Sport Event Video Stitching Using Fixed Camera Mount Geometry
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for stitching video recordings from mobile user devices at sport events suffer from parallax errors due to unknown relative positioning and angling between cameras, particularly when using different types of devices and mounts, resulting in low-quality panoramic videos.
Innovation Solution
A method that utilizes fixed relative camera positioning and angling, combined with identifying and utilizing physical elements like goal posts or pitch markings, to estimate feature locations and improve stitching accuracy by aligning video recordings based on these elements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If video recordings from multiple mobile user devices are stitched together to provide panoramic coverage of a sport area, then the coverage area is improved, but stitching artefacts such as parallax errors occur due to unknown relative positioning and angling between cameras
Solution Approach 1:
The patent introduces a mount as an intermediary device that physically connects multiple mobile user devices and defines their relative positioning and angling. This mount acts as a mediator that ensures fixed geometric relationships between cameras, thereby enabling accurate stitching without parallax errors while maintaining the ability to cover the entire sport area
Solution Approach 2:
The patent changes the parameter of relative camera positioning from unknown to fixed by using a mount with defined geometric parameters. The mount establishes specific angular relationships and positional offsets between cameras, transforming the stitching problem from one with variable parameters to one with controlled, known parameters
2Manufacturing precision
If dedicated sport camera assemblies with multiple image sensors are used to cover the entire sport area, then stitching accuracy is improved, but cost and device complexity increase
Solution Approach 1:
The patent makes mobile user devices universal by enabling them to serve dual purposes: their original functions plus sport event recording. By using mounts to attach multiple standard mobile devices, the system achieves dedicated sport camera assembly performance without requiring specialized equipment, thereby reducing device complexity and cost
Solution Approach 2:
The patent uses copies of existing mobile user devices instead of creating specialized sport camera assemblies. By mounting multiple standard smartphones or tablets, the system replicates the functionality of dedicated multi-camera systems while leveraging the ubiquity and affordability of consumer mobile devices
3Adaptability or versatility
If mobile user devices with different types and mounts are used for recording, then adaptability and ease of operation are improved, but relative positioning and angling between cameras become unknown, worsening stitching accuracy
Solution Approach 1:
The patent applies local quality by ensuring that while the mount can accommodate different device types (providing adaptability), the specific local positioning and angling parameters for each camera mounted on it are precisely defined and fixed. This allows different mobile devices to be used while maintaining the geometric precision needed for accurate stitching
Data Source
Figure 1
Figure 2a
Figure 2b
AI summary
The present disclosure relates to a method for stitching video recordings of a sport event. The method comprising the steps of: providing a first video recording and a second video; determining a first and second pluralities of feature points in the first and a second video recordings, wherein the first and second pluralities of feature points have a common set of feature points; identifying an element portion of a physical element present in at least one of the video recordings, wherein the element portion is associated with a physical size; estimating a feature location of feature points of the common set based on an in-frame position of the element portion in the video recordings and on the physical size of the element portion; and stitching the first video recording and the second video recording based on the feature location to obtain a panoramic video of the sport event.