Loading...

Design And Implementation Of A Hand Gesture Recognition System

Tavakol Elahy, Maryam | 2016

746 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: English
  3. Document No: 49580 (55)
  4. University: Sharif University of Technology, International Campus, Kish Island
  5. Department: Science and Engineering
  6. Advisor(s): Babaie Zadeh, Masoud
  7. Abstract:
  8. This thesis discusses a real-time vision-based framework for the purpose of hand region detection and hand gesture recognition. Our proposed methods include detecting hand regions in the cluttered background, based on Viola-Jones object detection algorithm and improving the classification of detected hand gestures regions in a novel contour-based framework. Our studies have demonstrated that deformability and high degree of freedom (DoF) of human hand as a non-rigid object besides diversity of skin color types, undeniable effect of cluttered background complexity, scalability and being robustness against rotation are the main reasons for considering some simplifications in visionbased methods that makes it more achievable for researchers to detect hand region and recognize different hand gestures. To the best of our knowledge, despite the fact that simplifications provide more accurate results in even shorter time, they usually prevent researchers from proposing an applicable hand gesture recognition framework for real environments. So we have put our goal to meet some drawbacks of previous works by understanding and tackling the aforementioned challenges without lowering the quality of the system as a real environment application which remarkably increases the applicability of our hand gesture recognition method. Our hand region detection proposed framework defines a new and appropriate enclosing area for extracting Haar-like features from hand region as our object of interest using Viola-Jones object detection method. Our hand ges ture recognition proposed method then provides a framework to create a contour model of hand and define four substantial descriptors according to the structure of human hand besides being equipped with the knowledge of distinguishability level of each feature and utilizing four classifiers to find appropriate rules among feature values of each hand gesture. Experimental results show that our proposed frameworks outperform many other hand region detection and hand gesture recognition methods. We have evaluated our results so that more than 99 percent accuracy in detection and more than 90 percent accuracy in recognition phase assure us of the possibility of achieving hand region detection and hand gesture recognition goals without any environmental simplifications. Keywords— real-time hand region detection, hand gesture recognition, human computer interaction, Viola-Jones object detection, hand gesture classification, Adaboost classifier, Haar-like features
  9. Keywords:
  10. Human Computer Interaction (HCI) ; Sign Language Recognition ; Hand Gestures ; Adaboost Algorithm ; Real-Time Hand Region Detection ; Viola-Jones Object Detection ; Haar-like Features

 Digital Object List

 Bookmark