Stereoscopic Image Disparity Correction for Virtual Cameras
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional stereoscopic display technologies face difficulties in rendering objects close to virtual cameras, leading to excessive disparity between images for the left and right eyes, making it challenging for users to view objects that project significantly towards or recess from the display surface.
Innovation Solution
An image processing system that generates stereoscopic images by correcting the disparity of objects based on their distance from virtual cameras, using intermediate virtual cameras for interpolation and coordinate transformations to adjust the positions of objects in the images for the left and right eyes, thereby reducing the degree of projection and maintaining natural object movement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If conventional stereoscopic display is used with objects close to virtual cameras, then the object is displayed with great disparity projecting from the display surface, but it becomes difficult for users to view the object
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting the disparity value based on the object's distance from the virtual camera. When an object is closer than a predetermined distance, the system reduces its disparity value to a smaller range, thereby modifying the stereoscopic parameters to ensure comfortable viewing while preserving depth information.
Solution Approach 2:
The system implements dynamics by making the disparity adjustment adaptive rather than static. The disparity correction amount varies dynamically according to the object's position in depth, with closer objects receiving greater correction. This dynamic approach allows the system to maintain natural depth perception for distant objects while preventing excessive projection for close objects.
2Ease of operation
If disparity correction is applied to objects, then the degree of projection is reduced and viewability improves, but the positional relation among objects in the depth direction must be maintained
Solution Approach 1:
The patent applies local quality by selectively correcting disparity for specific objects based on their individual characteristics and positions. Each object's disparity is adjusted independently according to its distance from the virtual camera, allowing localized optimization of viewability without affecting the overall spatial relationships among multiple objects in the scene.
Solution Approach 2:
The system uses an intermediate virtual camera positioned between the left and right virtual cameras as a mediator for disparity calculation. By referencing positions from this intermediate camera, the system maintains accurate depth relationships among objects while applying correction, effectively using the intermediate camera as a reference frame that preserves spatial composition.
Data Source
Figure 1
Figure 2(a)~2(d)
Figure 3
AI summary
Stereoscopic image generation means for taking images of a virtual three-dimensional space in which an object is located, with a virtual camera, and generating a stereoscopic image that is composed of an image for a left eye and an image for a right eye between which a disparity is provided, and display control means for causing a display section to display the stereoscopic image generated by the stereoscopic image generation means, are provided. In generating the stereoscopic image, the stereoscopic image generation means performs correction of decreasing a disparity of the object.