NCTU

Idendity Authentication Using Gesture Signature Recognition in 3D Environments

2017/08/04
CodeM15-KCF-10-1
Professorprofessor Kuo-Chin Fan
ApplicationMedicine and health care and Surveillance and security
FunctionImage detection, Image recognition, Information of depth sensor, Object/human tracking, and Stereo vision
Technical BenefitEasy operation/ reduce participation of operators, Improve system integration, and Improve system intelligence
Technology StatusCan be transferred

The using of gestures as the communication media for Human Computer Interaction(HCI) is gradually realized in high-end 3C commercial products. As to the using of 3D gesture signature as the identity authentication scheme, it is still under the preliminary stage world widely. Especially for 3D Chinese gesture recognition, it is admitted as a much more difficult problem than the gesture recognition of other languages because of the uniqueness and complexity of Chinese characters. In this project, we propose a 3D gesture signing verification system. The 3D gesture signing verification system relies mainly on traditional CCD cameras with an eye to verifying the identity of gesture signer in 3D environments. The main modules of this project include gesture detection, real and virtual stroke extraction of 3D gesture, and feature space transformation for 3D gesture.

3D gesture signature procedure.

Spec. request

  1. platform: Window 7
  2. tools: C/C++
  3. hardware specification: PC Core i7, LeapMotion

 

demand

  1. Intel Core 2 Q9400以上

Ram: at least 2GB