Light Field Image Encoding via Segmented Depth Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Light field imaging technologies face challenges in generating high-quality images efficiently due to the large data structures required, which can be slow to render, display, and transmit, necessitating a method to reduce data size and processing power while maintaining high-resolution display quality.
Innovation Solution
A system and method that acquire and encode light field images using a computing system with a camera array and display, employing raster-based encoding to generate and decode light field images and videos, enabling compact transmission and high-quality display on weaker computing systems by compressing data using techniques such as zigzag compression and metadata storage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If many high resolution images are used to generate light field images, then image quality is improved, but data structure size increases and processing speed decreases
Solution Approach 1:
The patent segments the light field data into multiple view images and depth maps, processing them separately through different encoding paths. This segmentation allows high-resolution image quality to be maintained while reducing the overall processing burden by dividing the data into manageable components that can be compressed and transmitted efficiently.
Solution Approach 2:
The patent uses depth maps as simplified copies or representations of the actual 3D scene geometry. Instead of transmitting and processing all high-resolution view images at full quality, the system uses depth information as a compact proxy that enables efficient compression while preserving the ability to reconstruct high-quality light field images on the receiving end.
2Manufacturing precision
If many high resolution images are used to generate light field images, then image quality is improved, but data structure size increases
Solution Approach 1:
The patent extracts depth information from the light field data as a separate component. By taking out the depth map as an independent element, the system can compress the color images and depth information separately using different encoding strategies, significantly reducing the total data size while maintaining the quality needed to reconstruct high-resolution light field images.
Solution Approach 2:
The patent changes the representation parameters by converting 3D spatial information into 2D depth maps and using compressed video formats. This parameter transformation allows the same visual information to be stored and transmitted in a much more compact form, reducing data structure size by up to 90% while preserving image quality through efficient decoding processes.
3Loss of information
If traditional light field encoding is used, then complete light field data is preserved, but processing power and storage requirements increase
Solution Approach 1:
The patent implements dynamic encoding where the level of compression and processing can be adjusted based on available resources. The system can adaptively choose between different compression ratios and processing depths, allowing complete light field data to be preserved when needed while enabling reduced-processing modes for systems with limited computational power, thus resolving the contradiction between data completeness and processing requirements.
Data Source
AI summary
A system and method can include receiving a set of views, encoding the set of views, and displaying the set of views such that they are perceived as a holographic image.