Research And Application Of A Parallel Model On GF Satellite Image Preprocessing

With the rapid development of Chinese remote sensing technology and the rapid construction of the integrated observational network of air,space,and earth,the spatial-temporal resolution,spectral resolution,and observation angles of remote sensing images are increasingly accurate,the count of remote sensing data is exponentially increasing.Remote sensing images are playing an increasingly important role in social life and economic construction,and the demand for remote sensing products in various fields is also increasing.What is more,the demand for data timeliness is gradually increasing,especially in the areas of disaster warning,environmental detection,military investigation,and so on,which caused tremendous pressure on remote sensing data processing.In the flow of remote sensing data processing,remote sensing data preprocessing is a necessary part of each remote sensing data processing,and plays an important role in the production of subsequent remote sensing products.In some degree,improving the efficiency of remote sensing preprocessing can relieve the pressure of remote sensing data processing.With the continuous development of observation techniques,many types of sensors generate different types of remote sensing data,and multi-sourced remote sensing data makes it difficult to develop general-purpose preprocessing systems.Based on the above issues,the main research work and innovations of this paper are as follows:(1)The remote sensing image preprocessing system is designed.The entire system architecture is composed of five layers.It fully considers the system scalability and ease of use while maintaining the independence of each functional module.The system-predefined and user-defined processes based on process visualization solve the data processing tasks of the system in different situations.Combined with the algorithm call interface in the XML file model database,the formed XML interface can support batch processing of different production processes.(2)A remote sensing data processing model is proposed.Based on the storage of XML files,the model processes the remote sensing data by adapting the algorithm to form an XML file interface driver,which has good scalability.Through parallel analysis of the model,it is considered that the model has certain advantages in improving processing efficiency and optimizing parallel strategies.(3)Research on GF satellite data preprocessing parallel strategy.Aiming at the geometry correction and quick-view flow in the GF satellite data preprocessing process,combined with the principle of GF satellite image data and algorithm,the data partitioning of the parallelization strategy is analyzed in depth,and the remote sensing image is rationally divided and divided into bands,parallelism improves the efficiency of remote sensing data preprocessing.(4)The above model is applied to the design of remote sensing preprocessing system.Taking GF satellite data preprocessing as an example,combined with related parallel strategies,experimental analysis shows that the proposed remote sensing data processing model can improve the efficiency of remote sensing image preprocessing to some extent.