Temporal Mapping Algebra for Satellite Imagery Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for analyzing temporal data, such as satellite imagery, typically operate on a per-pixel basis, discounting significant data and downsampling, which limits the effectiveness of data processing and analysis.
Innovation Solution
Temporal mapping and analysis (TMA) uses temporal map algebra to treat time series of raster data as three-dimensional data sets, considering spatial and temporal dimensions, allowing for improved data processing and analysis through the creation of temporal data cubes and custom composites using local, focal, and zonal functions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional per-pixel analysis methods are used, then processing simplicity is maintained, but data utilization efficiency deteriorates due to discounting significant data and downsampling
Solution Approach 1:
The patent extends conventional two-dimensional map algebra to three-dimensional temporal map algebra by adding the time dimension. This allows processing of temporal data cubes that incorporate multiple time steps, enabling analysis of temporal patterns and trends while utilizing all available data without downsampling. The third dimension (time) transforms the data structure from static 2D rasters to dynamic 3D temporal cubes, resolving the contradiction by maintaining data completeness while enabling sophisticated temporal analysis.
Solution Approach 2:
The patent segments temporal data into discrete time steps within the temporal cube structure, allowing independent processing of each time step while maintaining temporal relationships. This segmentation enables efficient parallel processing of different time steps or spatial regions, improving productivity without requiring complex sequential processing of entire datasets.
2Loss of information
If full temporal data is processed, then analysis completeness is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary actions by pre-processing raster data and organizing it into temporal data cubes before analysis. This includes optional preprocessing steps such as cloud masking, quality assessment, and data normalization that are performed once during data preparation. By preparing data in advance and structuring it into reusable temporal cubes, the system avoids redundant processing during analysis, reducing processing time while maintaining complete data utilization.
Solution Approach 2:
The patent enables dynamic adjustment of processing parameters such as temporal window size, composite types (min, max, mean, median), and spatial resolution. Users can optimize these parameters based on specific analysis requirements, balancing data completeness with processing time. For example, analyzing shorter temporal windows or using coarser spatial resolutions reduces processing time while still capturing essential temporal patterns.
3Measurement precision
If temporal data cubes are created with full resolution, then spatial detail is preserved, but storage requirements increase
Solution Approach 1:
The patent implements local quality by allowing different spatial resolutions in different regions of the temporal data cube. Users can specify higher resolution for areas of interest and lower resolution for less critical regions. This selective resolution approach preserves spatial detail where needed while reducing storage requirements in other areas, optimizing the balance between measurement precision and storage volume.
Solution Approach 2:
The patent enables partial processing by allowing users to analyze only specific time steps, spatial regions, or parameter combinations rather than processing the entire temporal data cube. This partial action approach reduces storage requirements by creating and maintaining temporal cubes at full resolution only for necessary analyses, while using lower-resolution or subset data for other purposes.
Data Source
AI summary
A compositing process for selecting spatial data collected over a period of time, creating temporal data cubes from the spatial data, and processing and/or analyzing the data using temporal mapping algebra functions. In some embodiments, the temporal data cube is creating a masked cube using the data cubes, and computing a composite from the masked cube by using temporal mapping algebra.