Capsule Camera Image Stitching Redundancy Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Image stitching of capsule camera images from the gastrointestinal tract is inefficient due to non-ideal image quality, deformation, and inconsistent camera movement, leading to high redundancy and prolonged review times for medical professionals.
Innovation Solution
A method involving high frame-rate image processing with deformable registration and motion estimation, where bridging images with lower spatial resolution assist in motion model derivation and stitching, reducing the number of stitching outputs by removing redundant images and conserving system resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If image stitching is performed on all high frame-rate images captured by the capsule camera, then the field of view is expanded and more GI tract surface is visualized, but the number of images to be reviewed increases significantly, prolonging review time
Solution Approach 1:
The patent extracts and removes redundant images from the sequence by identifying and eliminating images that do not contribute new information to the stitched output. This is achieved through analyzing motion models and determining which images are substantially redundant, thereby reducing the total number of images that need to be reviewed while maintaining complete coverage of the GI tract surface.
Solution Approach 2:
The patent discards redundant images that provide no additional value to the stitched output, while recovering and preserving only the essential images needed to represent the complete field of view. This selective discarding reduces review time while the stitching process recovers the complete anatomical information through image fusion.
2Measurement precision
If all high frame-rate images are processed through image registration and stitching, then registration quality is improved, but computational complexity and processing time increase
Solution Approach 1:
The patent extracts only the necessary images for processing by identifying redundant images beforehand. By removing redundant images from the processing pipeline, the computational complexity is reduced while maintaining registration quality on the essential images that contribute to the final stitched output.
Solution Approach 2:
The patent applies partial processing by performing image registration and stitching only on the non-redundant subset of images rather than all captured images. This partial action approach maintains registration quality where needed while avoiding excessive processing of redundant data.
3Measurement precision
If the capsule camera captures images at high frame rate to capture motion details, then image quality and motion resolution are improved, but the number of captured images increases, leading to higher data storage requirements
Solution Approach 1:
The patent extracts and removes redundant images from the high frame-rate sequence, keeping only the essential images that provide unique information. This extraction process reduces the data volume significantly while preserving the motion resolution and diagnostic information contained in the original high frame-rate capture.
Solution Approach 2:
The patent discards redundant high frame-rate images that duplicate information already captured in previous or subsequent images. By selectively discarding these redundant frames, the data storage requirement is reduced while the diagnostic quality and motion detail are recovered through the stitching of remaining unique images.
Data Source
AI summary
A method and apparatus of processing images captured from human gastrointestinal (GI) tract by a capsule camera are disclosed. High frame-rate images captured from human gastrointestinal (GI) tract by a capsule camera are received for processing. The high frame-rate images comprise first images at a first spatial resolution corresponding to a regular frame rate and second images at a second spatial resolution, the first images and the second images are interleaved, and the second spatial resolution is lower than the first spatial resolution. Motion models among the high frame-rate images are derived by applying image registration to the high frame-rate images. The high frame-rate images are stitched according to the motion models to generate stitching outputs comprising stitched images and non-stitched images. The stitching outputs are provided.