A design on recommendations of sensor development platforms with different sensor modalities for making gesture biometrics-based service applications of the specific group

Abstract

Gesture biometrics sensors will be extremely helpful for constructing a gesture-based application system. Compared with traditional acoustic sensor-based speech recognition applications and RGB image sensor-based surveillance applications, proper utilizations of the gesture biometrics sensor will absolutely be able to create an innovative and practical application and speed up the maturity of gesture recognition. Currently, three well-known gesture biometrics sensor platforms are Kinect, Leap Motion, and Myo. Kinect and Leap motion sensor devices belong to the categorization of 3D image sensors, containing both RGB and depth sensors. Myo armband is categorized into the type of wearable sensor devices. Nowadays, related studies on application system establishments using Kinect, Leap Motion, or Kinect have been frequently seen in the recent years. However, most of all these related gesture-based application systems are designed for end users. Extremely few investigations using these popular gesture biometrics sensors are done specifically for the system developer of a gesture-based application system. In fact, a correct and proper adoption of the gesture biometric sensor platform will speed up the development of the desired system and increase the practicality of the designed system. In this paper, a recommendation scheme of Kinect, Leap Motion, and Myo sensor platforms is presented for supporting developments of a mid-air gesture application. Compared with the conventional application system design procedure, the presented sensor recommendation scheme incorporated between two stages of functional operation and system design specifications will have a great help to the system developer to establish his desired system. In addition, a case study on system developments of efficient web-based streaming media control by disabled persons using the presented sensor platform recommendation scheme is also reported in this paper.