Helmet Detector for the Automated Teller Machine

This project investigates the development of a helmet detection system for Automated Teller

Machines (ATMs) using OpenCV (Open Source Computer Vision Library) and Rasberry –pi

3 B+module.

Raberry pi 3B+

Pi camera

Power supply

SD card

Electric door lock

LCD display

Image acquisition with Rasberry pi camera

Utilizing a Rasberry pi camera to capture live feed images for real time analysis & processing.

Preprocessing with openCV

Employing openCV for image enhancement , noise reduction & feature extraction to prepare

imagesfor accurate detection.

XML haarcascade models

Implementing XMLmodels for deep learning based helmet detection, leveraging neural networks

for precise identification.

Alert generation based on detection outcomes

Generating immediate alerts upon detecting the presence or absence of helmets in the monitored

area to ensure security compliance.

cv2: For image processing, capturing frames from the camera, and detecting objects (helmets) using a Haar Cascade classifier.

RPi.GPIO: For controlling hardware components (like the solenoid lock and helmet signal) connected to the Raspberry Pi’s GPIO pins.

time: For tracking the timing of events (like helmet detection duration) and creating pauses (delays) in the program.

smbus: For sending commands and data to the I2C-based LCD display via the I2C bus.

This project makes a significant contribution to the discourse on the use of cutting-edge technologies to address contemporary security challenges in public spaces. The helmet recognition system shows the potential to use computer vision and machine learning to improve security measures at ATMs.

This project demonstrates the feasibility of developing a real time helmet detection system using OpenCV. Documented development, implementation, and evaluation of the system provides valuable insight into its practical use and effectiveness. The system offers a cost effective and practical solution to enhance ATM security.