Research and application of large-scale vehicle video hash technology

In today's network environment, the application and dissemination of multimedia are becoming more and more extensive, and people are increasingly inseparable from the internet and multimedia in their daily lives, with a wide range of video applications emerging. But with this comes the problem of video security protection. Many images and videos on social networking platforms are stolen and re-published by other users for secondary creation. Therefore, it is not only necessary to protect the copyright of original videos, but also to ensure the information security during the storage and application of these video data. This paper proposes a video hashing algorithm based on the combination of global hashing based on invariant moments and partial hashing based on singular value decomposition, and the algorithm is used to detect the similarity of large-scale vehicle videos. By comparing with other video hashing algorithms, this paper's algorithm shows better results in terms of storage space, running time and recognition accuracy.