Faster abnormality localization and recognition in a secured video bitstream by implementation of video encryption techniques

Abstract

Cloud video storage uses an encrypted format to protect user data. It means encrypted video processing is an essential part of secured cloud storage. In order to detect suspicious or anomalous behavior, video surveillance must have encrypted cloud access. The primary goals of this research are to estimate parameters and detect abnormalities in an encrypted video bitstream. Various typical properties of video encoding frameworks and format-compliant encryption algorithms are investigated to identify abnormalities in an encrypted video bitstream using format-compliant encryption. The encrypted bitstream is decrypted to get three different kinds of enhancement features: the sizes of macroblocks, partitions of macroblocks, and the magnitude of the motion vector. The identification and localization methods now do not include video decryption or complete decompression. The proposed strategy has been created to implement the video encryption scheme efficiently and is compatible with various video encryption techniques. The experimental findings demonstrate that, in comparison to other methods, the proposed approach provides optimal running time and detection rate performance.