Cloud-assisted secure video transmission and sharing framework for smart cities.

With the innovations and increasing popularity of smart cities, video analytics (e.g., from social media, entertainment, surveillance, smart health monitoring, and crowd management) are used in a range of application domains to provide safety, security, and well-being for residents. As these video analytics are shared through highly interconnected devices, sensors, and other smart city stakeholders, security and integrity is a concern for secure video content. Without proper techniques to address the security threats of such transmission and sharing, it would be very difficult for a smart city government to provide safety and quality-of-life to its residents. To this end, this paper proposes a cloud-assisted framework for secure video transmission and sharing, where mobile clients have limited capabilities; however, users need to share videos seamlessly without sacrificing integrity and quality. In the proposed framework, once the video is captured by a smart phone, the key frames are detected. The key frames are detected by applying a genetic algorithm (GA). A discrete wavelet transform (DWT) based watermarking algorithm is used to insert the watermark into the key frames. A two layer protection mechanism based on error correcting codes is applied to the signature or identity of an individual to generate the watermark that is inserted into the video to protect it from any transmission loss, distortion or attacks. Watermarked video is then transmitted to the cloud. The watermark is also shared to a number of clouds in a space efficient and secret-robust way. To check the integrity of the video, the original user can at any time download the video, extract the watermark, and verify it. To test the quality of the watermarked video, imperceptibility analysis is carried out using PSNR and structural similarity index measure (SSIM). The experiment results demonstrate that the watermarking algorithm does not degrade the quality of the video.