Real-time Shadow Correction in Hyperspectral Retinal Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for detecting and correcting shadowing in hyperspectral retinal images captured by hyperspectral sensors are slow and non-real-time, leading to time-consuming reacquisition of images in clinical settings due to poor image quality caused by uneven illumination or insufficient pupil dilation.
Innovation Solution
A processor-based method for real-time detection and correction of shadowing in hyperspectral retinal images, which involves calculating a shift vector to determine shadow severity, generating alerts for reacquisition, and using soft masks and tensors to improve image quality, enabling immediate correction and reacquisition of images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional methods are used for shadow detection in hyperspectral retinal images, then shadowing can be detected and corrected, but the process is slow and does not provide real-time results
Solution Approach 1:
The patent transforms the shadow detection problem from analyzing intensity variations to calculating geometric shift vectors. By changing the detection parameter from intensity-based metrics to displacement-based measurements, the system achieves both real-time performance and accurate shadow identification through straightforward geometric computations.
Solution Approach 2:
The patent replaces complex iterative shadow correction algorithms with a geometric shift vector calculation approach. This substitution eliminates time-consuming computational steps while maintaining detection accuracy, enabling real-time shadow identification in hyperspectral retinal images.
2Loss of time
If shadowing is not detected in real-time, then image processing can be simpler, but time-consuming reacquisition is required in clinical settings
Solution Approach 1:
The patent performs shadow detection immediately upon image capture rather than delaying processing. By implementing preliminary shadow analysis using the shift vector method, the system identifies problematic images in real-time, allowing for immediate reacquisition decisions without time-consuming post-processing delays.
Solution Approach 2:
The system automatically detects shadows and triggers reacquisition decisions without requiring manual review. The shift vector calculation autonomously identifies shadowed regions, and the system self-manages the reacquisition process, eliminating delays associated with human intervention while maintaining simple operational workflows.
3Manufacturing precision
If conventional shadow correction methods are applied, then image quality can be improved, but the correction process is difficult and time-consuming
Solution Approach 1:
The patent extracts shadow information as a separate geometric shift vector distinct from the image intensity data. By isolating shadow detection into an independent geometric calculation, the system simplifies the overall processing workflow while maintaining the ability to identify and address shadowing issues effectively.
Data Source
AI summary
A method for real-time detection and correction of shadowing in hyperspectral retinal images may include capturing receiving, using a processor, a hyperspectral image of a retina of a patient, detecting, by the processor, a shadow in the hyperspectral image, determining, by the processor that the shadow of the hyperspectral image exceeds a threshold, and in response to determining that the shadow of the hyperspectral image exceeds the threshold, initiating, using the processor, a capture of an additional hyperspectral image of the retina of the patient. Various other methods, systems, and computer-readable media are also disclosed.