Design And Implement Of Video Objective Quality Evaluation System Based On Druid

With the development of the information age,the way people accept information has changed from simple text information to more intuitive short video.After people upload short video through the app,the application platform will compress the video to save bandwidth,but video compression is damaged,affecting the user's perception.From the point of video application,the increasing video volume relies on the human eye to subjective evaluation efficiency is very ineffective.At the same time,developers of video applications often need A/B experiments to determine the impact of coding optimization on video quality when trying coding optimization,which often costs a lot of time and has a risk.In view of the above problems,the video objective quality evaluation system based on Druid is designed and implemented.The paper mainly compeletes the fllowing works:(1)This system uses a full-reference objective quality evaluation method.Based on the original video,the quality of the encoded and compressed video is evaluated.The pixel difference and fusion index are integrated into the system module,so that the video quality can be measured numerically.(2)This system implements an asynchronous task model.The Web service implemented in Go language is used as the task publisher,and the video objective quality evaluation module developed by Python is used as the task processor.The system uses Gearman middleware to solve problems of calling between different programming languages and uneven load during task scheduling in a distributed environment.(3)Based on Druid which uses Kafka as its data intake,the system implements the analysis and storage of objective video quality evaluation data,and uses the visualization framework Imply to implement the visualization platform.The interface provides authority management,monitoring alarms and other functions.(4)The system has been tested and verified from functional perspectives.The test results show that the system is fully functional and the accuracy of the video objective quality evaluation meets the system requirements.The system has good practicality value.