Pixel Value Change Detection for Lesion Regions in Endoscopic Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image processing techniques for detecting lesions in body cavity images struggle with accurately identifying candidate lesion regions due to environmental differences and false positives, often relying on color information that is prone to changes in imaging environments.
Innovation Solution
An image processing apparatus and method that calculates pixel value change amounts with surrounding pixels and inter-surrounding pixel change amounts to detect candidate lesion regions, using weighted change amounts and feature extraction to determine if a region is a real lesion, while excluding non-target pixels and regions caused by organ wall grooves or food residues.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If color information comparison techniques are used to detect lesions, then lesion detection can be performed, but detection accuracy deteriorates due to environmental differences and false positives
Solution Approach 1:
The patent changes the detection parameter from color information (which varies with imaging environment) to pixel value change amounts relative to surrounding pixels. This parameter transformation makes the detection metric invariant to environmental conditions such as lighting changes, camera variations, and imaging conditions, thereby resolving the contradiction between reliability and measurement precision.
Solution Approach 2:
The patent creates a local reference model by copying the pixel values from surrounding pixels to establish an expected pattern. By comparing the center pixel against this locally copied reference rather than against fixed color thresholds, the system adapts to local environmental conditions and reduces false positives caused by global environmental variations.
2Productivity
If conventional lesion detection methods are used, then processing can be performed, but detection speed deteriorates due to complex processing requirements
Solution Approach 1:
The patent segments the image processing task into simple, independent pixel-level operations. Instead of performing complex regional analysis or color space transformations on entire images, the method divides the problem into individual pixel comparisons with their immediate neighbors, enabling parallel processing and significantly improving detection speed while reducing computational complexity.
Solution Approach 2:
The patent applies partial action by focusing computational effort only on pixels that show significant value changes relative to their surroundings, rather than processing every pixel with equal complexity. This selective approach identifies candidate lesion regions quickly without requiring exhaustive analysis of the entire image, thereby improving processing efficiency.
Data Source
AI summary
An image processing apparatus for performing an image processing on a body cavity image captured in a living body includes: a storage unit which stores information including image information of the body cavity image; a change amount calculator which reads out the image information of the body cavity image from the storage unit and calculates, in the read body cavity image, a pixel value change amount of a pixel of interest with a plurality of surrounding pixels located around the pixel of interest; and a candidate lesion region detector which detects a candidate lesion region in the body cavity image based on a calculation result of the change amount calculator.