Epipolar Plane Image View Extrapolation With Super-Resolution Disparity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to simultaneously perform super-resolution, disparity estimation, and view extrapolation of target images effectively using epipolar plane images.
Innovation Solution
A method involving creating an epipolar plane image (EPI) from a target image and depth map, applying Gaussian blur, warping a super-resolution EPI with a disparity, and using a Generative Adversarial Network (GAN) to restore the EPI, thereby enabling simultaneous super-resolution, disparity estimation, and view extrapolation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If existing methods are used to perform super-resolution, disparity estimation, and view extrapolation, then each operation can be performed separately, but the overall efficiency and consistency are reduced due to multiple separate processing steps
Solution Approach 1:
The patent combines super-resolution, disparity estimation, and view extrapolation operations into a single unified neural network architecture that processes the epipolar plane image simultaneously. This merging of previously separate processing steps into one integrated system improves processing efficiency and ensures consistency across all operations while managing system complexity through a cohesive architectural design.
2Productivity
If a unified neural network architecture is used to simultaneously perform super-resolution, disparity estimation, and view extrapolation, then processing efficiency improves, but the complexity of the network architecture increases
Solution Approach 1:
The unified neural network architecture is designed to perform multiple functions simultaneously - super-resolution, disparity estimation, and view extrapolation - all within a single multi-functional system. This universal approach allows the network to process multiple tasks in parallel, improving overall productivity while the modular design of the network helps manage the inherent complexity by organizing different functions within a cohesive framework.
3Manufacturing precision
If the EPI is directly used without super-resolution, then processing is simpler and faster, but the image quality and extrapolation accuracy are reduced
Solution Approach 1:
The super-resolution process is applied as a preliminary action to the epipolar plane image before performing disparity estimation and view extrapolation. By enhancing the image quality in advance through super-resolution, the subsequent processing steps benefit from higher accuracy inputs, leading to improved final results. This preliminary enhancement justifies the additional processing complexity by ensuring high-quality output throughout the pipeline.
Data Source
AI summary
The present disclosure relates to a method of inferring an epipolar plane image and extrapolating a view of a target image using the epipolar plane image. A view extrapolation method using an epipolar plane image according to an embodiment of the present disclosure includes: creating an Epipolar Plane Image (EPI) using a target image and a depth map corresponding to the target image; creating a super-resolution EPI and a disparity of the EPI on the basis of the EPI; creating an EPI mask by warping the super-resolution EPI in accordance with the disparity; and creating a restored EPI by applying the EPI mask to the warped super-resolution EPI, and creating an extrapolated image for the target image using the restored EPI.