A Survey on Crime Detection using CCTV Systems
A. Jeba Sheela, S. Balaji, B Balaji, U Hemanth Kumar · 2023
In order to improve security measures across numerous domains, such as public safety, transportation, and critical infrastructure protection, the use of closed-circuit television (CCTV) systems for threat analysis has become increasingly important. The technical abstract gives a summary of the most recent algorithms and techniques used in threat analysis when using CCTV. This study emphasizes the importance of spatial and temporal reasoning, machine learning models like Support Vector Machines (SVM), Haar-Cascade, ViF (Violent Flows), OViF (Oriented Violent Flows), RINOC (Real-time Image and Video Object Classification), and LRT (Live-Streaming Real-Time). Insights into the diverse methodologies that can be applied to enhance the effectiveness of CCTV-based threat analysis, the efficiency of CCTV systems and their impact on detecting and reducing criminal activities are discussed. The comparative outcomes of different detection approaches and their ensembles have been empirically discussed.