Endoscope Image Processing for One-Eye Blur Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Three-dimensional endoscopes with stereo cameras face issues due to 'one-eye blurring' caused by dirt on one optical system, leading to discomfort and fatigue for users as they struggle to recognize high-frequency image deterioration.
Innovation Solution
An image processing apparatus that compares left and right images to detect blurred regions, sets corresponding regions, and determines if notification is needed based on overlap with regions of interest, thereby alerting users only when discomfort is likely.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If frequency components in entire left and right images are computed to detect one-eye blurring, then detection capability is improved, but user fatigue increases due to unnecessary warnings in non-critical regions
Solution Approach 1:
The patent applies local quality by dividing the image into multiple regions and computing frequency components only for specific regions of interest rather than the entire image. This allows detection to be focused on areas where blurring would be most critical, reducing unnecessary detections in non-important areas and thereby reducing user fatigue from false warnings.
Solution Approach 2:
The patent segments the image into multiple regions and processes each region separately. By computing frequency components for divided regions rather than the entire image, the system can identify blurring in specific areas while ignoring regions where such blurring would not impact the user experience, thus reducing unnecessary warnings.
2Measurement precision
If notification is provided for any detected blurring, then detection accuracy is improved, but productivity decreases due to interruptions during treatment
Solution Approach 1:
The patent applies local quality by identifying regions of interest and only providing notifications when blurring is detected in those specific regions. This selective notification approach maintains detection accuracy for critical areas while avoiding interruptions caused by notifications in non-critical regions, thereby preserving treatment efficiency.
Solution Approach 2:
The patent uses partial action by computing frequency components for divided regions rather than the entire image, and by providing notifications only for blurring in regions of interest. This partial processing and selective notification reduces the overall number of interruptions while maintaining adequate detection accuracy for treatment-critical areas.
3Reliability
If full image processing is performed to detect one-eye blurring, then detection reliability is improved, but processing time increases
Solution Approach 1:
The patent segments the image into multiple regions and processes each region separately rather than processing the entire image at once. This segmentation reduces the computational burden per processing cycle while maintaining detection reliability through cumulative analysis of multiple regions, thereby reducing overall processing time.
Solution Approach 2:
The patent performs partial processing by computing frequency components for divided regions rather than the entire image in each frame. This partial processing approach reduces computation time while maintaining adequate detection reliability for the regions that matter most for treatment.
Data Source
AI summary
An image processing apparatus for an endoscope includes processing circuitry. The processing circuitry acquires a left image and right image. The processing circuitry generates a three-dimensional display image based on the left image and right image. The processing circuitry sets a first region in a part of the left image or right image. The first region is a region having a blur. The processing circuitry sets a second region in the left image or right image. The second region is a region where the user is performing treatment. The processing circuitry computes an overlapping region between the first region and the second region. The processing circuitry determines whether or not to notify the user based on the overlapping region. The processing circuitry generates information for notifying.