Robust hand gesture recognition based on finger-earth mover's distance with a commodity depth camera

2011 ◽  
Author(s):  
Zhou Ren ◽  
Junsong Yuan ◽  
Zhengyou Zhang
Keyword(s):  
Gesture Recognition ◽  
Hand Gesture Recognition ◽  
Hand Gesture ◽  
Depth Camera ◽  
Earth Mover’S Distance ◽  
Earth Mover's Distance

Robust Part-Based Hand Gesture Recognition Using Finger-Earth Mover’s Distance

2017 ◽  
Vol 7 (7) ◽  
pp. 131
Author(s):  
Ashwini Kolhe ◽  
R. R. Itkarkar ◽  
Anilkumar V. Nandani
Keyword(s):  
Gesture Recognition ◽  
Computer Games ◽  
Real Life ◽  
Hand Gesture Recognition ◽  
Hand Gesture ◽  
Earth Mover’S Distance ◽  
Language Recognition ◽  
Sign Language Recognition ◽  
Hand Gestures ◽  
Earth Mover's Distance

Hand gesture recognition is of great importance for human-computer interaction (HCI), because of its extensive applications in virtual reality, sign language recognition, and computer games. Despite lots of previous work, traditional vision-based hand gesture recognition methods are still far from satisfactory for real-life applications. Because of the nature of optical sensing, the quality of the captured images is sensitive to lighting conditions and cluttered backgrounds, thus optical sensor based methods are usually unable to detect and track the hands robustly, which largely affects the performance of hand gesture recognition. Compared to the entire human body, the hand is a smaller object with more complex articulations and more easily affected by segmentation errors. It is thus a very challenging problem to recognize hand gestures. This work focuses on building a robust part-based hand gesture recognition system. To handle the noisy hand shapes obtained from digital camera, we propose a novel distance metric, Finger-Earth Mover’s Distance (FEMD), to measure the dissimilarity between hand shapes. As it only matches the finger parts while not the whole hand, it can better distinguish the hand gestures of slight differences. The experiments demonstrate that proposed hand gesture recognition system’s mean accuracy is 80.4% which is measured on 6 gesture database.

Sign in / Sign up

Export Citation Format

Share Document