Stereoscopic Video Adjustment for Viewer-Aligned 3D Depth
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional stereoscopic imaging techniques often result in unnatural depth perception and viewer discomfort when images captured for a cinema are viewed at home, as they require specific camera orientations and setups that are not easily replicable.
Innovation Solution
A method using graphics processing units (GPUs) to process images from left and right cameras with a common field of view, adjusting virtual cameras to align with viewer dynamics, allowing for real-time adaptation of stereoscopic videos without strict camera positioning requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional stereoscopic imaging uses fixed camera positions and orientations for cinema viewing, then 3D viewing experience is optimized for cinema audiences, but the same setup causes unnatural depth perception and viewer discomfort when viewed at home
Solution Approach 1:
The system dynamically adjusts stereoscopic image parameters based on viewer position and camera geometry. Virtual cameras are repositioned and reoriented in real-time to match the actual capture camera configuration, allowing the stereoscopic display to adapt from static cinema setup to dynamic home viewing conditions, eliminating fixed viewing position constraints
Solution Approach 2:
The system changes multiple parameters simultaneously including virtual camera positions, orientations, interocular distance, and convergence points. By modifying these parameters based on the relationship between capture cameras and actual viewer position, the system transforms incompatible cinema stereoscopic images into comfortable home viewing experiences
2Measurement precision
If stereoscopic images are captured with parallel cameras for optimal 3D effect, then depth perception is improved, but camera setup complexity and positioning requirements increase
Solution Approach 1:
The system creates virtual camera copies that replicate the capture camera positions and orientations. By generating polygon meshes from captured images and placing virtual cameras at corresponding positions, the system preserves accurate depth perception without requiring precise physical camera alignment, as the virtual setup compensates for any physical misalignment
Solution Approach 2:
The system introduces virtual cameras and polygon meshes as intermediaries between the physical capture cameras and the final stereoscopic display. This intermediary layer allows flexible processing and adjustment of camera geometry, enabling accurate depth perception from non-parallel capture cameras without direct physical camera reconfiguration
3Productivity
If stereoscopic video is processed with fixed camera geometry, then processing speed is maintained, but the system cannot adapt to different viewer positions and camera orientations
Solution Approach 1:
The system performs preliminary processing by generating polygon meshes from capture camera images and pre-calculating virtual camera positions and orientations. This preparatory work enables rapid real-time rendering and adjustment when viewer position changes, as the foundational geometric data is already prepared and can be quickly transformed without full reprocessing
Solution Approach 2:
The system creates a universal processing framework that handles multiple camera geometries (parallel, convergent, divergent) and viewer positions through a single unified approach. The virtual camera and polygon mesh system serves multiple functions: capturing images, storing geometric relationships, enabling real-time adjustments, and generating final stereoscopic output, all within one flexible pipeline
Data Source
AI summary
The present disclosure relates to an image modification process used to generate a chosen type of stereoscopic video of the common field of view shared by two cameras without requiring any pre-set and arrangement normally needed for stereoscopic display. The process can use left and right video streams of images originating from a wide range of camera types sharing some portion of their respective fields of view that are not limited to a specific relative orientation (convergent, parallel or divergent) and can be used to adapt in real time the stereoscopic video to be displayed to a viewer where his dynamic inputs are constantly updated. The proposed method can be used to optimize, correct and or allow the use of various apparatuses (webcams, cameras, scanners, etc.) that may come out of the production lines with imperfects to the capture cameras.