Aerial Image Illumination Correction via Patch Similarity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current digital image processing systems face issues with uneven ambient illumination, particularly in aerial images, leading to undesirable shadows and non-uniform color intensity, resulting in images of less than desirable quality.
Innovation Solution
A computer-implemented method that identifies patches of pixels in images, calculates similarities based on pixel intensities, and estimates illumination values to remove illumination variation, using an image processing server with modules for initialization, similarity calculation, optimization, and image enhancement to correct pixel intensities and eliminate shadows.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If aerial images are captured for map systems, then geographic coverage is improved, but illumination uniformity deteriorates due to shadows from clouds or objects
Solution Approach 1:
The patent extracts the illumination component from the image by comparing pixel patches across different regions. By identifying similar patches and analyzing their intensity differences, the method separates the illumination variation from the actual image content, allowing removal of shadows and non-uniform lighting effects while preserving the geographic information.
Solution Approach 2:
The patent applies local quality analysis by examining small patches of pixels throughout the image. Each patch is compared with similar patches to determine local illumination characteristics. This localized approach allows the method to handle varying illumination conditions in different regions of the aerial image while maintaining overall geographic coverage.
2Ease of manufacture
If images with uneven illumination are used, then data acquisition is simplified, but image quality deteriorates due to non-uniform color intensity
Solution Approach 1:
The patent implements a feedback mechanism where the system continuously compares pixel patches, calculates illumination values based on patch similarities, and uses this information to correct illumination variations. The illumination map generated from patch comparisons provides feedback that guides the correction process, iteratively improving image quality while maintaining the simplicity of the original data acquisition method.
3Measurement precision
If illumination values are calculated for all pixels, then illumination removal accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent divides the image into multiple small patches and processes them independently. By segmenting the image into manageable units and comparing only similar patches, the method reduces the computational burden of calculating illumination values while maintaining accuracy. This segmentation approach allows parallel processing and reduces the number of comparisons needed compared to analyzing all pixels uniformly.
Solution Approach 2:
The patent applies partial action by calculating illumination values only for representative patches rather than every single pixel. By selecting key patches and interpolating or propagating their illumination characteristics to surrounding areas, the method achieves sufficient illumination removal accuracy without the excessive computational cost of processing every pixel individually.
Data Source
AI summary
An image comprising varying illumination is selected. Patches of pixels from among the plurality of pixels with the image are identified. Similarities between pairs of patches of pixels based on pixel intensities associated with the pairs of patches of pixels are calculated. Illumination values for the plurality of pixels within the image based on the calculated similarities between the pairs of patches of pixels is calculated. The illumination variation from the image is removed based on the calculated illumination values for the plurality of pixels within the image.