Abdullah Erzin - Interactive Portfolio

About Me

I'm Abdullah Erzin,

A passionate engineer with a strong background in computer vision, AI, and robotics. With a degree in Mechatronics Engineering from Marmara University, I've developed and implemented innovative solutions across various fields, from autonomous navigation systems to advanced AI models. My career journey spans roles as a Software Engineer, AI Developer, and Autonomous Vehicle Engineer, showcasing my ability to tackle complex challenges and deliver impactful results. Highlights of my work include developing a cleaning robot's autonomous movement system, achieving high-accuracy face recognition algorithms, AI supported surveillance systems. I excel in using cutting-edge technologies like TensorFlow, PyTorch, OpenCV, and NumPy, and I'm skilled in programming languages such as Python and C++. Beyond engineering, I enjoy exploring the open roads on my motorcycle and diving into the underwater world as a certified scuba diver. My journey reflects a relentless drive to innovate and push the boundaries of technology while creating practical, efficient solutions.

Skills

Programming Languages

Python, Halcon, C++, C#, JavaScript, HTML, CSS, SQL

Computer Vision & AI Frameworks

TensorFlow, PyTorch, OpenCV, Scikit-learn, NumPy

Specialized Skills

Stereo Camera Calibration, YOLO Object Detection, Depth Mapping & 3D Vision, SLAM & Autonomous Navigation, Gradient Boosting Models, Custom ML Pipelines

Hardware & Embedded Systems

Jetson (Nano/Xavier/Orin), UART Communication Protocols, Sensor Fusion (Stereo Cameras, Gyroscopes, Ultrasonic), Thermal Cameras

Development Tools & Platforms

Git, SQL/MSSQL, Desktop Application Development, Mobile App Integration

Soft Skills

Leadership, Cross-functional Collaboration, Problem Solving, Algorithm Optimization, Ability to Fight with Jetson Devices and survive

Experience

Computer Vision Engineer

Pek Automotive (Ljubljana, Slovenia) | Apr 2025 - Present

Developed an advanced stereo camera calibration system for agricultural robots, reducing calibration time by 60% while achieving industry-leading precision of ±5mm deviation, significantly improving operational efficiency and robot accuracy in field conditions.
Improved a YOLO neural network's detection accuracy by developing a secondary ML pipeline. This pipeline uses a gradient boosting model with depth data to correct false positives in bundled apples and detect partially visible apples.
Engineered mathematical transformation algorithms to resolve spatial misalignment between camera 3D coordinates and robotic arm positioning, enabling precise computer vision-guided targeting and eliminating systematic positioning errors in agricultural applications.
Enhanced object detection capabilities by integrating depth analysis with YOLO models through custom linear regression algorithms, improving apple detection accuracy in real-world field conditions and implementing adaptive camera control systems that optimize depth mapping performance under varying sunlight conditions.

Software Engineer

Robsys Robotic Systems | Mar 2024 - Mar 2025

Developed an autonomous moving system from scratch for a cleaning robot, including mapping and navigation with arrays, and managed robot communication independently, ensuring no external dependencies.
Integrated real-time data and camera streaming into a mobile app for remote monitoring and control.
Designed a custom communication protocol via UART for seamless data exchange with the robot's motherboard.
Leveraged a Jetson device with AI models for real-time data acquisition and precise control through direct communication with the robot's motherboard.
Collaborated with cross-functional teams to develop a robust communication system that streams live data and camera feeds to a mobile app, enhancing user experience by providing real-time monitoring and remote control capabilities

AI Developer

Bomensoft (Remote) | Jan 2024 - Feb 2024

Collaborated with team members to optimize AI models and add new features to backend significantly enhancing efficiency and performance
Revamped the website's log structure using JavaScript, HTML, and Python, improving data connectivity and analysis.
Expanded website functionality by developing new pages and implementing additional features.

Computer Vision Engineer

DGR Project | May 2023 - Dec 2023

Developed a high-security face recognition algorithm achieving 98% accuracy and optimizing memory usage to 4kb for per person using image pre-processing.
Led the development of a live streaming project, reducing processing time to under 30ms.
Aligned package dependencies and optimized AI models for Jetson platforms.
Spearheaded the decelopment of an innovative live streaming project, aiming to reduce processing time to under 30ms, managing both software and computer vision model development

Autonomous Vehicle Development Engineer

ZGN Autonomous & Robotics | Dec 2022 - May 2023

Optimized pathfinding algorithms, reducing processing time by 30% and increasing efficiency.
Designed and implemented algorithms enabling robots to map areas using construction blueprints.
Contributed to autonomous driving algorithm development projects

Machine Vision Engineer

Gensys Automation & Machine Vision | Dec 2021 - Dec 2022

Developed desktop applications using C#, Halcon, and MSSQL for seamless machine vision program functionality.
Conducted on-site project setups and delivered two TUBITAK-funded projects.
Developed NDT (Non-Destructive Testing) program via thermal camera and halogen lamps to detect cracks inside of the materials

About Me

Skills

Programming Languages

Computer Vision & AI Frameworks

Specialized Skills

Hardware & Embedded Systems

Development Tools & Platforms

Soft Skills

Experience

Computer Vision Engineer

Software Engineer

AI Developer

Computer Vision Engineer

Autonomous Vehicle Development Engineer

Machine Vision Engineer

Contact

Details