Thermal Image Skeleton Extraction for Noise Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Inexpensive small thermal image sensors suffer from poor image resolution, contrast, outline sharpness, and Signal-to-Noise (S/N) ratio, while large sensors are expensive, and existing sharpening methods amplify noise, reducing visibility.
Innovation Solution
An image processing device that generates a sharp thermal image with a high S/N ratio by using a background image generator to create a skeleton image from averaged and sharpened thermal frames, and an image corrector to correct thermal images using this skeleton image, thereby enhancing image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If inexpensive small thermal image sensors are used, then cost is reduced, but image resolution, contrast, outline sharpness, and S/N ratio deteriorate
Solution Approach 1:
The patent applies preliminary action by capturing multiple thermal images before processing and performing averaging processing in advance. This pre-processing step reduces noise and improves S/N ratio before the final image is generated, allowing inexpensive sensors to achieve image quality comparable to expensive sensors without requiring higher-end hardware
Solution Approach 2:
The patent segments the image processing into distinct steps: capturing multiple frames, calculating feature quantities, ranking images, determining middle images, averaging selected frames, and extracting skeleton components. This segmentation allows each processing stage to optimize specific aspects of image quality, resolving the contradiction between low sensor quality and high output image quality
2Manufacturing precision
If sharpening processing is applied to thermal images, then outline sharpness is improved, but noise is amplified and visibility is reduced
Solution Approach 1:
The patent performs averaging processing on multiple thermal images before applying sharpening. This preliminary averaging reduces noise in the input images, so that subsequent sharpening operations enhance outline sharpness without amplifying noise as much as sharpening would on single noisy frames
Solution Approach 2:
The patent introduces an intermediate processing step where skeleton components are extracted from the averaged image. These skeleton components serve as a mediator that captures essential outline information while filtering out noise, which are then used to correct the original thermal image, achieving sharpness without noise amplification
3Measurement precision
If multiple thermal images are processed to generate average image, then S/N ratio is improved, but processing time increases
Solution Approach 1:
The patent applies partial action by selecting only a subset of thermal images for averaging based on feature quantity thresholds. Instead of averaging all captured frames, it identifies and processes only those images with feature quantities within a specific range, reducing processing time while still achieving effective noise reduction and S/N ratio improvement
Solution Approach 2:
The patent performs preliminary ranking and selection of thermal images based on feature quantities before averaging. This pre-screening step identifies the most suitable frames for averaging, ensuring high S/N ratio improvement while minimizing the number of frames that need to be processed, thus reducing overall processing time
Data Source
AI summary
A background image generator (21) stores, in a storage device (3), a skeleton image obtained by calculating a feature quantity for each of multiple first thermal images (Din1) obtained by imaging by a thermal image sensor (1) in the same field of view or multiple sorted images (Dc) generated from the first thermal images, generating an average image from the first thermal images or sorted images, sharpening the average image, and then extracting a skeleton component. An image corrector (22) corrects, by using the skeleton image stored in the storage device (3), a second thermal image (Din2) obtained by imaging by the thermal image sensor in the same field of view as the first thermal images, thereby generating a corrected thermal image. It is possible to generate a sharp thermal image with a high S/N ratio.