Professional Documents
Culture Documents
Hell 3
Hell 3
Bachelor of Technology
in
by
October, 2016
Certificate of Examination
Roll Number: 116CS0224, 716CS1046
Name: Tarinee Prasad Sahoo, Kumar Mapanip Saheb Sahoo
Title of Dissertation: Gesture Control Gaming
We the below signed, after checking the project report mentioned above and the official
record book (s) of the student, hereby state our approval of the project report submitted in
partial fulfillment of the requirements of the degree of Bachelor of Technology in Computer
Science and Engineering at National Institute of Technology Rourkela. We are satisfied with
the volume, quality, correctness, and originality of the work.
Arun Kumar
Principal Supervisor Member, DSC
Ashok K Turuk
Head of the Department
Department of Computer Science and Engineering
National Institute of Technology Rourkela
Prof.
Supervisors’ Certificate
This is to certify that the work presented in the project report entitled Gesture Control
Gaming submitted by Tarinee Prasad Sahoo, Kumar Mapanip Saheb Sahoo, Roll
Number 116CS0224, 716CS1046, is a record of original research carried out by him under
our supervision and guidance in partial fulfillment of the requirements of the degree of
Bachelor of Technology in Computer Science and Engineering. Neither this project report
nor any part of it has been submitted earlier for any degree or diploma to any institute or
university in India or abroad.
Arun Kumar
Associate Professor
Dedication
This project is dedicated to all those people who have helped us in different stages of life.
We’re thankful to our teachers, friends and family who supported us during the tenure of
this project.
Signature
Declaration of Originality
I, Tarinee Prasad Sahoo, Kumar Mapanip Saheb Sahoo, Roll Number 116CS0224,
716CS1046 hereby declare that this project report entitled Gesture Control Gaming presents
my original work carried out as a undergraduate student of NIT Rourkela and, to the best of
my knowledge, contains no material previously published or written by another person, nor
any material presented by me for the award of any degree or diploma of NIT Rourkela or
any other institution. Any contribution made to this research by others, with whom I have
worked at NIT Rourkela or elsewhere, is explicitly acknowledged in the dissertation. Works
of other authors cited in this dissertation have been duly acknowledged under the sections
“Reference” or “Bibliography”. I have also submitted my original research records to the
scrutiny committee for evaluation of my dissertation.
I am fully aware that in case of any non-compliance detected in future, the Senate of NIT
Rourkela may withdraw the degree awarded to me on the basis of the present dissertation.
Certificate of Examination ii
Dedication iv
Declaration of Originality v
Acknowledgment vi
Abstract vii
1 Introduction 1
1.1 Why gesture? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 How Gesture Recognition works? . . . . . . . . . . . . . . . . . . . . . . 1
1.3 Stages of Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.4 Related Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
References 9
Dissemination 10
Index 11
viii
Chapter 1
Introduction
The main aim behind this project is to make an interaction between human and computer
using various applications running on computer by aiming basic shapes made by hand.Our
hand movements have an important role while interacting with other people, as they convey
very rich information in many ways. Gesture recognition could be a quite sensory activity
computing program that permits computers to capture and interpret human gestures as
commands. Gesture Recognition is largely the power of a computer to grasp gestures and
execute commands supported those gestures .
Many companies are working on gesture recognition system . Most of them are working
with additional sensors like wireless gloves with sensors or the most popular Microsoft
Kinect which uses a Depth camera which basically consists of some system that segment
the body , and based on the segmentation they produce the getsure . On the contorary we
decided to feed a neural network with a large amount of data and seeing whether we can
achieve the result without using much of hardware . We only need a laptop camera to build
a gesture recognition system .
1
• The Programmed package then identifies meaningful gestures from a planned gesture
library wherever every gesture is mapped to a associated command.
• The package then associates every real�time gesture, interprets it , uses the library
to spot meaningful gestures that match with the given gesture. Once the gesture has been
understood, classification is performed.
3
Chapter 2 Algorithms Used and Working
and background pixels, otherwise a poor threshold is chosen. On the opposite hand, selecting
regions that area unit overlarge will violate the idea of roughly uniform illumination.
This technique is a smaller amount computationally intensive than the Chow and Kaneko
approach and produces smart results for a few applications.
4
2.2.1 CNN Parameters
• First layer : Convolution layer with 5 * 5 stride with ‘relu’ activation function .
• Second layer : Maxpooling layer with 2 * 2 pooling window .
• Third layer : Convolution layer with 5 * 5 sliding with ‘relu’ activation function .
• Fourth layer : Maxpooling layer with 2 * 2 pooling window
• Flattening of outputs from fourth layer
• Fifth layer : Full connection layer with 256 input nodes and ‘relu’ Activation .
• Sixth layer : Four nodes with sigmoid function
Chapter 3
Progress and Conclusion
Setup
We developed our algorithms on Python3 . And we employed Tensorflow 1.4,Keras 2.0.9
to build the network and the loss function . We used CV2 3.3.0 for image processing and
drawing bounding boxes as well as text. For mathematical computations Numpy 1.2.1 was
used . For network model training NVIDIA NC6 in AWS
Dataset
In order to train a convolutional neural network, a lot of data is required for good accuracy .
All the dataset has been prepared and trained by us.
Progress
3.1 Data Collection
• We used OpenCv module of python for image processing, image is extracted from live
video with camera frame by frame .
• Images are then converted to grayscale.
• Adaptive mean thresholding with binary thresholding is used and Mask is applied.
• The final image that was saved was a 64x64 grayscaleimage.
6
Chapter 3 Progress and Conclusion
• The model consists of two CONV BOX followed by a fully connected layer.
• One CONV BOX consists of three layers
Convolution Operation - > Relu - > Max Pooling
• So, this operation was carried out twice.
• Then flattening of the data is performed and fed into Fully connected ANN.
Fig.
1
The plot Fig. 1 shows the steps performs in training the neural network with a appropriate
set of hyperparameter with proper tuning .
Result
Fig.
2
The plot Fig. 2 shows the validation loss for the training set .
7
Fig. 3
The plot Fig. 1 shows the validation loss vs number of epochs and accuracy vs number of
epochs for the test set .
• https://www.youtube.com/watch?v=FTr3n7uBIuE
• http://cs231n.github.io/convolutional-networks/
• https://ujjwalkarn.me/2016/08/11/intuitive-explanation-convnets/
• http://colah.github.io/posts/2014-07-Conv-Nets-Modular/
• http://andrew.gibiansky.com/blog/machine-learning/convolutional-neural-networks/
0
This reference format follows ASME style. You are advised to follow one reference format of any
dominant journal of your field.
9
Dissemination
An Empirical Evaluation on Semantic Search Performance of Keyword-Based and
Semantic Search Engines: Google, Yahoo, Msn and Hakia (by Duygu Tümer,
Mohammad Ahmed Shah, Yıltan Bitirim)
layout, 1