Video Scene Detection for Motion Sickness Preview in HMD Viewing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing head-mounted displays (HMDs) can cause visually induced motion sickness due to factors like motion differences between the viewer and the video, non-smooth movement, and 3D image incompatibilities, which existing detection technologies fail to adequately address before viewing, especially for inexperienced users.
Innovation Solution
An image processing apparatus detects scenes likely to cause motion sickness by analyzing factors such as viewpoint motion, object motion, left-right image incompatibility, and 3D unnaturalness, generating a partial video showcasing these characteristics to warn viewers before watching the full content.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If video content is processed to detect and remove visually induced motion sickness scenes, then viewer comfort is improved, but video content completeness and entertainment value deteriorate
Solution Approach 1:
The system performs preliminary detection of visually induced motion sickness scenes in the video content before the viewer watches. It generates a characteristic video showcasing these potentially harmful scenes, allowing viewers to make informed decisions about whether to watch the full content or skip to safer portions, thus preventing motion sickness without forcibly removing content.
Solution Approach 2:
A characteristic video is created as an intermediary between the original video content and the viewer. This intermediate representation displays only the visually induced motion sickness scenes, serving as a warning preview that enables viewers to avoid uncomfortable portions without eliminating the original content's completeness.
2Object-affected harmful factors
If skilled camera operators are used to shoot VR videos, then visually induced motion sickness is reduced, but production cost and complexity increase
Solution Approach 1:
The system enables video content to evaluate itself by automatically detecting visually induced motion sickness scenes through algorithmic analysis of video data. This self-detection capability eliminates the need for skilled camera operators during production, as the post-processing system can identify problematic scenes regardless of how the video was originally shot.
Solution Approach 2:
The patent replaces the mechanical dependency on skilled human camera operators with an automated computational system. The detection unit uses algorithms to analyze video content and identify visually induced motion sickness scenes, substituting human expertise with machine-based automated analysis.
3Object-affected harmful factors
If frame thinning is applied to process visually induced motion sickness scenes, then viewer comfort is improved, but video quality and smoothness deteriorate
Solution Approach 1:
Instead of applying frame thinning that removes portions of scenes, the system extracts and separates visually induced motion sickness scenes from the rest of the video content. The characteristic video displays only these extracted problematic scenes as a warning, while the original video remains intact for viewers who choose to watch it.
Data Source
AI summary
An image processing apparatus includes one or more memories and one or more processors. The one or more memories and the one or more processors are configured to detect a specific scene that causes visually induced motion sickness from a video for a predetermined display device, generate a partial video corresponding to a predetermined period including the detected specific scene in the video, and present the generated partial video.