An Adaptive Geographic Meshing and Coding Method for Remote Sensing Data
Xiaohui Huang,Jiabao Li,Jining Yan,Lizhe Wang
DOI: https://doi.org/10.1088/1755-1315/1004/1/012006
2022-01-01
IOP Conference Series Earth and Environmental Science
Abstract:Spatial indexing techniques, inherently data structures, are generally used in portals opened by institutions or organizations to efficiently filter RS images according to their spatial extent, thus providing researchers with fast Remote Sensing (RS) image data discovery ability. Specifically, space-based spatial indexing approaches are widely adopted to index RS images in distributed environments by mapping RS images in two-dimensional space into several one-dimensional spatial codes. However, current spatial indexing approaches still suffer from the boundary objects problem, which leads to multiple spatial codes for a boundary-crossing RS image and thus alleviates the performance of spatial indexes built on top of these spatial codes. To solve this problem, we propose an adaptive geographic meshing and coding method (AGMD) by combining the famous subdivision model GeoSOT and XZ-ordering to generate only one spatial code for RS images with different spatial widths. Then, we implement our proposed method with a unified big data programming model, (i.e., Apache Beam), to enable its execution in various distributed computing engines (e.g., MapReduce, and Apache Spark, etc.) in distributed environments. Finally, we conduct a series of experiments on real datasets, the archived Landsat metadata collection in level 2. The results show that the proposed AGMD method performs well on metrics, including the following aspects: the effectiveness of the storage overhead and the time cost are up to 359.7% and 58.02 %, respectively.