VR Video Encoding with Foveated Quality and Adaptive Field of View
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing virtual reality systems face challenges in providing adequate video quality due to the close proximity of the video to the user's eyes, leading to noticeable artifacts and impairments.
Innovation Solution
The system employs a backend architecture that adjusts quantization parameters, applies gradient scaling, and adaptively shapes the field of view to improve video quality in virtual reality systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If existing video compression techniques are used for virtual reality, then the video can be transmitted and displayed, but the video quality is insufficient and artifacts are noticeable due to close proximity to the eye
Solution Approach 1:
The patent applies different quantization parameters to different regions of the video frame based on the human visual acuity model. The central region (foveal vision) uses higher quality encoding with lower quantization parameters, while peripheral regions use lower quality encoding with higher quantization parameters. This local differentiation of quality matches the non-uniform sensitivity of the human eye, improving perceived video quality while reducing overall compression artifacts.
Solution Approach 2:
The patent dynamically adjusts quantization parameters based on the visual importance of different video regions. By changing the quantization parameter values according to the human visual acuity model and lens distortion characteristics, the system optimizes video quality in critical viewing areas while maintaining acceptable quality in less critical areas, thereby resolving the contradiction between compression efficiency and artifact reduction.
2Adaptability or versatility
If video is displayed close to the eye in virtual reality, then immersive experience is improved, but visual acuity requirements increase making artifacts more noticeable
Solution Approach 1:
The patent implements region-specific quality optimization by applying different quantization parameters to different portions of the video frame. The central foveal region, which corresponds to the area of highest visual acuity and most importance for immersion, receives superior quality encoding. This local quality enhancement ensures that the immersive experience benefits from high quality video where the user's attention is focused, while tolerating lower quality in peripheral regions.
Solution Approach 2:
The system dynamically adapts the video encoding parameters based on the user's viewing characteristics and the importance of different video regions. By making the quantization parameters dynamic and region-dependent rather than uniform, the system optimizes the balance between immersive experience and video quality, ensuring high quality where needed while maintaining overall compression efficiency.
3Ease of manufacture
If uniform quantization parameters are used across the entire video frame, then encoding is simple, but video quality is degraded in the central viewing area where visual acuity is highest
Solution Approach 1:
The patent replaces uniform quantization parameters with region-dependent quantization parameters based on the human visual acuity model. The encoding process divides the video frame into different regions (central foveal region and peripheral regions) and applies appropriate quantization parameters to each region. This local quality approach significantly improves video quality in the central viewing area while maintaining reasonable encoding complexity through systematic region classification.
Solution Approach 2:
The system changes the quantization parameters from uniform to non-uniform based on spatial location and visual importance. By implementing parameter changes that reflect human visual characteristics, the patent improves video quality in critical central regions without requiring completely complex encoding algorithms, as the parameter changes follow a systematic model-based approach.
Data Source
AI summary
A system and method for improving the quality of video for virtual reality systems uses several different techniques to improve the quality of the video. The different techniques may include quantization parameter maps, gradient scaling, using analytics to identify most view scenes and encode the most viewed scenes with better quality and adaptively shaping the field of view.