Illumination-Corrected Image Generation via Coefficient Component Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing remote sensing techniques fail to accurately estimate and correct the environmental fluctuation component from observation images, particularly the coefficient component, which is essential for obtaining accurate information about the earth's surface, as they rely on pre-stored atmospheric transmittance values and do not account for illumination variations.
Innovation Solution
An image processing device and method that reads observation images across multiple wavelength bands, eliminates reflection absorption band areas, calculates the coefficient component proportional to illumination and atmospheric transmittance, and generates an illumination-corrected image by dividing pixel luminance values by the calculated coefficient component.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If pre-stored atmospheric transmittance values are used for correction, then the correction process is simple, but the measurement precision deteriorates because illumination variations cannot be accounted for
Solution Approach 1:
The system uses the observation image data itself to calculate the coefficient component through statistical processing (standard deviation calculation), making the correction process self-sufficient without requiring external pre-stored atmospheric transmittance values. This eliminates the trade-off by making the system both simple to operate and highly accurate.
2Measurement precision
If coefficient component calculation is performed for each wavelength band, then the measurement precision improves, but the device complexity increases
Solution Approach 1:
The system segments the correction process by wavelength band, calculating the coefficient component separately for each wavelength band used in observation. This segmentation allows accurate wavelength-specific correction while maintaining clear organizational structure, managing the complexity through systematic breakdown rather than overwhelming simultaneous processing.
3Measurement precision
If reflection absorption band areas are eliminated before calculation, then the measurement precision improves, but the loss of information increases
Solution Approach 1:
The system extracts and eliminates only the problematic reflection absorption band areas from the observation image before performing coefficient component calculation. This selective extraction removes the source of calculation instability while preserving all other valid observation data, achieving stable estimation without excessive information loss.
Data Source
AI summary
Provided is an image processing device including: an image read-in unit 51 which reads in one or more observation images which retain observation results from one or a plurality of wavelength regions; a reflection-absorption band region deletion unit 52 which deletes, from each of one or more of the observation images, observation results with respect to a reflection-absorption band region, and generates reflection-absorption band-deleted images; a coefficient constituent proportional value computation unit 53 which, using the reflection-absorption band-deleted images, derives a proportional value of a coefficient constituent for each wavelength band used in the observation; and a coefficient constituent deletion unit 54 which, on the basis of the obtained proportional values of the coefficient constituents for each of the wavelength bands, deletes the coefficient constituents from each of the observation values included in the one or more observation images, and generates illumination-corrected images.