Recognition and tracking of features in the image (internship project)
Company: Courage + Khazaka Electronics

Period: August - Oktober 2023 (3 Monate)

Tasks completed:

* Sensor-based mapping and localization;

* Mean Gaussian filtering and median filtering;

* OpenCV image reads and displays pixel operation;

* Image restoration, brightness enhancement, skin whitening;

* Image: zoom, cut, shift, mirroring, affine map and rotation;

* object recognition and calculation of the image coordinate of each point;

Focus:

Implement an image orientation invariant algorithm using affine transformation

Detect and track objects using imaging algorithms. These points will later be used for camera calibration.

Automatic camera calibration and 3D pixel position determination. (bachelor thesis project)
Company: Courage + Khazaka Electronic

Period: Jan 2024 - Juni 2024 (6 Months)

Tasks completed:

* Development of an automatic calibration algorithms in csharp

* Creation of a calibration process with one image and with multiple image

* Implementation of a method to optimize the calibration parameter and reduce the projection error

* Estimation of the best suitable 3D Position of image points using triangualtion

Focus:

FreeCad, Math.NET, MatLab, CSharp, User graphical interface, OpenCV, Kalman filter, KNN, CNN & SVM recognition.

Image Processing Project - Implementation of 2D-FT and High-pass Filtering in C++
Company: University project in imaging

Period: Aug - Nov, 2023 (4 Months)

Project Overview:

This project involves the implementation of an image processing application in C++ that performs 2D Fourier Transformation (2D-FT) and high-pass filtering. The goal is to develop an algorithm that enables efficient and precise image processing. The implementation was carried out in both C++ and MATLAB to ensure a thorough examination of the steps and to ensure that the implementation does not deviate from the given structure.

Tasks Completed :

Implementation of five class methods in C++:

* Setting the Maximum k-values: The method void set_kMax(); sets the maximum k-values for the Fourier transformation based on the given parameters.

* Performing 2D Fourier Transformation of an Image: The method void FTofImage(Image img); performs a 2D Fourier transformation on a given image.

* Filtering with a High-pass Filter: The method void FilterWith_NormK_Filter(void); applies high-pass filtering based on the normalized k-values.

* Inverse Fourier Transformation: The method Image IFT(void); performs an inverse Fourier transformation and returns the result as an image.

* Setting the k-Space Distance: The method void set_Delta_k(double FoVx, double FoVy); sets the k-space distance based on the given parameters.

Focus of the Work :

*The implementation of the image processing algorithms focused on efficiently implementing 2D-FT and high-pass filtering in C++. MATLAB was additionally integrated to verify the implementation steps

Tools :

* C++, Matlab ;

Key Skills :

* Image Processing ;

* Implementation of Fourier Transformation ;

* High-pass Filtering ;

* Integration of MATLAB in C++ projects ;

* Understanding of 2D image processing algorithms ;

Result :

The generated image matrices and the matrices with the real and imaginary parts of the Fourier-transformed original image were read in MATLAB and graphically displayed using imagesc(). It was verified that the implemented filter led to a de-blurring of the original image. The k-space matrix was analyzed to ensure it looks sensible and the implementation is correct.

Note :

The k-space matrix represents the spatial frequency domain in Fourier space. In medical imaging, for instance, k-space is utilized to represent spatial frequency information in MRI images. Analyzing the k-space matrix helps ensure that the Fourier transform and filtering have been implemented correctly.

Reconstruction of a CT scan / Implementation of the Radon transformation in C++
Company: University project in imaging

Period: März - Mai, 2023 (3 Months)

Basic Flow of Radon Transformation :

The task involved implementing the Radon Transformation in C++ for a given image object. The attenuation coefficients in the image were considered as image points. The Field of View (FoV) of this image is 220 × 220 mm². The Radon transformation to be implemented provides the corresponding Sinogram, which a CT scanner would generate as the measurement result.

Tasks Completed :

* Definition of classes for the Image and the Sinogram;

* Loading the image from the file Object.txt;

* Computing the Sinogram;

* Saving the generated Sinogram to a text file;

* Implementing the algorithm to ensure that each pixel of the image contributes to the sum exactly once along a ray;

Matlab Integration :

* Testing the algorithm in Matlab before implementation in C++;

* Reading the generated Sinogram matrix and the original image into Matlab;

* Graphical representation of the Sinogram using imagesc();

* Generating an image from the Sinogram using the iradon() function.

