Stratospheric Image Processing for Cloud Removal and Geolocation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for processing stratospheric images are ill-suited due to altitude, atmospheric conditions, and imagery capture mechanics, leading to inadequate color processing and geolocation challenges.
Innovation Solution
An apparatus and method utilizing a processor with a memory and specially-trained machine learning models to remove obfuscatory data, geolocate images, and stitch them into a mosaic, including color correction and blur detection, suitable for stratospheric altitudes between 30,000 to 100,000 feet.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional aerial imaging techniques are used for stratospheric images, then existing processing methods can be applied, but the methods are ill-suited due to altitude, atmospheric conditions, and imagery capture mechanics
Solution Approach 1:
The patent applies parameter changes by adjusting color correction parameters and processing algorithms specifically for stratospheric altitudes (30,000 to 100,000 feet). The system modifies conventional processing parameters to account for unique atmospheric conditions at high altitudes, including specialized color correction for stratospheric imaging and adjustments for atmospheric scattering effects, thereby improving both adaptability and processing accuracy.
Solution Approach 2:
The patent introduces an intermediary processing layer that includes a blur detection model and cloud detection model as mediators between raw image capture and final map-ready output. These intermediary models specifically detect and filter obfuscatory data caused by atmospheric conditions, serving as a buffer that protects the final output from degradation while maintaining processing accuracy.
2Measurement precision
If full atmospheric calibration with Top of Atmosphere reflectance measurements is applied, then satellite image processing requirements are met, but the approach is not suited for stratospheric sensors
Solution Approach 1:
The patent extracts and removes the complex Top of Atmosphere reflectance measurement requirements from the processing pipeline. Instead of requiring full atmospheric calibration, the system extracts only the essential color correction needs and applies simplified, stratosphere-specific color correction algorithms, thereby reducing device complexity while maintaining measurement precision for the specific application.
Solution Approach 2:
The patent changes the calibration approach by using stratosphere-specific color correction parameters rather than conventional satellite calibration methods. This parameter change simplifies the calibration system by focusing on the specific atmospheric conditions at stratospheric altitudes, reducing the complexity of required calibration equipment and procedures while maintaining adequate measurement precision.
3Productivity
If conventional color processing is used for aerial imaging at 1,500 to 5,000 feet, then minimal color processing is required, but significant color deviations occur at stratospheric altitudes
Solution Approach 1:
The patent applies parameter changes by implementing stratosphere-specific color correction algorithms that account for significant color deviations at high altitudes. The system adjusts color processing parameters to compensate for atmospheric scattering and lighting conditions unique to stratospheric imaging, thereby improving color accuracy without excessively reducing processing speed.
Solution Approach 2:
The patent applies preliminary action by performing color correction and blur detection in advance during the image processing pipeline. The blur detection model and cloud detection model perform preliminary filtering to identify and remove obfuscatory data before final map generation, ensuring color accuracy is maintained while keeping the overall processing efficient.
4Quantity of substance
If image data with clouds and blur is included in processing, then complete image coverage is achieved, but the resulting images contain obfuscatory data that reduces quality
Solution Approach 1:
The patent extracts and removes obfuscatory data from the image processing pipeline using specialized detection models. The blur detection model and cloud detection model identify and extract only the useful portions of images, separating clear data from blurred or cloud-obscured regions, thereby maintaining image quality while preserving complete image coverage where possible.
Solution Approach 2:
The patent applies local quality by treating different regions of images differently based on their quality assessment. The detection models evaluate local areas of images to identify blurred or cloud-covered regions, and the system applies appropriate processing or filtering to each local region, maintaining high quality where data is clear while managing quality degradation in obscured areas.
Data Source
AI summary
In an embodiment, an apparatus for processing of high-altitude images is presented. An apparatus for processing high-altitude images may include a processor and a memory communicatively coupled to the processor. A memory may contain instructions configuring a processor to obtain image data. A processor may be configured to perform image processing on image data specific to an atmosphere associated with the image data. A processor may be configured to remove, through application of a machine learning model, obscure image data from processed image data to generate viable images. A processor may be configured to geolocate viable images based on obtained georeference data to generate geolocated image data. A processor may be configured to mosaic geolocated image data to produce a georeferenced map.