Babatunde Ogunfemi

I began my computer vision experience developing experimental chess game tracking software to help with research that a professor of a university was performing. This involved anazlying video footage in realtime with and without markers (we experiemented with different approaches including using AR markers, color markers and no markers at all), calibration and hand motion detection and tracking. This project was OpenCV + C++ Eventually in the future I moved to working on facial recognition software towards the OpenBR project, specifically a project with the intention of using specific facial "landmarks" to improve facial recognition technology. This project was OpenCV + Python. Most recently, I worked on a 3D real-time hazard detection system with the intent to help prevent falls by senior citizens. This system involved many sensors (including a Microsoft Kinect v2, ambient light sensor and topology sensor) to help detect 3D spaces, separate detected objects, filter out noise and extraneous objects (like non-hallway walls) and analyze the potential for hazard given input of a particular patients illnesses and disabilities. The architecture was largely designed by me (including the complete computer vision architecture), the sensor fusion was implemented by me, as well as the majority of the computer vision code (filters and algorithms). This involved OpenCV + Python + Point Cloud Library (PCL) + Java.