Object-oriented design for parallel processing of massive remote sensing data

When remote sensing data of earth observation reach the P level, distributed processing of global remote sensing data and improvement of digital image processing speed have become a research focus. In this article, a system of object-oriented metadata definition is developed for describing parameter inversion model and dependency of the remote sensing data. Based on the system, a space-time partitioning multi-granularity model is established and then applied to partitioning massive remote sensing data for parallel processing. Our results show that the model can provide data-based parallel processing and improve the capabilities of the massive remote sensing data processing.