Dynamic Camera Pair Selection for Stereoscopic Views
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing mediated reality systems face challenges in generating high-quality stereoscopic views from sparse image data captured by a multicamera array, particularly in surgical applications, due to complex computations and poor image quality.
Innovation Solution
A method for selecting a dynamic pair of cameras based on the tracked position of a display device and/or an instrument within the scene, cropping images based on a known calibration and desired disparity, and combining with 3D reconstructed views to generate stereoscopic and magnified views.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If reconstructed images are generated using image data from multiple cameras positioned at different angles and orientations, then the completeness of scene coverage is improved, but the computational complexity increases and image quality deteriorates when based on sparse image data from relatively few cameras
Solution Approach 1:
The system segments the task by selecting only the necessary pair of cameras for generating stereoscopic views, rather than processing all cameras in the array. This selective approach reduces computational complexity while maintaining adequate image quality for the specific viewing task.
Solution Approach 2:
The system uses a partial subset of cameras (specifically pairs of cameras) to generate stereoscopic views, rather than requiring all cameras to be used. This partial action approach achieves the desired stereoscopic effect with reduced computational burden and acceptable image quality.
2Quantity of substance
If reconstructed images are generated using image data from multiple cameras positioned at different angles and orientations, then the completeness of scene coverage is improved, but the image quality deteriorates when based on sparse image data from relatively few cameras
Solution Approach 1:
The system applies local quality by selecting camera pairs that are specifically positioned to provide optimal stereoscopic views for the current scene content and viewing requirements. This localized selection ensures high image quality for the specific viewing task rather than uniformly processing all cameras.
Solution Approach 2:
The system dynamically selects which camera pairs to use based on real-time analysis of the scene and viewing requirements. This dynamic adaptation allows the system to optimize image quality for each specific situation rather than using a fixed set of cameras, achieving better image quality from sparse data when needed.
3Adaptability or versatility
If a dynamic pair of cameras is selected based on tracked position of display device and instrument, then the relevance of image data is improved, but the processing time increases
Solution Approach 1:
The system performs preliminary tracking of display device and instrument positions to pre-determine which camera pairs will be most relevant for the current viewing configuration. This preliminary action allows the system to efficiently select appropriate camera pairs without requiring extensive real-time processing, thus reducing overall processing time while maintaining high adaptability.
Data Source
AI summary
Methods for selecting pairs of cameras in a multicamera array configured to image a scene to generate stereoscopic views, and associated systems and devices, are disclosed herein. In some embodiments, a representative method includes selecting a pair of cameras in the multicamera array from which to generate a stereoscopic view, and receiving images of the scene from the selected pair of the cameras. The method further includes cropping the images based on a known calibration of the cameras and a desired disparity, and displaying the cropped images on a display device to generate the stereoscopic view. The desired disparity can correspond to an interpupillary distance of a user.