Column-Level Index for Multi-Resolution Raster Data Access
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional database systems experience performance degradation and slow access speeds when performing multi-resolution access on large numbers of raster objects, as they only support fast access at the raster object level, leading to numerous small object accesses.
Innovation Solution
A method and apparatus for accessing raster data that determines a target pixel resolution and calculates a sampling level, using a column-level index generated by down-sampling raster data in columns, to improve access speed and efficiency, especially for large-scale queries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If conventional database systems access raster data at the raster object level, then each individual raster object can be accessed, but access performance degrades when performing multi-resolution access on large numbers of raster objects due to numerous small object accesses
Solution Approach 1:
The patent merges multiple raster objects into a single merged raster object that contains pixel data from multiple source raster objects. This consolidation allows the system to access multiple raster objects through a single unified structure, eliminating the need to individually access numerous separate raster objects and thereby improving access performance for multi-resolution queries.
Solution Approach 2:
The patent introduces a new dimension of organization by creating a column-level index structure that aggregates raster data across multiple raster objects. Instead of accessing raster objects in the traditional row-by-row manner, the system accesses data through a column-level index that spans multiple objects, fundamentally changing the access dimension from individual object level to aggregated column level.
2Reliability
If multiple raster objects are accessed to achieve multi-resolution access, then comprehensive data coverage is obtained, but the number of access operations increases leading to performance degradation
Solution Approach 1:
The patent performs preliminary actions by pre-merging multiple raster objects into a unified merged raster object and pre-building a column-level index structure. This preparation work is done in advance, so that when multi-resolution access is needed, the system can directly query the pre-organized structure without having to individually access and combine multiple raster objects at query time, significantly reducing access time while maintaining complete data coverage.
3Adaptability or versatility
If conventional raster data access methods are used, then simple queries can be executed, but complex multi-resolution queries on large datasets experience slow performance
Solution Approach 1:
The patent segments the access process into two distinct phases: first, determining the target pixel resolution and calculating the sampling level based on query parameters; second, using the pre-built column-level index to efficiently retrieve data at the calculated sampling level. This segmentation allows the system to handle complex multi-resolution queries by breaking them down into manageable steps, maintaining query versatility while improving execution speed through the optimized index structure.
Data Source
AI summary
A method, an apparatus, an electronic device, and a computer storage medium for accessing raster data are provided. The raster data access method comprises: executing a corresponding window query input to determine a target pixel resolution corresponding to a query window according to a raster data access request; calculating a sampling level according to the target pixel resolution and a stored pixel resolution of a raster object, and determining whether the sampling level is greater than a threshold; and accessing raster data corresponding to the query window according to a column-level index if the sampling level is greater than the threshold, the column-level index being an index generated by down-sampling according to raster data of the raster object in a column. Using the raster data access method can achieve quick access to raster data.