Composite video quality assessment method, device, equipment and medium

Abstract

The invention discloses a synthetic video quality evaluation method, device and equipment and a medium. The method comprises the following steps: acquiring a video request message and a response message, and extracting video file information and a response state code; judging the process type of the video according to the video file information and the response state code; performing video synthesis association on the video request message and a subsequently received video request message according to the process type and the response state code; judging whether the video is transmitted completely or not, and outputting a synthetic video receipt if the video is transmitted completely; and according to the synthetic video document, calculating a deep message detection DPI index. According to the invention, deep message detection is utilized to synthesize video service receipts, the receipts of dispersed transaction levels are connected in series, and related experience evaluation information is output by taking a complete video service as a unit.

Description

technical field [0001] The present invention relates to the technical fields of big data and Internet business, and in particular to a synthetic video quality evaluation method, device, equipment and medium. Background technique [0002] There are mainly three schemes for evaluating video quality: [0003] Solution 1: When evaluating user service perception based on HTTP services, the original in-depth message detection method is based on a single request (that is, transaction-level) output of business documents, including transaction-level download traffic, response delay and other information, and a complete The video is split into numerous transaction-level document outputs, each transaction-level document calculates the evaluation indicator, and the evaluation method of the complete video becomes the average value of the transaction-level document indicators. [0004] Solution 2: Self-operated video reporting Video awareness, through mobile self-operated video APP, inte...

AI Technical Summary

Problems solved by technology

[0006] When evaluating user service perception based on HTTP services, the original in-depth message detection method is based on a single request (that is, transaction level) to output business documents, including transaction-level download traffic, response delay and other information. Transaction-level index data is evaluated Web browsing, instant messaging and other small-packet services can truly reflect user perception, but because the data basis of this method is transaction-level, it is not applicable to video services where one service consists of N transactions. Browsing services have large differences in data packet size and resource scheduling, and transaction-level analysis cannot accurately evaluate the user's perception during the entire video service process

[0007] Reporting video service experience indicators through self-operated videos can only analyze the video experience of self-operated businesses, but cannot analyze other mainstream videos such as Youku, Tencent, iQiyi, etc.

[0008] Use the test app for testing, 1) limited by the video resource type of the app, for example, the SpeedVideo video test app only has resources above 720P and only the resources of a single website (such as Youku), and cannot cover more video resource types; 2) Testers are required to conduct road tests or fixed-point tests, which increases the cost of personnel input; 3) The video experience results are in csv format, which requires manual analysis, and the analysis efficiency is low

Images