Suspect detection and tracking through image processing

Project Title

Project Area of Specialization

Artificial Intelligence

Project Summary

Security systems at entrance of majority of offices, building, institutes etc. are manual and guards are deployed to check the identification of visiting people. Sometimes it gets hard for them to check the identities of all in rush hours. So, the chances of penetration by unauthorized people is high during these hours. Moreover, penetration using other means is also possible. To overcome these security lapses, a system is required which will help the security staff to identify and capture the intruder.

It is a real-time system which can identify the people in the pictures taken by cameras, having the ability to track and report such persons or suspects to security staff using android app.

Project Objectives

Main objectives of our project are:

Face Detection
Face Recognition
Identification of suspect
Tracking the suspect
Real-time streaming of suspect on web app.
Maintaining database.
Identification of weapons.

Project Implementation Method

The area will be continuously monitored through cameras
Database will be maintained.
If the picture of person doesn’t match with our database or weapon is detected, then the person will be considered as suspect.
The cameras will track the suspect and his/her movement will be reported.
When the range of one camera ends it will signal the next one to track the suspect and this process continues.
We also design and implement android app or web app for live streaming of suspect.
All activities of suspect will be monitored from main central office.

We use different algoritthms for face detection,recognition and weapon detection and if person is considered as suspect then we track this person using cameras and end user will watch all these things on LED through web app.