Terminal Video Quality Evaluation Using Movement-Aware Frame Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current no-reference video quality evaluation methods in the time domain on terminal sides suffer from high evaluation errors, overlook movement impacts, and provide single indicators, failing to accurately assess video quality in real-time streaming services.
Innovation Solution
A method that calculates the significant movement area proportion of video frames, performs frozen, scenario-conversion, jitter, and ghosting frame detections, and determines video quality using scenario information weights and distortion coefficients, effectively reducing evaluation errors and highlighting movement impacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If full-reference evaluation methods are used to detect frame loss by aligning tested video with original video, then detection capability is improved, but adaptability to real-time streaming services deteriorates
Solution Approach 1:
The patent extracts and removes the dependency on original reference video from the evaluation system. By using only the tested video itself to detect frame loss through frame difference analysis, the system eliminates the need for original video alignment while maintaining detection capability, thus adapting to real-time streaming where original video is unavailable.
Solution Approach 2:
The patent creates a virtual reference by using the previous decoded frame as a substitute for the original video frame. This copying approach allows the system to perform reference-based evaluation without actually needing the original video, enabling adaptation to streaming scenarios while maintaining measurement precision.
2Device complexity
If simple frame difference methods are used to evaluate time domain quality, then computational complexity is reduced, but measurement precision deteriorates
Solution Approach 1:
The patent segments the quality evaluation process into multiple distinct components: frame difference calculation, significant movement area detection, and weighted quality scoring. This segmentation allows each component to be optimized independently, maintaining low computational complexity while improving overall measurement precision through targeted analysis of different aspects.
Solution Approach 2:
The patent applies local quality assessment by identifying and weighting significant movement areas differently from static areas. Instead of uniform frame difference evaluation, the system focuses computational resources on regions with meaningful changes, improving measurement precision without proportionally increasing overall complexity.
3Ease of operation
If traditional no-reference methods only calculate luminance difference between frames, then ease of operation is improved, but measurement precision deteriorates due to overlooking movement impact
Solution Approach 1:
The patent introduces dynamic adaptation by detecting significant movement areas and adjusting the evaluation strategy accordingly. The system dynamically identifies regions with substantial changes between frames and focuses quality assessment on these areas, moving from static uniform evaluation to dynamic region-specific evaluation, thereby improving precision while maintaining operational simplicity.
Solution Approach 2:
The patent changes the evaluation parameters from simple luminance difference to a composite metric that includes movement-aware weighting. By introducing significant movement area proportion as a new parameter and using it to weight the quality calculation, the system improves measurement precision while keeping the overall method straightforward to implement.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
Disclosed are a method and device for valuating quality of a video in a time domain on a terminal side. The method comprises that: a significant movement area proportion of each video frame is calculated, video frames are divided into absolute regular frames and suspected distorted frames according to the significant movement area proportion of each video frame; a frozen frame detection, a scenario-conversion frame detection, a jitter frame detection, and a ghosting frame detection are performed on the suspected distorted frames; the video is split into scenarios according to the result of the scenario-conversion frame detection, scenario information weight of each scenario is calculated, and the quality of the video in time domain on the terminal side is determined. The present disclosure solves the problems for the no-reference technology in time domain on the terminal side in the related art that the evaluation error is big, the movements is overlooked, and the indicator is single, thus increasing the closeness of the evaluation result to subjective perception, expanding an evaluation system of time domain distortions of the video, and reducing the probability of misjudgments.