Autonomous Maneuvering of UAV to Follow the Target Identified in Ground Camera

An autonomous aerial tracking system, designed to follow a target that is identified in the ground camera. The solution secures an outdoor territory, by identifying the suspicious person, and then using a UAV to follow the target, re-identified in aerial view. In the event of a terrorist att

Project Title

Autonomous Maneuvering of UAV to Follow the Target Identified in Ground Camera

Project Area of Specialization

Computer Science

Project Summary

An autonomous aerial tracking system, designed to follow a target that is identified in the ground camera. The solution secures an outdoor territory, by identifying the suspicious person, and then using a UAV to follow the target, re-identified in aerial view. In the event of a terrorist attack, the solution would ensure outdoor security by following the target autonomously, until the reinforcement arrives. The hardware-software integration enables to perform the three basic tasks; 

1. Manual identification of the target in ground camera

2.  Re-identification of the target in drone camera

3. Autonomous maneuvering of the drone to follow the target

Project Objectives

This project contains three objectives;

1. We aim to design a robust system capable of tracking down the trajectory of the suspect or target identified anomalous and following him. 

2. Aerial reidentification and ground reidentification are being studied as two separate domains for a decade, we aim to merge the two domains. Our objective is to train a model that is capable of reidentifying humans from aerial and ground perspective, and based on the re-identification it follows the target.

3. Maneuvering of UAV to follow the target without any collision and occlusion.

Project Implementation Method

The project is divided into three main modules;

1. Target detection: The target person is selected manually from the ground surveillance camera. The identified target is then sent to the Ground Control Station (GCS), which then sends the GPS coordinates to the UAV to arrive at the location. 

2. Re-identification of target in UAV: The UAV captures a full frame as it arrives at the location, and sends it to the GCS. GCS then applies human detection on the full frame and performs image matching to match each person with the target image.  

3. Autonomous maneuvering of UAV to follow target: The UAV starts following the target re-identified in aerial view autonomously, whilst performing collision detection.

Benefits of the Project

1. Our project can be deployed at any place such as institutions, bases, etc., where the goal is to maintain a secure territory. 

2. By using a drone, the target can be followed beyond the surveillance camera's range and the continuous movement of the target can be monitored until the reinforcement arrives.

Technical Details of Final Deliverable

The project is a software integrated hardware, comprising of three Major Modules as;

1. Human Detection

2. Camera Handshaking

3. Autonomous Maneuvering of UAV

Human Detection: Using Algorithms for human detection, human individuals in Surveillance footage of ground and aerial surveillance cameras will be enclosed in a bounding box to acquire ground perspective images.

Camera Handshaking: Ground perspective images after conversion in aerial perspective will be used in re-identification in aerial images. A solution will be proposed for the ground to aerial re- identification.

Autonomous Maneuvering: Once the target is reidentified, a programmable drone will autonomously maneuver to follow the trajectory of the target. After tracking the drone will fly back to its base station.

Final Deliverable of the Project

HW/SW integrated system

Core Industry

IT

Other Industries

Security

Core Technology

Artificial Intelligence(AI)

Other Technologies

Robotics

Sustainable Development Goals

Industry, Innovation and Infrastructure

Required Resources

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