欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
多尺度分割是高分辨率遥感信息计算的重要基础，是高分辨率遥感影像图谱认知中“图”提取的关键技术。当前提出的多尺度分割方法普遍存在着占用内存大，耗费计算资源、计算时间长的缺点，并且这些问题随着遥感数据量的增大、算法的改进等进一步加剧。针对这种情况，根据当前集群计算技术的发展，以均值漂移的多尺度分割方法为例，实现了一种基于集群计算环境的多尺度分割算法，集中解决任务分配和结果回收以及数据并行的方式，统计了算法所消耗的时间，对其的效率进行了分析，通过实验说明了集群化对提高多尺度分割效率的有效性。
英文摘要：
Multi_scales segmentation is important basis for high resolution RS information computation and key technologies for graphics information extraction.The existed multi_scales segmentation algorithms are usually memory cost,computation-intensive. What＇s more,these problems will become serious as the data accumulating and algorithms improving.To solve these problems,a parallel algorithm for mean shift multi_scales segmentation based on cluster is proposed and implemented,statistics the processing time ,then analyzes and proves the effectiveness of the algorithm.
同期刊论文项目
同项目期刊论文
Bosten lake wetlands water level automatic retrieval and trends analysis based on ICESat GLAS global
ESTIMATION OF IMPERVIOUS SURFACE BASED ON INTEGRATED ANALYSIS OF CLASSIFICATION AND REGRESSION BY US
Wetland mapping in the Balqash Lake Basin Using Multi-source Remote Sensing Data and Topographic fea
A New Approach to Improve the Cluster-based Parallel Processing Efficiency of High-Resolution Remote
位置：