Copyright Storage Method of Dance Short Video Based on Blockchain

Abstract

Blockchain technology is widely used in the field of digital rights protection. The traditional digital rights protection scheme is inefficient, highly centralized, and has the risk of being tampered with. At the same time, blockchain cannot save all original files of digital resources due to its storage size limitation. Firstly, this paper proposes a dynamic limb action recognition algorithm based on dance pose prediction (RADPP). This algorithm recognizes the dance action in dance short video based on deep learning. The algorithm extracts the features of dance action accurately and forms a log file that can represent the short dance video. The algorithm stores the dance pose file into the blockchain to the purpose of preservation. On the one hand, the above methods improve the efficiency of short dance video storage, reduce the degree of centralization of storage, and eliminate the risk of copyright information being easily tampered with; at the same time, the log file calculated by deep learning technology for short video not only ensures the privacy of copyright information, but also ensures the feasibility of storing video information into blockchain. Experiments show that the method proposed in this paper is lighter and more efficient than the existing copyright storage methods, and this method can provide technical support for the media resource management department.