University Surveillance System

Project Title

University Surveillance System

Project Area of Specialization Artificial IntelligenceProject Summary

Our Surveillance system will monitor and detect any sort of anomaly, such as, Smoking, Eating/Drinking in class, Gun/Knife/fork detection, Face detection (feature based), ID card detection, fight/scuffle detection (in cafeteria only). The system will also give the time-in and time-out of students upon entering and exiting the university. The idea is to replace the university’s manual procedure of supervision into an automated surveillance system that will provide an extra edge to security operations.

Project Objectives

Intruder Detection . Guns/ Knives Detection. Mask Detection. Face Detection by guard’s camera. All detections will be reported to the Admin.

• Intruder Detection .

• Guns/ Knives Detection.

• Mask Detection.

• Face Detection by guard’s camera.

• All detections will be reported to the Admin.

Project Implementation Method

Python, numpy, opencv, tensorflow, keras, dlib, multi ip-camera, firebase, flask, Angular

Python, numpy, opencv, tensorflow, keras, dlib, multi ip-camera, firebase, flask, Angular

Benefits of the Project

• Student’s IN and OUT time would be recorded in the system.

• If any student is found smoking inside the campus the system would detect the activity through camera feed and an alert would be generated on the system with a still picture of the exact time for the security officers to investigate.

• The system would be able to detect eating or drinking activity in a class (containing 40 students at a time) through the camera feed. The system would detect any activity and an alert would be generated on the system with still pictures of the exact moment.

• The system would be able to detect Gun, knife or fork within the campus through the camera feed. As soon as a detection takes place an emergency alert would be generated and sent to the admin and a quick notification alert would be sent to mobiles of the security officers. A still picture of the detection would be saved on the system as soon as a weapon is detected.

• The system would be able to detect fights or quarrels inside the cafeteria through input from the camera feed. An alert would be generated and sent to the system.

• All these detection pictures and videos will be stored on local storage so that they can be viewed later on.

• System admin would be able to see live camera feed and can also view live detections.

Technical Details of Final Deliverable

Train Model on given features(university student dataset). Model Testing. Camera and other hardware setup. Connecting camera to the web app Communication between camera stream and our model.

For FYP 2:

Show detection results in the app. Testing the security system.

• Train Model on given features(university student dataset).

• Model Testing.

• Camera and other hardware setup.

• Connecting camera to the web app

• Communication between camera stream and our model.

• Show detection results in the app.

• Testing the security system.

Final Deliverable of the Project Software SystemCore Industry ITOther IndustriesCore Technology Artificial Intelligence(AI)Other TechnologiesSustainable Development Goals Industry, Innovation and InfrastructureRequired Resources

Python, numpy, opencv, tensorflow, keras, dlib, multi ip-camera, firebase, flask, Angular