Multi-Spectral Imagery Atmospheric Correction via Iterative Radiance Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing satellite imaging tools using multi-spectral imagery struggle to accurately model atmospheric effects like haze and aerosol loading, leading to improper adjustments in imagery due to the limitations of generic atmospheric models that fail to capture these effects.
Innovation Solution
A method and system that identify low intensity regions in multi-spectral images to estimate upwelling path radiance, iteratively adjust visibility values using a radiation transfer model, and repeat calculations until a termination condition is met to accurately model atmospheric effects and adjust imagery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If generic atmospheric models are used for multi-spectral imagery, then the modeling process is simple and fast, but the accuracy of atmospheric effects modeling is insufficient
Solution Approach 1:
The patent replaces manual or simple generic atmospheric modeling with an automated iterative system that uses radiation transfer models and image data analysis to calculate atmospheric parameters, substituting complex computational processes for simplified generic models
Solution Approach 2:
The patent implements an iterative feedback mechanism where the system calculates atmospheric parameters, applies corrections to the imagery, compares the results with reference values, and adjusts the parameters accordingly until convergence is achieved, thereby improving modeling accuracy through continuous refinement
2Measurement precision
If hyperspectral imaging is used for atmospheric characterization, then atmospheric effects can be accurately modeled, but the data requirements and system complexity increase significantly
Solution Approach 1:
The patent extracts and utilizes specific features from multi-spectral imagery (such as dark regions and spectral signatures) to derive atmospheric parameters, obtaining sufficient atmospheric characterization information without requiring the complete hyperspectral data set
Solution Approach 2:
The patent makes multi-spectral imagery serve multiple functions: it is used both for surface observation and for atmospheric characterization, eliminating the need for separate hyperspectral sensors by maximizing the utility of existing multi-spectral data
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach allows for precise adjustment of multi-spectral imagery by determining the necessary parameters for atmospheric models, effectively accounting for haze and other atmospheric effects, resulting in improved image quality and accuracy.
Implementation Method 1
calculating second values for upwelling path radiance for each of the different images, each of the second values based on a radiation transfer model
Data Source
AI summary
A method for processing remotely acquired imagery data includes identifying a region of lowest intensity in each of a plurality of different images for a plurality of different spectral bands and estimating first values for upwelling path radiance for each of the different images according to an intensity in the region in each of the different images. The method further includes selecting a visibility value and calculating second values for the upwelling path radiance for each of the different images, each of the second values based on a radiation transfer model, the radiation model generated using the visibility value and the meta-data. The method can also include comparing the first and the second values for the upwelling path radiance for the different images, adjusting the visibility value based on the comparison, and repeating the calculating, the comparing, and the adjusting steps until a termination condition is met.