3D Image Reconstruction Using Selective Movement Compensation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for improving image quality in user-selectable stereoscopic media, such as 3DTV broadcasting, face challenges in efficiently transmitting and receiving additional information for reconstructing high-resolution images, especially in environments with low bit rates and varying image resolutions.
Innovation Solution
The proposed solution involves an apparatus and method that generate and receive additional information for reconstructing high-resolution images using movement compensation, conditional disparity, and movement compensation techniques, based on previous frames and encoded images, to enhance image quality by selecting appropriate decoding modes and utilizing disparity and movement information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If additional information is transmitted for reconstructing high-resolution images, then image quality is improved, but transmission bandwidth and complexity increase
Solution Approach 1:
The patent extracts only the essential additional information needed for high-resolution reconstruction (motion vectors, disparity vectors, block classification data) rather than transmitting complete high-resolution frames. This selective extraction maintains image quality improvement while minimizing bandwidth consumption.
Solution Approach 2:
The transmitting device performs preliminary processing to generate and classify blocks requiring additional information before transmission. By pre-identifying and preparing only the necessary additional data for specific blocks, the system avoids transmitting redundant information, thus improving image quality where needed while conserving bandwidth.
2Manufacturing precision
If multiple types of additional information are transmitted for different blocks, then reconstruction accuracy is improved, but device complexity increases
Solution Approach 1:
The patent divides the image into multiple blocks and applies different types of additional information to different blocks based on their characteristics. This segmentation allows the system to improve reconstruction accuracy for specific blocks that need it while avoiding unnecessary processing complexity for blocks that don't require additional information.
Solution Approach 2:
Different blocks are assigned different types of additional information (motion compensation for some blocks, disparity compensation for others, none for certain blocks) based on local image characteristics. This local quality approach improves reconstruction accuracy where needed while reducing overall processing complexity by not applying uniform complex processing to all blocks.
3Measurement precision
If additional information is transmitted for all blocks, then image quality is improved, but transmission efficiency decreases
Solution Approach 1:
The patent applies additional information transmission selectively to only those blocks that benefit from it, rather than transmitting additional information for all blocks. This partial action approach maintains image quality for blocks that need enhancement while improving transmission efficiency by avoiding redundant data transmission for blocks that can be adequately reconstructed without additional information.
Data Source
AI summary
A method and apparatus for generating additional information used to reconstruct an additional image through steps of: generating information for movement compensation based on the original right image of the stereoscopic image and the previous frame of the right image; and generating first additional information for reconstructing the right image to a high resolution based on the original right image and the information for movement compensation are provided.