Adaptive Histogram Co-Registration for Satellite Imagery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Accurate co-registration of images from different spectral bands in satellite imagery is hindered by physical offsets and optical instability, leading to warping and distortion, which impede object classification and material identification efforts.
Innovation Solution
A computer-implemented method that retrieves images from different sensor bands, creates signal intensity histograms, cross-correlates them to create a correlation map representing Mutual Information, and uses the peak to register the images, with bin widths adjusted based on noise levels and signal intensity, allowing for registration of images with different spatial resolutions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If images from different spectral bands are captured using physically offset sensors, then multi-spectral information is obtained, but image alignment accuracy deteriorates due to warping and distortion
Solution Approach 1:
The patent introduces an intermediary processing system that uses elevation models and coordinate transformation algorithms to mediate between the physically offset sensor views and produce aligned images. The system acts as a mediator that reconciles the geometric differences caused by sensor offsets through computational methods.
Solution Approach 2:
The patent transforms the registration problem from direct pixel-to-pixel alignment to a parameter-based approach using elevation models, viewing angles, and coordinate transformations. By changing the parameters used for registration (from simple geometric alignment to physics-based modeling), the system achieves accurate co-registration despite physical sensor offsets.
2Ease of manufacture
If traditional registration methods are used with fixed histogram bins, then processing is simple, but registration accuracy deteriorates in the presence of noise and varying signal intensities
Solution Approach 1:
The patent transitions from static, fixed histogram bins to dynamic, adaptive histogram bins that automatically adjust their width based on local signal intensity and noise characteristics. This dynamic adaptation allows the registration method to maintain high accuracy across varying image conditions without requiring manual tuning or complex preprocessing.
Solution Approach 2:
The patent implements a feedback mechanism where the histogram bin width is determined by the local signal-to-noise ratio. The system measures the noise level and signal intensity, then uses this information to adjust the bin width accordingly, creating a self-regulating system that optimizes registration accuracy based on actual image conditions.
3Adaptability or versatility
If images with different spatial resolutions are registered, then multi-resolution analysis is enabled, but registration sensitivity deteriorates
Solution Approach 1:
The patent applies different histogram bin widths to different regions of the image based on local signal intensity and noise characteristics. This local quality approach allows the system to maintain high registration sensitivity in all regions by adapting the bin width to the specific conditions of each local area, rather than using a uniform bin width across the entire image.
Data Source
AI summary
Techniques for improving image co-registration. One aspect relates to improving Mutual Information techniques by making the histogram bin widths used therein depend on the amount of signal noise. Another aspect relates to populating the bins by integrating the signal probability in each bin. A third aspect relates to converting top-of-the-atmosphere image data to surface reflectance data, and then using surface reflectance data in the Mutual Information technique for determining a correlation surface. The fourth aspect relates to registering higher-resolution images with lower-resolution images by down-sampling the higher-resolution image. The last aspect relates to a technique for determining the accuracy of the co-registration by synthesizing a perfectly-registered image in a second wavelength band from an image in a first wavelength band and then using Mutual Information between the two images to create a correlation surface.