Research On Optimization Method Of Remote Sensing Image Data Storage Based On Cat Swarm Algorithm

With the advent of the era of big data,remote sensing image data has gradually demonstrated the characteristics of massive,multi-source,multi-scale,and multi-time equalization,which brings new challenges and considerations to the management and use of image data.Under the premise of avoiding data loss,how to organize and manage these remote sensing image data reasonably and efficiently,quickly and accurately find remote sensing image data that meet the different needs of users from a large number of remote sensing data sources has become an urgent problem to be solved today.Aiming at the challenge and thinking of remote sensing image data,in order to solve the problem of inefficient storage and query,this paper proposes a method of classification before storage,and designs an optimization method of remote sensing image storage using cat swarm algorithm,which effectively improves the storage performance and query speed of data.The main work of this paper is as follows:(1)Analyzing the storage and application requirements of existing image data,aiming at the problem of large data loss and low efficiency in image pyramid construction,the pyramid layering method and the segmentation strategy based on regular and irregular images were improved before pyramid construction.The cat swarm optimization algorithm is used to construct the pyramid.The method searches around four adjacent image blocks at the bottom of the pyramid to get the upper image of the pyramid.The idea of algorithm optimization is based on the similarity and number of image blocks.Experiments show that this method effectively improves the efficiency of pyramid construction and reduces the loss of image data.With the increase of the number of image blocks,the data integrity of pyramid becomes better and better.(2)In order to solve the problem of inefficiency and low query speed of traditional image data storage methods,this paper stores and querys image data through HBase database,redesigns the table structure of HBase.It combines the information of landmark code and quadtree index ID as RowKey,and the information of longitude and latitude coordinates of image data as column families.It effectively improves the query speed of image data.At the same time,the method of parallel inputting and querying image data is proposed,which reduces the time in the process of storage and query.In parallel storage,this paper introduces the pre-partitioning method to avoid the hot issues of HBase;in parallel query,a filter method is introduced to solve the problem of full table scanning for querying data other than RowKeys.Experiments show that the improved RowKey not only achieves the purpose of data filtering,but also improves the reading efficiency.It has good feasibility and scalability.In this paper,cat swarm algorithm is used to construct pyramid for classified image data by parallel computing technology,and the combination of landmark code and quadtree index coding is used to improve the database RowKey method to solve the problem of low speed of image data storage and query,and improve the efficiency of image management and use.