Video Quality Estimation Using Network Parameters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video data quality estimation methods are ineffective in scenarios where a specific video stream is unavailable, as they rely on specific statistic parameters like V_CCF, V_PLEF, and V_IR, which are not accessible in such cases.
Innovation Solution
A video data quality estimation method and apparatus that calculates statistic parameters using a fitting model and probabilistic model to estimate video quality, even when specific video streams are unavailable, by deriving parameters from available data at the network planning layer and applying them to quality models for encoding, packet-loss, and rebuffering.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If specific statistic parameters (V_CCF, V_PLEF, V_IR) are used for video quality estimation, then measurement precision is improved, but adaptability deteriorates because these parameters are not accessible when specific video streams are unavailable
Solution Approach 1:
The patent changes the parameters used for quality estimation from specific video stream parameters (V_CCF, V_PLEF, V_IR) to network planning layer parameters (bitrate, resolution, frame rate, packet loss rate) that are universally accessible. This parameter substitution resolves the contradiction by maintaining measurement capability while improving adaptability to scenarios where specific video stream data is unavailable.
2Measurement precision
If packet header parsing is performed to obtain packet loss information, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent extracts only the essential packet loss rate parameter from the complex packet header information, rather than performing full packet header parsing. This selective extraction maintains sufficient measurement precision for quality estimation while significantly reducing the complexity of the quality estimation apparatus by avoiding detailed packet header analysis.
3Measurement precision
If multiple quality models (encoding, packet-loss, rebuffering) are integrated, then measurement precision is improved, but ease of operation deteriorates
Solution Approach 1:
The patent merges the multiple quality models (encoding quality model, packet-loss quality model, rebuffering quality model) into a unified quality estimation framework. By combining these models and using a standardized set of network planning layer parameters for all three models, the patent achieves comprehensive video quality measurement while maintaining operational simplicity through parameter unification and integrated processing.
Data Source
Figure 1~2
Figure 3~4
AI summary
The present invention relates to the field of digital signal processing technologies, and discloses a video data quality estimation method and apparatus. The video data quality estimation method includes: obtaining a parameter used to obtain video quality of the video, where the parameter used to obtain video quality of the video includes at least one of a content complexity factor of the video, an average video frame impairment rate of the video, a sequence impairment rate of the video, or a packet-loss event frequency of the video; and obtaining quality of the video according to the parameter used to obtain video quality of the video. According to the present invention, quality estimation can be performed on video data in an application scenario in which a specific video stream is unavailable.