Composite LWIR Image Processing for Gas Turbine Diagnostics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The precision of thermal imaging for gas turbine engine diagnostics is limited by the movement and vibration of engine parts, as well as variations in camera fields of view, leading to challenges in accurately comparing thermal images against expected heat profiles.
Innovation Solution
A method that identifies geometric features on the moving workpiece, designates a master feature, and adjusts secondary images for offset, rotation, and scaling to align with actual coordinates, allowing for the creation of a composite thermal image that minimizes differences between pixel and actual coordinates, thereby enhancing diagnostic accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If thermal imaging is performed on moving workpieces with vibration, then diagnostic capability is enabled, but measurement precision deteriorates due to image misalignment
Solution Approach 1:
The system performs preliminary actions by capturing multiple thermal images at different positions before processing, and pre-identifies geometric features in advance. This allows the image registration process to have reference points ready before alignment occurs, improving the precision of thermal image comparison despite workpiece movement and vibration.
Solution Approach 2:
The system creates a composite image that is a synthesized copy combining multiple individual thermal images. This composite image serves as a stable reference that replicates the thermal patterns while eliminating the effects of workpiece movement, enabling accurate diagnostic comparison without being affected by vibration-induced misalignment in individual frames.
2Measurement precision
If multiple thermal images are captured to improve diagnostic accuracy, then measurement precision improves, but device complexity increases due to image processing requirements
Solution Approach 1:
The system extracts only the essential geometric features from each thermal image, separating these stable reference points from the rest of the image data. By focusing processing only on identifying and aligning these extracted features rather than processing entire images, the system achieves accurate registration while reducing computational complexity.
Solution Approach 2:
The system introduces geometric features as intermediary reference elements that mediate between multiple thermal images. These features serve as intermediate markers that simplify the alignment process by providing clear, easily identifiable correspondence points across different images, reducing the complexity of direct image-to-image comparison.
3Manufacturing precision
If geometric feature identification is used to align images, then manufacturing precision of image registration improves, but difficulty of detecting and measuring increases due to feature identification requirements
Solution Approach 1:
The system employs geometric features that serve multiple functions: they act as alignment references for image registration, provide diagnostic information about workpiece geometry, and enable tracking of workpiece position and orientation. This multi-functionality reduces the need for separate detection systems, making feature identification more straightforward despite the increased precision requirements.
Data Source
AI summary
A method for thermally imaging a moving workpiece of a gas turbine engine comprises identifying a plurality of geometric features to construct a composite image. The geometric features include at least one integral thermal feature of the moving workpiece, and at least one artificial feature applied to the workpiece for diagnostic purposes. One of the plurality of geometric features is identified as a master feature, and the remainder of the plurality of geometric features are located relative to the master feature with relative actual coordinates. A pixel location of the master feature is identified or each image, and the remainder of the plurality of geometric features are located relative to the master feature with relative pixel coordinates. Offset, rotation, and scaling of the secondary images are varied to minimize a relative difference between the relative pixel coordinates and the relative actual coordinates. The offset, rotated, and scaled secondary images are combined with the reference image to form a composite image.