Target Tracking on 2D Platform Based on Kalman Filter

MSc student: Bakir Brkic & Omar Hassan

Mentor: Assist. Prof. Dr. Tarik Namas

The project was intended to test the general properties of the Kalman Filter algorithm, how applicable and reliable it is in target tracking. For the purpose of the project, we developed complete hardware and software solutions consisting of cross-platform software used for generating the test data, collecting the real-world data, processing the acquired data (image recognition), predicting the target position (Kalman Filter script) and finally controlling the hardware components. The hardware that was used was based on a repurposed 2D laser engraver.

Objectives:

· Implement Kalman Filter function that predicts and corrects position data of a given target

· Represent the comparison between Kalman filter results and actual raw data

· Implement a cross-platform application that enables users to generate test data by determining a variety of factors, such as the trajectory form and the desired level of noise

· Implement a cross-platform application for the real-world data acquisition through Camera API

· Implement an image processing script for determining the position of targets in two-dimensional space

· Implement a software capable of hardware control

· Select and design a communication protocol that is going to unite all parts of the system