ADVANCED GESTURE RECOGNITION USING MEMS MIRRORS FOR PRECISION CONTROL

Abstract

This thesis presents an advanced gesture recognition system utilizing Micro-Electro Mechanical Systems (MEMS) mirrors for precise control applications. The proposed system employs an OAK-D Pro camera integrated with computer vision techniques, real-time image processing algorithms, and MEMS scanning mirrors to accurately detect and track hand gestures. Through a wide-ranging methodology incorporating gesture definition, data collection, algorithm implementation, and thorough testing, we achieve a both high-reliability and responsive system. Experimental results suggest that our approach achieves a low mean tracking error of less than 4 mm with robust tracking low latency. The accuracy of system is evaluated on a systematic basis and it outperforms the available solutions. Applications are discussed, showcasing its tremendous potential in the medical, automotive and consumer electronics industries. Some technical hurdles including light dependent movement sensing based and system latency get identified and strategies are recommended that can advance positive experience. Finally, this work also discusses future research perspectives such as hard ware aspects and algorithmic improvements, highlighting the technology’s potential in terms of integration and social consequences.