Research On Key Technology Of The Big Data Management System For Remote Sensing Image And Its Implementation

With the rapid development of remote sensing technology, the variety of remote sensing image data has increased exponentially, and the growth of geometric series is presented at the same time. With the application of remote sensing image data and the increasing demand of business, more and more requirements are put forward for the management of remote sensing image, and the efficient storage and fast retrieval have become the urgent problem to be solved. And in Internet technology continuous innovation today, cloud storage as a new storage model for data provide unlimited storage capacity, application the technology to manage and store the massive remote sensing image, is also a hot spot of current research. The rapid retrieval of spatial data can’t be separated from the support of efficient indexing technology. It is of great value and significance to study the efficient indexing mechanism for the retrieval of spatial data. In addition to the requirements of retrieval performance, another technical index of remote sensing image storage management is to retrieve accuracy, such as from massive remote sensing image to pinpoint the image data which has a certain ground object feature. Thus, remote sensing image retrieval based on semantic features is also the research point in this paper.Based on the above problems, this thesis makes a series of research on remote sensing image data management and retrieval. Therefore, the main research work and achievements include:(1) Design different analytical methods to extract the required remote sensing image metadata, the metadata is composed of the data table of the remote sensing image data management system, based on these parameters of remote sensing images to complete a variety of ways to query and retrieval. In addition, remote sensing data as a spatial data, the merits of its retrieve performance is directly affected by the indexing mechanism. Therefore, this paper also designed experiments to verify the performance of different spatial indexing mechanism. The performance of the index tree is tested by using two dimensional spatial data, the experiment is based on different data level and different index terms, performance testing and verification of different indexing techniques are implemented by Java program in a visual way.(2) Design of semantic feature retrieval for remote sensing image, based on the proposed scheme, completing extraction of ground object features including in remote sensing images, these features are mapped to a set of semantic ID strings, which are stored in the database as the metadata of the remote sensing image management system. First of all, based on Watershed Transform and Full Lambda-Schedule Merge block algorithm, to segment the original remote sensing image; through the SVM classification algorithm, the feature extraction and classification of remote sensing images are completed. Finally, the classified features are mapped to semantic ID.(3) Development of remote sensing image data management system. According to the requirement analysis of the system, the architecture of the system is designed, according to this framework, a data management system for remote sensing image based on Hadoop is developed. The paper focuses on the interactive process of the Hadoop system. Based on the HDFS, remote sensing image data to efficiently store and manage. Remote sensing image data management system is a B/S architecture model system, through the design of each function module, eventually complete the image of the query and retrieval, fast browsing, as well as local downloads.