* Comparing the original image with the image generated from the Sinogram using iradon();

* Discussing possible differences;

Focus of the Work :

*Developing a C++ application to perform the Radon transformation of a given object;

* Reading the generated Sinogram matrix and the original image into Matlab;

Tools :

* Visual Studio, Matlab, Text File I/O (fstream, ifstream, ofstream) ;

Result :

The implementation of the Radon transformation in C++ and the integration with Matlab were successfully completed. The focus was on the correct implementation of the algorithm to ensure accurate Radon transformation. Through Matlab integration, the algorithm was validated, and the results were presented graphically and analyzed.

CSharp (C#) Programming Training
Company: Udemy

Period: Feb 2023 - März (1,5 months)

Overview:

I have successfully completed a C# Programming course on Udemy. The training covered various aspects of C# programming, from basics like data types and variables to more advanced concepts like Unity Game Development and WPF applications. Below you will find the completed tasks and the focus of my work during the training.

Completed Tasks and Focus of Work:

1. Basic C# Programming:

   Developing a solid foundation for C# programming through practical applications of data types and variables.

   Practical Application: Knowledge of data types and variables is crucial for developing software in medical technology, such as processing patient data.

2. Efficient Code Writing:

   Efficient use of functions/methods to solve practical problems and optimize code flow.

   Practical Application: Efficient code is essential for implementing computation-intensive medical algorithms, such as in image processing systems for diagnosis.

3. Decision Making in Programming:

   Learning decision-making concepts in programming to determine logical paths within the code.

   Practical Application: Decision structures like IF-ELSE and SWITCH are used in medical technology to make adjustments to medical devices based on various patient scenarios.

4. Task Automation:

   Automating tasks and repetitions through a comprehensive understanding of loop constructs in C#.

   Practical Application: Loops are used in medical technology to automate repetitive processes, such as analyzing large data sets from medical sensors.

5. Object-Oriented Programming (OOP):

   Creating robust and maintainable applications through a deep understanding of OOP principles in C#.

   Practical Application: Object-oriented programming is used in medical technology to develop modular and maintenance-friendly software for medical devices, such as for imaging procedures or patient monitoring systems.

6. Advanced C# Topics, Collections, Debugging, Inheritance, Polymorphism, Text Files, Events, Delegates, LINQ, Threads, Unit Testing, WPF, Unity Basics, and Unity Game Development:

   Practical Application: The advanced C# topics as well as the knowledge in collections, debugging, inheritance, polymorphism, text files, events, delegates, LINQ, threads, unit testing, WPF, and Unity basics and game development find application in medical technology to:

   • Develop effective and reliable software for medical devices
   • Support medical imaging and diagnosis
   • Development of highly flexible and scalable applications through in-depth knowledge of knowledge of polymorphism and the processing of text files in C#.
   • Manage and analyze patient data
   • Improve the performance and accuracy of medical systems

Focus:

C# Programming, Medical Technology, Software Development.

Supervision of student tutorial groups in Mathematics I and III at the University of Koblenz (Academic staff)
Company: Hochschule Koblenz

Period: March 2022 - Feb 2023 (12 Months)

Skills acquired and tasks completed:

* Communication skills: Interacting with students in the tutorial group develops the ability to explain complex mathematical concepts in an understandable way and to respond to questions from participants.

* Teamwork: Working as a tutor often requires collaboration with other tutors or the main lecturer to ensure effective teaching and maximise student learning.

* Empathy and sensitivity: By putting yourself in the student's shoes and considering their individual difficulties and needs, you can create a supportive learning environment and enhance the learning process.

* Deeper understanding of the subject: Regular exposure to the content of Maths 3 will deepen and consolidate students' own understanding of the topics.

* Problem solving skills:* Working with students to solve mathematical problems strengthens their ability to analyse complex problems and develop solution strategies.

* Didactic skills: Tutoring experience helps to develop effective teaching methods and to teach mathematical concepts in different ways to suit the different learning styles of student.

Focus:

Mathematical modeling, Statistical analysis, Algorithm development, Machine Learning.