Motion Estimations of Hand Movement Based on a Leap Motion Controller

Multimedia technologies have increased the demand for interactive tools that offer touchless interaction. In response to this demand, leap motion controller (LMC) has emerged as a touchless motion-tracking device for precise hand movement analysis. This work introduces a novel method for hand movement estimation using an LMC. The data captured by LMC are processed and analyzed to generate a hand skeleton. In our approach, points of interest (POIs) are identified at each finger, and their optic flow is determined based on their spatial–temporal positions in the 2-D space. The method considers the feature extraction mechanism for the hand fingers’ phalanges and tips, where these features contribute to the motion estimation vector that enables hand motion tracking, in addition to enriching the feature recognition and identification of dynamic gestures since these vectors are calculated over time. We rigorously tested our model with several motion types, including translations, rotations, and complex movements. The results demonstrate exceptional performance with a slight error compared with synthetic optic flow. The average root mean square error (RMSE) in the spatial-temporal domain is ±4.74°, ±4.61°, ±5.07°, and ±3.94° for left, right, up, and down motions, respectively, thus highlighting the accuracy and reliability of our model in estimating optic flow and tracking hand motion.