Remote Sensing Image Storage in HBase Using Google S2 Grids
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional remote sensing image management systems face challenges with massive, heterogeneous, and distributed data storage due to lack of unified standards, leading to inefficient data management, increased development costs, and wastage of bandwidth resources, as well as inability to share or recover data effectively across different satellite systems.
Innovation Solution
A method for storing remote sensing big data in an HBase database using the Google S2 algorithm, which involves pre-cropping images based on grid levels, serializing image blocks, and establishing a table storage model with a unique RowKey to facilitate efficient storage and management of multi-source data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If traditional file-based management manner is used for remote sensing images, then data storage is simple, but data management and distribution efficiency deteriorates with massive data
Solution Approach 1:
The patent segments remote sensing images into multiple image blocks based on Google S2 grid levels. Each image block corresponds to a specific grid cell and can be independently stored and managed in the HBase database. This segmentation enables efficient retrieval and distribution of specific regions without managing entire large-scale images, directly improving data management efficiency while maintaining storage simplicity through structured organization.
Solution Approach 2:
The patent introduces a new dimensional organization by mapping image blocks to HBase RowKeys using a multi-dimensional encoding scheme that includes S2 grid level, latitude, longitude, and image identification. This dimensional transformation allows the system to efficiently manage massive remote sensing data in the HBase distributed database, improving both management efficiency and distribution capability while maintaining storage simplicity.
2Adaptability or versatility
If no unified standard is established for storing various types of remote sensing data, then data format diversity is preserved, but development costs and construction periods increase
Solution Approach 1:
The patent creates a universal storage model in HBase that can accommodate multiple types of remote sensing data (optical images, SAR images, hyperspectral data) with different formats and resolutions. The unified RowKey design and flexible column family structure allow the system to store and manage diverse data types through a single standardized interface, reducing system complexity while preserving adaptability to various data formats.
Solution Approach 2:
The patent uses parameter-based organization where image blocks are categorized and stored according to multiple parameters including S2 grid level, spatial coordinates, resolution, and satellite identification. This parameter-driven approach allows the system to handle diverse remote sensing data formats by organizing them according to standardized parameters, reducing system complexity while maintaining versatility.
3Ease of manufacture
If metadata plus image files management manner is used, then data storage is straightforward, but data management accuracy and sharing capability deteriorates
Solution Approach 1:
The patent merges image data and metadata into a unified HBase storage structure where both are stored together with consistent RowKeys. Image blocks and their associated metadata (acquisition parameters, processing information, spatial references) are co-located in the same database records, improving data management accuracy by ensuring data integrity and enabling efficient joint retrieval without the inconsistencies of separate metadata and file storage.
4Loss of energy
If remote sensing images are stored without pre-processing and cropping, then storage bandwidth is saved, but data transmission efficiency and utilization deteriorates
Solution Approach 1:
The patent performs preliminary cropping of remote sensing images into standardized image blocks based on Google S2 grids during the storage phase. This pre-processing creates uniformly sized, georeferenced image blocks that can be directly retrieved and used without additional processing. The pre-cropped image blocks improve data transmission efficiency and utilization by enabling direct access to specific regions of interest without transmitting entire large-scale images.
Data Source
AI summary
A method of storing remote sensing big data in a Hbase database. The method includes: obtaining a remote sensing image, and pre-cropping the remote sensing image based on a Google S2 algorithm; selecting a grid level based on a resolution of the remote sensing image; demarcating a coverage scope of the remote sensing image; calculating grids covered within the coverage scope; cropping the remote sensing image according to the grids, to yield a plurality of image blocks, and serializing the plurality of image blocks; establishing an Hbase table; establishing a table storage model containing RowKey, and calculating the RowKey according to attributes of an image block; and integrating data structure of multi-source remote sensing image data according to attribute expression of a zoned remote sensing image in a plurality of dimensions, and storing serialized image blocks in an Hbase database according to the RowKey.