My name is Oluwaseun Ojeleye. I hold a Bachelor's degree in Computer Engineering (Turkey) along with three Master's degrees (Finland, France and Japan), all with a profound focus on the captivating realm of Imaging and Light in Extended Reality (IMLEX). I was born and raised in Lagos, Nigeria. I am very passionate about programming and engineering as a whole. My main research interests lie in the fields of data and image processing, deep learning, computer vision, VR, AR, XR, and embedded systems.
Data Processing, Image Processing, Machine Learning, Deep Learning, Computer Vision, VR, AR, Embedded Systems and IoT
C, C++, CUDA, Python, Javascript, Java, SQL, PHP, Haskell, Assembly Language, Arduino, VHDL, Shell Scripting
Linux, GitHub, Sublime Text, QT5, Okular, Opera
This project involves spectral and color reconstruction of Hyperspectral Images based on Principal Component Analysis (PCA).
App
Code
PDF
This project involves the development of a desktop application designed for camera calibration using rectangular markers for augmented reality (AR) applications.
This software implementation leverages advanced image processing and computer vision algorithms to accurately compute both the intrinsic camera parameters (projection matrix) and extrinsic parameters (rotation and translation) required for precise rendering of virtual objects within a real-world environment.
The implementation is based on a straightforward approach, which involves the following steps:
• Acquisition of video frames from the camera.
• Detection of a marker in the image using advanced computer vision techniques.
• Computation of the homography matrix using the points of the detected square and a template image.
• Decomposition of the homography matrix to obtain the extrinsic parameters, including the rotation matrix and translation vector.
• Rendering of virtual objects in the world coordinate system using the computed parameters.
Demo Video
Code
This project is a comprehensive image processing tool that offers parallelization techniques for both the CPU and GPU to optimize image processing tasks. It is composed of four sub-projects:
• Shader (GLSL and WebGL-based image processing).
• Multithreading using OpenMP.
• CUDA-based image processing.
• CUDA with shared memory for improved performance.
Each sub-project offers a diverse range of options for image processing tasks that demand high performance and scalability. With its focus on parallel computing, this project is an indispensable tool for any image processing project that requires quick and efficient processing.
App
Shader based Code
This project involves the development of a ROS based teleoperation Robot that models indoor evironment in virtual reality space. Microcontrollers used: Raspberry Pi 3, OpenCR and Intel Joule 570X. Sensors and actuators such as Lidar and Dynamixel motors were interfaced with the microcontrollers.
Demo Video
Code
This project involves the implementation of a desktop‐based Multilayer Perceptron simulation and visualization tool. It was developed using Visual C++ and it can be used for learning and understanding Neural Network concepts.
Code
This project consist of the implementation of objects detection and classification program which detects and classifies objects(such as rice, beans, etc.) based on their geometric shape. It includes the implementation of machine and image processing techniques and methods such as Noise Filter, Image Segmentation, Morphological Image Processing, Labeling, Bounding, Feature Extraction, Feature Analysis and Supervised Learning.
PDF
Code
This is the development of a graphical image labeling tool that allows users to create, import, view, correct and generate annotations based on formats such as YOLO Format
Demo Video
Code
This project consists of my codes for hyperspectral image processing and analysis. I have used different hyperspectral camera for taking spectral images of different materials and objects for various analysis such as analysing paintings'
blue dyes, etc. These cameras include:
• Specim Scanner • Specim IQ
• Nuance • Tunable Light Source
The most common operations I performed on the gathered hyperspectral images during these analysis are:
• White & Dark Correction • Segmentation
Code
The objective of this project is to learn how to analyse and visualize eye-tracking data. The dataset consists of rows (one row per one session and one subject) of raw gaze locations (x, y) recorded in time. Subjects have multiple rows/samples
in the dataset. A fixation algorithm was used to get a better understanding of what the participants were fixating or focusing on while partaking in the experiment.
Code
This project consists of the implementation of a real-time face detector and gender classifier using an optimized Darknet Library. This optimized version of the darknet library was created for inference only. For accurate detection of faces closer or farther from the camera, the model was created by applying transfer learning using YOLOv3 and training on a modified FDDB dataset.
Code
This project consist of the implementation of lines and circles detection program using Hough Transform. This project is the extension of the former project's (Object detecting and classifying application) components. Processes such as Smoothing, Edge detection, Non-maximum Suppression, Hysteresis thresholding, etc. are implemented in this project.
PDF
Code
In this project, I wrote a simple neural network library that implements a multilayer perceptron. This library was written in C++ and the implementation was tested by building a simple Neural Network Model to solve the XOR problem.
Code
This is a Database Management Project that involves the development of a social networking web application which graduates, alumni, and professors of a university’s department can use to connect and share information such as job, internship and other opportunities among one another from every corner of the world.
PDF
This project involves the development of a bot for Hex board Game.
Completed Tasks: