AI-DRIVEN AUTONOMOUS CRISIS REPORTING SYSTEM BASED ON ANOMALY DETECTION USING VIDEO SURVEILLANCE

The project is aimed towards designing and implementing a real-time anomly detection system. This system will be able to analyze large amount of data in the form of surveillance videos generated continuously by CCTV systems. Such a system will find great utility in the fields of security, safet

Project Title

AI-DRIVEN AUTONOMOUS CRISIS REPORTING SYSTEM BASED ON ANOMALY DETECTION USING VIDEO SURVEILLANCE

Project Area of Specialization

Artificial Intelligence

Project Summary

The project is aimed towards designing and implementing a real-time anomly detection system. This system will be able to analyze large amount of data in the form of surveillance videos generated continuously by CCTV systems. Such a system will find great utility in the fields of security, safety of personnel and infrastructure, military surveillance for hostile activities and helping in the prevention of any forced/stealthy intrusions etc. Moreover, timely reporting of incidents like explosions, fire and accidents will help in the provisions of appropriate aid to the concerned. These applications require the analysis of excessive amount of data which will consume a lot of man hours as it demands round-the-clock surveillance, which is tedious and might not be always possible. Additionally, there would be chances of error due to inefficient surveillance or failures to monitor due to human fatigue and negligence. The designed system will provide a feasible solution to all these problems.

Project Objectives

Crisis refers to a difficult or dangerous situation, or a state of emergency. Crisis reporting is critical for effective emergency response and issuing intime warning to the first responder. The project aims to implement a realtime and autonomous crisis reporting system using surveillance videos on a portable hardware with latest Deep Learning Algorithm.

Project Implementation Method

The research approach for conducting this project is to study the related literature of the project and its datasets for manifold anomalies. Moving forward on studying and understanding the concepts of Artificial Intelligence, specifically Deep Learning. Afterwards, stepping into the mechanisms of various available implementation platforms for training the model on YOLOv5 (You Only Look Once) by google. After that, the trained model will be implemented on a portable device Jetson AGX Xavier, and an automated alarm generation mechanism would be designed for real-time reporting. Finally, its accuracy and feasibility will be gauged.

Benefits of the Project

This research project conforms significantly with the PAF and National needs. This technology will be proven as a useful tool for surveillance due to its self-sufficiency. It is understandable that round-the-clock surveillance is a challenging task and carries the chances of error in case of any negligence which is unaffordable to an organization like PAF. This technology will aid in continuously tracking an area for any abnormal activity and in time reporting of a crisis. This project will assist in minimizing the number of man hours consumed and substitute them as an alternative for surveillance and target tracking, performing more competently in comparison.

Technical Details of Final Deliverable

• The Expected deliverables of the project are as follows • The data sets will be trained on YOLOv5, that will minimize the latency for real-time implementation.
•  The trained system will be implemented on Jetson AGX Xavier, that will enhance the overall performance of the model.
• Integrating emergency alarm over a network for real-time reporting

Final Deliverable of the Project

Software System

Core Industry

IT

Other Industries

Security

Core Technology

Artificial Intelligence(AI)

Other Technologies

Internet of Things (IoT)

Sustainable Development Goals

Industry, Innovation and Infrastructure

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com