Endoscope Image Processing for Specular Reflection Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Endoscope image processing systems face challenges in effectively managing specular reflection, which can hinder the accurate detection and discrimination of objects, such as polyps, due to the interference it causes in image clarity.
Innovation Solution
An image processing apparatus and method that utilize a processor to differentiate between search and discrimination operations, maintaining or correcting specular reflection regions in endoscope images based on user operation states, employing convolutional neural networks for enhanced image recognition and classification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If specular reflection regions are corrected in endoscope images, then measurement precision of object discrimination is improved, but loss of information in search phase increases
Solution Approach 1:
The image processing system dynamically switches between two modes: in the search phase, it maintains original images with specular reflection to preserve information for object location; in the discrimination phase, it applies correction processing to remove specular reflection and improve measurement precision. This dynamic adaptation resolves the contradiction by adjusting the level of correction based on the current operational phase.
Solution Approach 2:
The observation process is segmented into distinct phases (search and discrimination), with different image processing strategies applied to each phase. During search, uncorrected images are used; during discrimination, corrected images are used. This segmentation allows the system to optimize for information preservation in one phase and measurement precision in another, resolving the contradiction.
2Loss of information
If specular reflection regions are maintained in images, then information for object search is preserved, but measurement precision for object discrimination deteriorates
Solution Approach 1:
The system dynamically adjusts image processing based on operational phase: maintaining original images during search to preserve information, and applying correction during discrimination to improve precision. This temporal dynamic resolves the contradiction by applying different strategies at different times.
Solution Approach 2:
The observation workflow is divided into search and discrimination segments, with uncorrected images used for search and corrected images used for discrimination. This segmentation allows optimal performance for each specific task without compromise.
3Measurement precision
If image correction processing is applied continuously, then object discrimination accuracy is improved, but device complexity increases
Solution Approach 1:
Rather than continuously applying correction processing, the system dynamically activates correction only during the discrimination phase when high precision is needed. This reduces overall processing complexity while maintaining high accuracy when required.
Solution Approach 2:
Image correction is segmented to apply only during the discrimination phase rather than continuously throughout observation. This reduces computational burden and system complexity during the search phase while ensuring high precision during discrimination.
4Productivity
If specular reflection correction is applied during search operation, then object location efficiency decreases, but image processing completeness improves
Solution Approach 1:
The system dynamically adjusts processing intensity based on operational phase: using minimal processing during search to maintain efficiency, and applying full correction during discrimination to ensure quality. This resolves the contradiction by optimizing for speed when searching and quality when discriminating.
Solution Approach 2:
Image processing is segmented into light processing during search and heavy correction processing during discrimination. This allows efficient object location followed by thorough analysis, resolving the contradiction between search efficiency and processing quality.
Data Source
AI summary
An image processing apparatus includes a processor. The processor is configured to: obtain a presumption result by presuming an operation state of a user who performs an operation while the user observes an image obtained by picking up an image of a subject; perform processing for correcting a specular reflection region included in the image obtained by picking up the image of the subject; and when detecting, based on the presumption result, that a first operation related to search for a desired object in the subject is performed, perform control for causing the image in which the specular reflection region is maintained to be outputted as an observation image, and when detecting that a second operation related to discrimination of the desired object found in the subject is performed, perform control for causing the image in which the specular reflection region is corrected to be outputted as an observation image.