to reduce the effect of error accumulation on the mosaic effect of images, a method of selecting the best reference image is proposed. At the same time, Dijkstra algorithm is introduced into the searching of splicing paths, and the search matrix is constructed by using projection transformation error between images. The results show that the optimal path splicing method can effectively avoid the impact of the image with larger projection error on subsequent mosaic images, reduce the number of image projection transformations, and improve the mosaic effect of the images.