Digest Image Detection Using Optical Flow Region Connection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image processing techniques for detecting digest images from time-series images, such as those captured by a capsule endoscope, lack high coverage characteristics, leading to incomplete information recognition during medical diagnosis.
Innovation Solution
An image processing apparatus and method that connects regions depicting the same target across time-series images using optical flow calculation, calculates feature data based on connected region volumes, and determines a digest index value to detect images with high coverage characteristics by aggregating targets, enabling the detection of digest images with improved recognition capabilities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional image processing techniques are used to detect digest images from time-series images, then processing speed is improved, but coverage characteristic deteriorates leading to incomplete information recognition
Solution Approach 1:
The patent divides the image processing task into multiple stages: initial digest image detection using conventional techniques, followed by secondary processing to identify additional important images that were missed. This segmentation allows the system to balance processing speed with comprehensive information coverage by applying different processing strategies to different subsets of images.
Solution Approach 2:
The patent applies partial processing to the entire image set by first processing a subset to generate digest images, then performing additional processing on remaining images to ensure no important information is missed. This approach achieves adequate coverage without processing every image in exhaustive detail, balancing speed and completeness.
2Loss of time
If a predetermined number of images are extracted based on change amount, then processing time is reduced, but diagnostic information completeness deteriorates
Solution Approach 1:
The patent implements a feedback mechanism where the results of initial digest image detection are used to guide subsequent processing. Images that were not selected as digest images but show significant changes or match specific diagnostic criteria are flagged for secondary processing, ensuring that important diagnostic information is not lost while maintaining efficient processing of the majority of images.
Solution Approach 2:
The patent performs preliminary processing to identify and extract digest images based on obvious change metrics, then uses this preliminary result to guide further processing. This preliminary action quickly reduces the image set to manageable size while a secondary pass ensures diagnostic completeness by checking for missed important images.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The solution effectively detects digest images with high coverage characteristics, ensuring comprehensive information recognition and aggregation of targets, thereby enhancing diagnostic efficiency.
Implementation Method 1
connects regions depicting the same target among a series of images captured in time series with each other; the calculator 100 calculates feature data of each connected region
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
An image processing apparatus includes a corresponding region connecting unit (110) that connects regions depicting the same target between a series of images captured in time series, thereby sets at least one connected region; a connected region feature data calculation unit (120, 210) that calculates feature data of the connected region; a digest index value calculation unit (130, 320, 330) that calculates a digest index value corresponding to a degree at which the target depicted in the series of images in each image of the series of images, based on the feature data; and a digest image detector (140) that detects a digest image based on the digest index value.