Spherical Model Motion Compensation for Fundus Image Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current technologies face challenges in improving image quality of spherical bodies, such as the fundus oculi, due to the complexity of aligning three-dimensional images, which often require additional equipment like OCT tomographic images or affine transformations, making it difficult to achieve high-quality images with a simple configuration.
Innovation Solution
An image processing device that detects motion vectors using a three-dimensional spherical model to perform motion compensation and synthesis of captured images, allowing for accurate alignment and enhancement of image quality without the need for extensive equipment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If affine transformation is used for rotation of fundus oculi images, then image alignment is performed, but accurate alignment of three-dimensional spherical images is difficult to achieve
Solution Approach 1:
The patent applies a three-dimensional spherical model to represent the fundus oculi, transforming the alignment problem from two-dimensional affine transformation to three-dimensional spherical coordinate transformation. This allows accurate representation of the spherical geometry of the eyeball, enabling precise alignment that respects the curved surface nature of the fundus, thereby resolving the limitation of affine transformation which cannot accurately handle spherical geometry.
Solution Approach 2:
The patent transitions from two-dimensional image alignment to three-dimensional spherical model alignment. By introducing the third dimension (radial distance from the spherical center) and using spherical coordinates, the system can perform motion compensation in three-dimensional space, achieving more accurate alignment for spherical objects compared to conventional two-dimensional methods.
2Measurement precision
If three-dimensional images are formed using OCT tomographic images for alignment, then rotation direction alignment is improved, but device size increases
Solution Approach 1:
The patent makes the fundus camera capable of performing both image capture and three-dimensional spherical model construction functions. By calculating the spherical model from captured images using the known focal length and optical center, the system eliminates the need for separate OCT equipment, achieving multi-functionality with a single device while maintaining alignment precision.
Solution Approach 2:
The patent creates a three-dimensional spherical model as a virtual copy of the fundus oculi based on captured two-dimensional images. This virtual model serves as a reference for alignment without requiring physical OCT equipment, effectively copying the essential three-dimensional geometric information in a computationally efficient manner that reduces device complexity.
3Manufacturing precision
If multiple fundus oculi images are synthesized over a certain period, then image quality is improved, but image quality improvement is hindered if the fundus oculi moves during capture
Solution Approach 1:
The patent performs motion compensation before synthesis by detecting motion vectors between sequential images and applying correction transformations. This preliminary action of aligning images based on detected motion eliminates the negative effect of eyeball movement, ensuring that subsequent synthesis operations combine properly aligned images and achieve high image quality.
Solution Approach 2:
The system implements a feedback loop where motion vectors are continuously detected from sequential images, and motion compensation is applied based on this detected motion information. This closed-loop approach continuously corrects for eyeball movement, maintaining alignment accuracy throughout the capture period and enabling successful synthesis of high-quality images despite physiological movements.
Data Source
AI summary
There is provided an image processing device including a motion vector detection portion that performs comparison of a substantially spherical photographic subject such that, among a plurality of captured images including the photographic subject, an image as a processing target and another image as a comparison target are compared using each of the plurality of captured images as the processing target, and which detects a motion vector of a whole three-dimensional spherical model with respect to the processing target, a motion compensation portion that performs motion compensation on the processing target, based on the motion vector of each of the plurality of captured images that is detected by the motion vector detection portion, and a synthesis portion that synthesizes each of the captured images that are obtained as a result of the motion compensation performed by the motion compensation portion.