An Authentication Approach for Lossy Surveillance Videos

Nowadays, there is increased use of surveillance cameras for security purpose in public or in other organizations (banks, airports, traffic etc.). The videos that are captured from these cameras need to be secured, as the video content may contain confidential information or any other activities that need to be analyzed. For analyzing the video content, videos are need to be transmitted across the internet for different authorities. But, due to wireless transmissions there would be packet loss or any change in the video content. Different authentication methods are used to authenticate videos, but these techniques sacrifice video distinctiveness. Hence we describe a novel approach for securing these videos, which is called video fingerprint generation and matching technique. Our proposed approach ensures trust for videos that are transmitted via internet. Experimental results have shown concise and accurate video fingerprint generation.