ESKVS: efficient and secure approach for keyframes-based video summarization framework.

Video security has emerged as a critical study issue in multimedia security in recent years as videos are the most effective and widely used multimedia format. They immediately establish a connection with users. It is necessary to prevent this sensitive information from being stolen or destroyed in various domains, including the military, finance, and education. It can be achieved by either hiding its significance, turning it into a secret code through encryption, or doing both simultaneously. Because video data is generated, transmitting it securely is challenging, and there is resource wastage of memory, processing, and bandwidth. Users must spend a lot of time and effort scanning through this enormous amount of video information in search of the required information. Therefore, a secure keyframes-based generic video summarization (VS) model is proposed to generate a secure video summary. First, Secret Keyframes (SKs) are extracted from the video through proposed Probability VS (PBVS) and Extended (E-PBVS). Second, multi-secret image sharing is provided to the SKs by the proposed EBEMSS (Enhanced Blockwise Encryption based Multi Secret Sharing) scheme, which uses a polynomial congruence concept for keyframe security. The proposed model E-PBVS achieved an average F-score of 0.76 and 0.81 on two benchmark (OV and YT) datasets, respectively, showing its effectiveness in producing informative video summaries. Additionally, EBEMSS outperforms the other related security models. The proposed model outperforms with generated summary and security compared to other keyframe-based VS and MSS techniques. [ABSTRACT FROM AUTHOR]