Detecting and preventing criminal activities in shopping malls using massive video surveillance based on deep learning models

Video surveillance devices are a valuable tool in various contexts to automate different danger conditions and enable security guards to make effective decisions to improve asset safety. This article suggests detecting and preventing criminal activities in shopping malls (DPCA-SM) framework that detects suspected activity in shopping malls in real-time. The video monitoring approach makes some suggestions that create a comprehensive application capable of effectively tracking people's pathways and detecting measures in a shop setting. The proposed system utilizes the publicly accessible CAVIAR data collection to validate the proposed approach for monitoring occlusions with a performance of nearly 92% to assess the accuracy of the principal inputs of the proposed initiative. The alerts provided by the proposed framework are also evaluated on naturalism, private datasets, demonstrating that in a shopping mall setting, the professional surveillance cameras strategy can efficiently detect unusual activity.