Publication

Abstract : The significant issue of ATM (Automatic Teller Machine) fraud is getting worse every day. Security precautions are being developed to protect ATMs from these scams. Even though the bank utilized new and improved ATM security measures, the rate of robberies is increasing daily. Automated teller machine (ATM) security research looks for ways to secure ATM installations and provide several points of defense against physical and technical theft. It is crucial to use an automated monitoring system to safeguard ATMs from threats due to the rise in ATM robberies. Currently, a camera mounted on the ATM captures and transmits the video stream to the main server of the bank. This manual surveillance constantly consumes a large amount of bandwidth for transmission. Memory is being wasted, which delays emergency replies. As a result, it’s crucial to have early situational awarenessin order to stop a continuing burglary. The challenging task of combining image processing with video surveillance systems has been the subject of various attempts over the past few decades. Despite being vulnerable to criminal activities, automated teller machine (ATM) surveillance systems and image processing software have not yet been fully integrated. According to this study, it can tell if someone is acting strangely inside the ATM room. The suggested method can also count the number of people inside the ATM and determine whether they are carrying any weapons like guns. Based on the detection, our suggested system would send a warning message to the appropriate bank officials and surrounding police officials, removing the need for constant human monitoring, decreasing the storage of unneeded video feed, and communicating only an unusual condition, allowing for a faster response to danger by informing the respective officials.