Security standards for real time video surveillance and moving object tracking challenges, limitations, and future: a case study

• Published in: Multimedia tools and applications 2024-03, Vol.83 (10), p.30113-30144
• Main Authors: Kumar, Chandan, Singh, Shailendra
• Format: Article
• Language: English
• Subjects: Algorithms; Case studies; Computer Communication Networks; Computer Science; Data Structures and Information Theory; Encryption; Entropy; Frames (data processing); Learning; Lightweight; Moving object recognition; Multimedia Information Systems; Real time; Security; Special Purpose and Application-Based Systems; Surveillance; Tracking; Video data; Weight reduction
• ISSN: 1573-7721; 1380-7501; 1573-7721
• DOI: 10.1007/s11042-023-16629-7

Uses of video surveillance (VS) have exponentially increased using the internet as a platform. Therefore, security issues in such real time videos need to be addressed. Video may have multiple moving objects in a frame and different features in video lengths. Hence, designing content-based video encryption standards and lightweight crypto-encryption standards are needed to serve the real-time requirement for securing surveillance videos. This paper presents a survey and case study of various encryption standards of VS, which secures video data and used for object tracking (OT). In the first part, a fast and secure modified AES method is proposed. The performance is compared using NCPR and UACI measures with state-of-the-art encryption standards. Modified AES is lightweight and offers nearly 99% NCPR value and offers the fastest elapsed time. In the second part, the paper has proposed an entropy adaptive object learning model for only securing frames containing moving objects in the scene. Percentage frame adaption is achieved by the entropy threshold. The proposed entropy learning model based on content adaptive security standards has proven to save memory requirements by encrypting the desired frames only. In third part, weight average background subtraction (BS) approaches are used for evaluation which performs better for multi OT. The paper has designed learning and adapting the best RGB space to apply tracking. The qualitative expected outcomes are presented for real-time captured videos with different object motions. The performance comparison of the entropy and crypto weights is compared for selected light weight ciphers under the consideration of multiple object motion of various real time videos. The paper finally addresses various challenges, open issues and future scopes of VS systems. Graphical abstract