Conor Pickering

Designed and optimised algorithms for object detection and tracking on embedded camera systems using Python and C, enhancing real-time image processing performance and accuracy. Maintained and improved code for deploying machine learning models on embedded platforms. • Applied Agile principles (sprints, stand-ups, peer reviews) to deliver iterative improvements and maintain high development velocity. • Developed, tested, and deployed image processing and filtering algorithms for infrared sensors in C/C++ and Python, with comprehensive unit, integration, and system-level testing. • Achieved execution-time targets by analysing and optimising code down to the Assembly level, gaining deep expertise in memory management and storage optimisation. • Conducted experiments to derive new equations for luminance calculation from existing image datasets, enhancing accuracy of brightness analysis methods. • Debugged and extended legacy image analysis repositories, adding functionality for real-time video stream input from multiple sensors. • Integrated and debugged sensors and embedded boards, interpreting datasheets and resolving communication issues via UART and other low-level protocols. • Utilised Wireshark, serial monitors, and JTAG tools to analyse and debug device communication and embedded firmware behaviour. • Expanded algorithms to support a broader range of image formats, ensuring compatibility across multiple input devices. • Built a Python-based GUI application for a facility administrator to monitor and manage desk space utilisation in the office. • Resolved algorithmic issues causing video distortion, requiring in-depth understanding of memory allocation and buffer management in C. • Developed expertise in luminance and illuminance principles while designing methods to detect shadows and camera obstructions with high reliability.

Developing embedded systems using machine learning models