遥感影像并行分割的无缝拼接算法

发布时间：2018-05-05 11:19

本文选题：遥感影像 + 并行分割　；参考：《测绘科学》2017年04期

【摘要】：针对影像数据分块并行处理造成的接边问题,该文基于构建缓冲区的方法提出了一种新的拼接算法。该算法只在每一个数据分块的右侧和下侧构建缓冲区,消除拼接线的原则是保证分割边界的准确性和拓扑关系的正确性。数据分块完成后采用并行的策略对第一个块进行分割,然后将每一块的分割结果与相邻的块进行拼接,在缓冲区内按一定规则判断各块分割对象是否保留并对拓扑关系进行检查,拼接完各块后并行输出对象矢量化后的结果。通过实验验证,该算法不仅保证了分割边界的准确性,而且能够处理大影像,提高了工作效率。
[Abstract]:In order to solve the problem of edge joining caused by image data block parallel processing, this paper proposes a new stitching algorithm based on the method of constructing buffer. The algorithm only constructs buffers on the right and lower sides of each data block. The principle of eliminating the splicing lines is to ensure the accuracy of the segmentation boundary and the correctness of the topological relationship. After data partitioning is completed, the first block is segmented by a parallel strategy, and then the segmentation results of each block are spliced with adjacent blocks. According to certain rules in the buffer zone, the block partition object is determined whether to retain or not, and the topological relation is checked, and the result of the parallel output object vectorization is obtained after each block is stitched together. The experimental results show that the algorithm not only ensures the accuracy of the segmentation boundary, but also can process large images and improve the working efficiency.
【作者单位】：山东城市建设职业学院;山东省第一地质矿产勘查院;
【分类号】：TP751

【相似文献】