Synthetic EMR Profile Generation for Surface Anomaly Detection
Solution Overview
Problem
Conventional methods for detecting surface anomalies on aircraft parts, such as visual inspections and thermal imagery, are unreliable and impractical due to lighting and ambient condition variations, requiring numerous observations that are often infeasible for in-service aircraft.
Innovation Solution
A computer-implemented method using machine learning algorithms to generate synthetic electromagnetic radiation (EMR) profiles and images, which account for varying ambient conditions, allowing for fewer measurements to detect surface anomalies by converting between different EMR ranges and types, such as optical to thermal images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If thermal imagery is used to detect surface anomalies, then detection capability is improved, but the number of required images increases making the inspection infeasible
Solution Approach 1:
The patent uses generative adversarial networks (GANs) to create synthetic thermal images from optical images. The GAN generates realistic thermal profiles that mimic actual thermal imagery, allowing the system to perform anomaly detection without requiring numerous actual thermal images. This synthetic copying approach reduces the number of required observations while maintaining detection reliability.
Solution Approach 2:
The system performs preliminary training of the GAN model using paired optical and thermal images of normal (anomaly-free) surfaces. This preliminary action creates a synthetic database of thermal profiles under various ambient conditions, which can then be used for anomaly detection without requiring extensive actual thermal image collection during inspection.
2Reliability
If thermal imagery-based inspection is performed, then surface anomalies may be detected, but the inspection becomes impractical for in-service aircraft due to the large batch of images needed
Solution Approach 1:
The GAN-based synthetic image generation creates realistic thermal profiles from readily available optical images, eliminating the need to capture large batches of actual thermal images. This makes the inspection process practical for in-service aircraft where extensive image collection would be time-consuming and operationally disruptive.
Solution Approach 2:
The system changes the approach from collecting multiple actual thermal images to generating synthetic thermal images through computational modeling. By transforming the problem from physical measurement to computational synthesis, the inspection becomes feasible with far fewer observations while maintaining detection reliability.
3Reliability
If thermal inspection methods are used, then anomaly detection is possible, but they fail to account for ambient condition variations affecting thermal profiles
Solution Approach 1:
The GAN model is trained to account for dynamic ambient conditions including temperature, humidity, and lighting variations. The synthetic thermal images generated by the model reflect how thermal profiles change under different ambient conditions, allowing the anomaly detection system to distinguish between thermal variations caused by environment versus those caused by actual surface anomalies.
Solution Approach 2:
The system explicitly models and accounts for ambient condition parameters (temperature, humidity, lighting) in the synthetic image generation process. By incorporating these environmental parameters into the GAN training and generation, the system adapts to varying conditions while maintaining consistent anomaly detection performance.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enhances the accuracy and reliability of surface anomaly detection by reducing the number of required observations and accounting for changes in temperature, humidity, and lighting, providing a more efficient and effective inspection process.
Implementation Method 1
generating synthetic electromagnetic radiation (EMR) profiles for the object based on the measured EMR profiles
Data Source
AI summary
Systems and methods for detecting surface anomalies are disclosed. For example, a computer-implemented method for detecting surface anomalies on an object comprises receiving measured electromagnetic radiation (EMR) profiles for the object, generating synthetic EMR profiles for the object based on the measured EMR profiles, determining whether the object contains a surface anomaly based on the measured EMR profiles and the synthetic EMR profiles, and indicating a surface anomaly to a user via a display when a surface anomaly is detected. In another example, a system comprises a computing device comprising non-transitory memory with computer-readable instructions for receiving unpaired image data of an object of two different image types, predicting missing image data to generate paired image data of the two different image types, and determining whether the object contains a surface anomaly based on the paired image data. The computing device comprises a processor configured to execute the computer-readable instructions.



