Design And Implementation Of A Video Rendering Crowdsourcing Task Management Platform Based On Blockchai

With the rapid development of high-speed Internet,online collaboration gradually comes into sight.As a distributed problem-solving mechanism,crowdsourcing technology is a typical representative of large-scale social collaboration,innovatively applied to various fields.As application scenarios become increasingly complex,crowdsourcing tasks get more dynamic and diversified,which undoubtedly aggravates the processing difficulty of traditional crowdsourcing platforms.Therefore,constructing a novel crowdsourcing task management platform with fairness,reliability,security has far-reaching research significance and broad application value.As an innovative technology for trust establishment in a distributed environment,blockchain subverts the centralized architecture design of traditional systems.Considering the defects of traditional crowdsourcing platforms,such as single-point failure,poor service stability,and lack of fairness,the advantages of blockchain in collective consensus,crossvalidation,and other aspects make up for the shortcomings,so it naturally becomes the key to establishing decentralized crowdsourcing platforms.With video rendering as the application scenario of crowdsourcing,this thesis researches designing a blockchain-based crowdsourcing task management platform and focuses on the design of task assignment mechanism and result verification method used in practice for video rendering crowdsourcing.The main research contents are as follows:First,given the high cost,low efficiency,low credibility,and other common defects,this thesis introduces crowdsourcing and blockchain technology into the field of video rendering.The system model and architecture of the blockchain-based video rendering crowdsourcing task management platform are designed,realizing the whole crowdsourcing process without any trusted third party.The theoretical security analysis proves that the proposed system is effective against common malicious threats and possesses remarkable technical advantages compared to other similar systems.Moreover,with the design of three different types of smart contracts and the implementation of several functional algorithms,a prototype system based on Ethereum is realized,and many performance tests are carried out on it.The experimental results prove the proposed system has low time overhead and cost.Second,in order to improve the overall satisfaction of task assignment in crowdsourcing,an original task assignment mechanism is designed for video rendering crowdsourcing which contains task matching and task scheduling mechanisms.The task matching mechanism introduces the analytic hierarchy process(AHP)to creatively calculate the preference list of both tasks and workers and realizes the reasonable task matching combining the preference calculation results.The task matching mechanism proves to effectively improve the satisfaction of crowdsourcing participants.Based on the results of task matching,the task scheduling mechanism is proposed to realize the splitting and scheduling of a single task and lays a foundation for the result verification by cleverly designing the random crossing subtasks assigned for workers.Third,considering the characteristics of crowdsourcing applications,this thesis establishes an elastic reputation mechanism based on workers’ historical behavior and presents a reputation-based verification method for video rendering crowdsourcing task results.Specifically,the designed task scheduling mechanism is to make the parts assigned to different workers in a single task have random crossover.Combining the idea of cross verification and the reputation value of answer providers,the correct answers of the cross tasks are obtained,with which the final results of all workers can be verified.After experimental analysis,the result verification method proposed in this thesis is turned out to realize the high accuracy rate of verification in video rendering crowdsourcing tasks.