Automatic Threat Recognition using Surveillance cameras
Mr. Arun Prasath.S B.E. Arun Prasath.S B.E., , Alex .J B.E.,, Guru Ganesh .S B.E., Dr. T. Vijay Anand M.E., Ph.D. Mr.K. SUBR, Alex .J B.E. , Guru Ganesh .S B.E. , Dr. T. Vijay Anand M.E. , Ph.D. Mr.K.
Paper Contents
Abstract
The government has stated that SMART cities are crucial for the prosperity of our country. Cities that are SMART change social norms. cleanliness, traffic control, surveillance, monitoring and identifying human activity, contemporary infrastructure, and upgrading public amenities. CCTV camera installation in homes and workspaces is now typical in order to monitor human activity. Camera recordings needed to be stored on hard drives, and viewing the pictures was necessary to detect movement. Overall, it is an inactive strategy. Event detections are found after the event has already happened and had an impact. Event detection is essential for understanding the situation dynamically and taking effective action. These security cams cannot be consistently watched over by humans. It requires the workforce and their constant attention to determine whether the documented actions are anomalous or suspicious. As a result, this flaw is fueling demand for this operation's highly precise automation. Additionally, in order to determine whether the strange behavior is suspicious or atypical more quickly, it is essential to pinpoint which frames and segments of the recording contain the odd activity. The suggested project uses 80:20 of the specified input data for training and testing, respectively. The test findings are as follows: Validation precision of 99.96% and Validation cost of 2.7%.
Copyright
Copyright © 2023 Mr. Arun Prasath.S B.E., , Alex .J B.E.,, Guru Ganesh .S B.E., Dr. T. Vijay Anand M.E., Ph.D. Mr.K. SUBRAMANIAN. M.E.,. This is an open access article distributed under the Creative Commons Attribution License.