Feature extraction and tracking is a fundamental operation used in many geoscience applications. In this paper, we present a scalable method for computing and tracking features on distributed memory machines for large-scale geospatial data. We carefully apply new communication schemes to minimize the data exchanged among the computing nodes in building and updating the global connectivity information of features. We present a theoretical complexity analysis, and show that our method can significantly reduce the communication cost compared to the traditional method.
L. Yu, F. Zhu, H. Yu, J. Wang and K. S. Kuo, "Feature extraction and tracking for large-scale geospatial data," 2016 IEEE International Geoscience and Remote Sensing Symposium (IGARSS), Beijing, 2016, pp. 1504-1507. doi: 10.1109/IGARSS.2016.7729384
This material is based upon work supported by the National Science Foundation under Grant No. 1541043. Opinions, findings, conclusions or recommendations expressed are those of the authors and do not reflect the views of the NSF.