Translator Disclaimer
9 September 2015 Illumination-invariant hand gesture recognition
Author Affiliations +
In recent years, human-computer interaction (HCI) has received a lot of interest in industry and science because it provides new ways to interact with modern devices through voice, body, and facial/hand gestures. The application range of the HCI is from easy control of home appliances to entertainment. Hand gesture recognition is a particularly interesting problem because the shape and movement of hands usually are complex and flexible to be able to codify many different signs. In this work we propose a three step algorithm: first, detection of hands in the current frame is carried out; second, hand tracking across the video sequence is performed; finally, robust recognition of gestures across subsequent frames is made. Recognition rate highly depends on non-uniform illumination of the scene and occlusion of hands. In order to overcome these issues we use two Microsoft Kinect devices utilizing combined information from RGB and infrared sensors. The algorithm performance is tested in terms of recognition rate and processing time.
© (2015) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
América Ivone Mendoza-Morales, Daniel Miramontes-Jaramillo, and Vitaly Kober "Illumination-invariant hand gesture recognition", Proc. SPIE 9598, Optics and Photonics for Information Processing IX, 95980W (9 September 2015);

Back to Top