Gesture Recognition using a sensor fusion of Quaternions and Image Processing
Abstract
The vital objective of this work is to bridge the gap between the physical world and the digital world using an efficient fusion of the image sensors. A low cost Inertial Measurement Unit (IMU) to form a gesture based remote controller which can be interacted with any electronic display devices. In contrast to the conventional approach of predefined robotic gesture recognition, our method is more efficient in detecting and recognizing more natural human gestures. With reference to the previous works being carried out in the similar field, approaches such as either image based or motion sensor based techniques have been used for the same. Our research primarily focuses on the convoluted sensor fusion which bridges the gap of effort less interaction between digital world and the real world through simple gestures. The proposed approach can identify more natural gestures with less software complexity and hence is more fast and accurate.