I= GESTURE RECOGNITION BASED INTELLIGENT ALGORITHMS FOR VIRTUAL KEYBOARD DEVELOPMENT
Pakistan Research Repository Home
 

Title of Thesis
GESTURE RECOGNITION BASED INTELLIGENT ALGORITHMS FOR VIRTUAL KEYBOARD DEVELOPMENT

Author(s)
Hafiz Adnan Habib
Institute/University/Department Details
Department of Electrical Engineering/ University of Engineering & Technology Taxila
Session
2007
Subject
Electrical Engineering
Number of Pages
77
Keywords (Extracted from title, table of contents and abstract of thesis)
gesture recognition, intelligent algorithms, virtual keyboard, finger-joint wearable sensor gloves, thumb code, accelerometer based inputs, laser projected keyboards, gyroscope based sensing, vision based virtual keyboard, fuzzy logic

Abstract
Virtual keyboard design is an important research & development area in human computer interaction. A keyboard in portable devices is generally inadequate for text entry because of small size and attachment of ordinary keyboard adds to the size. Need for virtual keyboard also exist in human machine interfaces to a wide range of applications in markets as consumer electronics, medical/healthcare, industrial automation, automotive, and public information kiosks.

Current virtual keyboards designs appear in various forms such as finger-joint wearable sensor gloves, thumb code, accelerometer based inputs, laser projected keyboards and gyroscope based sensing. Each virtual keyboard has certain design characteristics. However, performance parameters for the evaluation of keyboards are same such as number of discrete keys, response time and failure rate. Other parameters included can be the ability to remap, key symbol mapping and space requirements. The cost of most of the successful virtual keyboard designs is high due to custom design and expensive technology.

This research provides a new perspective to view the problem of virtual keyboard design by incorporating gesture recognition by using a simple mono vision camera. Human hand and finger movements are considered as gesture movements making keystrokes on any surface. Human hand and finger movements are recorded in video sequence and gesture recognition algorithms estimate the key pressed. This algorithm uses aggregation of gesture information using fuzzy logic and intelligent learning based approach. Feasibility of mono-vision gesture based virtual keyboard is demonstrated. The proposed system shows lesser requirements in hardware due to software centric design and off the shelf components.

Download Full Thesis
894.48 KB
S. No. Chapter Title of the Chapters Page Size (KB)
1 0 Contents
83.42 KB
2 1 Introduction 1
236.67 KB
  1.1 Problem Statement 1
  1.2 Need For Virtual Keyboard 1
  1.3 Alternative Approaches For Text Entry 3
  1.5 Characteristics And Metrics 7
  1.6 Virtual Keyboard Designs 9
  1.7 Vision Based Virtual Keyboard 13
  1.8 Social And Economic Issues With Virtual Keyboard 15
  1.9 Gesture Recognition Based Virtual Keyboard 15
  1.10 Summary 16
  1.11 References 17
3 2 Gestures Recognition 19
288 KB
  2.1 Introduction To Gesture Recognition 19
  2.2 Types Of Gestures 21
  2.3 Application Domains Of Gestures 22
  2.4 Architecture Of Gesture Recognition System 23
  2.5 Hand Gesture Recognition 39
  2.6 Summary 40
  2.7 References 40
4 3 Fuzzy Logic 44
114.59 KB
  3.1 Introduction 44
  3.2 Intersection, Union And Compliment 46
  3.3 T-Norm And T-Conorm : 46
  3.4 Fuzzy Relation 47
  3.5 Fuzzy Composition 48
  3.6 Fuzzy Inference Engine 48
  3.7 Fuzzy Measures 51
  3.8 Fuzzy Measures 51
  3.9 Fuzzy Inferencing 51
  3.10 Performance Metrics 52
  3.11 References 53
5 4 Gesture Recognition Based Virtual Keyboard System 54
151.84 KB
  4.1 Architecture 54
  4.2 Algorithmic Design Of Mono Vision Gesture Based Virtual Keyboard 57
  4.3 Summary 65
  4.4 References 65
6 5 Experimental Setup And Results 67
81.77 KB
  5.1 Development Environment & Results 67
  5.2 Prototype Setup For Virtual Keyboard 70
  5.3 Summary 72
  5.4 References 73
7 6 Conclusion & Future Work 74
47.7 KB
  6.1 Conclusion 74
  6.2 Future Work 76