Deep Learning Based Autonomous Gun with real time UAV detection and tracking capabilities

Project Title

Deep Learning Based Autonomous Gun with real time UAV detection and tracking capabilities

Project Area of Specialization Artificial IntelligenceProject Summary

The main aim of the proposed project is to design and implement a real-time drone detection and tracking algorithm which can be employed in an Autonomous Gun based Shooting system.

As we face a lot of war threats from our neighboring countries, which are usually accompanied by Drone/Missile Attacks. This Project is based on an Autonomous gun or we can say an Anti-missile system, which will have the basic competence and intelligence to detect and track drones only and to avoid false positives such as birds or any other moving objects. An on-board solution will be developed, and the proposed system will be implemented on the Jetson Nano with a capability to reach near to real-time implementation. Hardware will be designed using motors and basic circuitry to follow the targeted drone. We will develop our own custom UAV dataset to train the machine learning/ deep learning algorithms.

Project Objectives

• Design and implementation of a robust and effective drone detection and tracking algorithm based on machine learning/ deep learning.
• Comparison, study, and selection of the required electronic equipment including motors and motor drivers.
• Assembly of the shooting system with the power distribution board, motors, motor drivers & camera.
• Optimization of the drone detection/ tracking algorithm to achieve real-time implementation.
• Complete integration of the hardware & software portions for the development of the complete prototype.

Project Implementation Method

• Comprehensive literature review will be carried out.
• A custom database for UAVs will be developed to train the drone-detection and tracking algorithm.
• Machine learning/ deep learning-based drone detection technique selected after the initial literature review will be implemented using Pytorch and our custom dataset.
• A tracking algorithm for the estimation of UAV trajectory will be designed, developed, and implemented.
• The control algorithm for controlling the motors using the feedback system will be designed, developed, and implemented.
• A power board will be designed to power the motors and cameras.
• Next, the proposed algorithms will be optimized to decrease the overall computational cost.
• Final integration of software and hardware will be implemented.

Benefits of the Project

1. To detect and track UAVs in real-time.
2. This methodology can be the basis for an improved security system that can detect and track the enemies and destroy them automatically.
3. Can be improved to develop a spy or a criminal detection system respectively for crowded areas.
4. Similar methodology can be used to detect and track other objects such as humans and vehicles etc.
5. Keeping in mind the current scenario of Covid-19, this methodology can be used to maintain the distance between the people

Technical Details of Final Deliverable

A machine learning/ deep learning-based drone detection and tracking algorithm will be designed and implemented on Jetson Nano to detect and track an incoming drone.

Final Deliverable of the Project HW/SW integrated systemCore Industry SecurityOther Industries Education , Legal Core Technology Artificial Intelligence(AI)Other Technologies RoboticsSustainable Development Goals Industry, Innovation and Infrastructure, Partnerships to achieve the GoalRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	73000
Motors	Equipment	6	1000	6000
Camera	Equipment	2	5000	10000
Motor Drivers	Equipment	6	500	3000
Power Board	Equipment	2	4500	9000
Jetson Nano	Equipment	2	20000	40000
hardware integration	Miscellaneous	1	5000	5000