Some Featured Projects
ROBOTICS:
Yıldız Rover
My team and I spent two years building a fully-functional rover, which we then participated into various contests, such as the ERC organized by the European Space Agency, the URC organized by The Mars Society, Teknofest - Turkey's biggest competition, and the ARC organized by UKET. We were thrilled to achieve impressive ranks in each of these competitions.
Air Conditioner
I made myself an air conditioner connected to the peltier system. I designed and assembled everything, it took me some time. But it works very well!!
IMAGE PROCESSING:
Extraction Automation
This project primarily involves extracting questions, answers, topic names, and test names. It then converts each question into a PDF file and an SVG image. The details of the question are incorporated into the file names. The purpose of developing this project is to build a question pool for an online study platform.
Aruco Detection with ZED Camera
As a component of the Yıldız Rover software, my friend and I have developed this project. Among the team's codes, this is the only portion that can be shared. The code is responsible for instructing the rover to autonomously navigate towards or pass between ArUco markers, contingent upon the count of detected markers.
WEB:
Camp System Website
In order to gain practical experience building web applications from scratch, this project was developed. The task is to develop a web application that simulates a camp database system over large area. Camps, reservations, accomadations, employee and customer, services, activities, foods information is stored and manipulated, accessed via UI. Collaborated with 2 friend.
ARTIFICIAL INTELLIGENCE:
Player Price Prediction and Web Scrapping
This project involves web scraping Super League football player data from the Mackolik website. We're extracting player statistics for consecutive two years to analyze market value changes. We aim to use this data to predict future market trends based on performance using supervised learning techniques like Naïve Bayes, Decision Trees, Random Forest, K-Nearest Neighbors, and Gradient Boosting. Collaborated with a friend.
Sensor Prediction
Using a comprehensive dataset of sequential sensor observations and room occupancy data in a hotel, this study preprocesses and structures the data before employing predictive models like K-Nearest Neighbors, Logistic Regression, and Decision Trees. The analysis incorporates data visualization techniques at each step to enhance transparency, offering a systematic approach to understanding sensor data in the hospitality sector. Collaborated with a friend.
Camera Height Estimation
This project is dedicated to developing a model for estimating camera height, building upon the "QR and Symbol Detection + Image Analysis" project. The main goal is to investigate the connection between six vectors originating from a tube's top circle center to its bottom circle center in an image and the camera-determined height.
Genetic Algorithm Example
This project uses a genetic algorithm to solve binary classification using a dataset. It identifies a set of N keywords for accurate sentence labeling. After preprocessing, a chromosome population is generated and evaluated for fitness. Selection, crossover, and mutation processes aim to find optimal solutions. The code provides fitness plots and parameter customization, including dataset choice, population size, gene count, and mutation probability. Collaborated with a friend.
ALGORITHM:
Puzzle Game
I visited an illusion museum where I encountered a challenging puzzle. Despite being unable to solve it on-site, the puzzle stuck in my mind. Determined to crack it, I developed a game as a way to tackle the problem. After some time and effort, I successfully managed to solve it. I recommend you to play :)
Arduino and Proteus Exchange Return Machine
Developed within the Arduino environment, the electronic circuit and coded system establish stable operation through serial communication with the user. Input values provided by the user drive the progression within the code, enabling the implementation of an algorithm for dispensing the minimum required number of banknotes. The system also includes a component that uses green or red lights to indicate the success of the operation to the user. Cooperated with a friend.
Phone Contacts Program Console
This Phone Contacts application is written in C and utilizes structs and file processing. It offers a list view of contacts in alphabetical order, allowing users to add, delete, edit, search, and display contacts. Contacts include full name, email, and phone number. The app confirms quitting and automatically saves and loads contacts to/from a file database.