AI Based Video and Image Analytics

Physical site security is a strategy for protecting against threats and mitigating the potential negative effects of security incidents. Potential threats must be identified and prioritized according to their criticality and potential impact. The abnormal activity analysis in video scene is very difficult due to several real world constraints. AI-based video surveillance systems are able to analyze video streams in real time and automatically detect suspicious behavior. Using machine learning, these systems can recognize normal behavior patterns and identify abnormal behavior indicating a threat or suspicious activity. This paper develops and discusses a solution for video surveillance based on artificial intelligence (AI), underlying deep learning implementation technology involved. The proposed solution meets the security requirements and allows security operators to react quickly to potential incidents and prevent dangerous situations.