摘要
针对海量国产卫星数据使用过程中优质数据筛选困难且费时费力的问题,设计并实现自动化优质数据筛选.构建基于资源一号02C、高分一号、高分二号等多源国产卫星影像智能优选模型,将卫星及传感器类型、空间范围、时相、空间分辨率和空间分析方法等各项指标要求抽象为数学模型,完成异构元数据的高效集成与管理,制定遥感影像评价指标体系和评价模型,面向用户的定制化需求,通过自适应优选规则和自主性权重设置,运用典型运筹学方法进行定量分析,实现目标区域的卫星影像自动最优化覆盖.实验结果表明,自动优选与人工精选的结果重合率在85%以上,执行效率提高了10倍以上,验证了自动优选方法的正确性和高效性.
In view of the difficulty and time-consuming problem of high quality data filtering in the utilization of massive domestic satellite image data, the authors designed and realized automatic high quality data filtering. The intelligent optimization model of domestic satellite image was constructed based on ZY1-02C,GF-1, GF-2. The requirements of satellite and sensor type, space range, time, spatial resolution and spatial analysis method were abstracted into mathematical model to complete the efficient integration and management of heterogeneous metadata. The evaluation index system and evaluation model of remote sensing image were developed, and the customization requirements of users were set up. By means of adaptive preference rules and autonomous weight setting, the typical operational research methods were used to quantitatively analyze the satellite images so as to optimize the coverage of satellite images. The experimental results show that the coincidence rate of automatic selection and artificial selection is higher than 85%, and the efficiency of implementation is improved by more than 10 times, which verifies the correctness and efficiency of the automatic optimization method.
出处
《国土资源遥感》
CSCD
北大核心
2017年第B10期13-20,共8页
Remote Sensing for Land & Resources
关键词
多源海量
国产卫星数据
优质数据自动筛选
影像智能优选
multi - source and massive
domestic satellite image data
high quality data automatic filtering
intel-ligent optimization of image
