HCI BASED APPLICATION FOR PLAYING COMPUTER GAMES | J4RV4I1014

Journal for Research | Volume 04 | Issue 01 | March 2018
ISSN: 2395-7549
All rights reserved by www.journal4research.org 48
HCI based Application for Playing Computer
Games
Mr. K. S. Chandrasekaran A. Varun Ignatius
Assistant Professor UG Student
Department of Computer Science & Engineering Department of Computer Science & Engineering
Saranathan College of Engineering-620012, India Saranathan College of Engineering-620012, India
R. Vasanthaguru B. Vengataramanan
UG Student UG Student
Department of Computer Science & Engineering Department of Computer Science & Engineering
Saranathan College of Engineering-620012, India Saranathan College of Engineering-620012, India
B. Vishnu Shankar
UG Student
Department of Computer Science & Engineering
Saranathan College of Engineering-620012, India
Abstract
This paper describes a command interface for games based on hand gestures and voice command defined by postures, movement
and location. The system uses computer vision requiring no sensors or markers by the user. In voice command the speech
recognizer, recognize the input from the user. It stores and passes command to the game, action takes place. We propose a simple
architecture for performing real time colour detection and motion tracking using a webcam. The next step is to track the motion of
the specified colours and the resulting actions are given as input commands to the system. We specify blue colour for motion
tracking and green colour for mouse pointer. The speech recognition is the process of automatically recognizing a certain word
spoken by a particular speaker based on individual information included in speech waves. This application will help in reduction
in hardware requirement and can be implemented in other electronic devices also.
Keywords: Computer Vision, Gesture Recognition, Voice Command, Human Computer Interaction
_______________________________________________________________________________________________________
I. INTRODUCTION
Computer games are one of the most successful application domains in the history of interactive systems even with conventional
input systems like mouse and keyboard. The existing system, mouse for instance seriously limits the way humans interact with
computers. Introduction of HCI techniques to the gaming world would revolutionize the way humans play games. It is a command
interface for games based on hand gestures and voice command defined by postures, movement and location. The proposed system
uses a simple webcam and a PC for recognizing the input from the user and thus uses natural hand movements for playing games.
This effectively reduces the cost of implementing HCI in conventional PCs. We propose a simple architecture for performing real
time colour detection and motion tracking using a webcam. Since many colours are detected it is important to distinguish the
specified colours distinctly. The next step is to track the motion of the specified colours and the resulting actions are given as input
commands to the system.
Speech technology is a very popular term right now. And Speech Recognition is the process of automatically recognizing a
certain word spoken by a particular speaker based on individual information included in speech waves. Using speech recognition,
we can give commands to computer and the computer will perform the given task. The main objective of this project is to construct
and develop a system to execute commands of operating system by using speech recognition system that is capable of recognizing
and responding to speech inputs rather than using traditional means of input(e.g. computer keyboard, mouse), thus saving time and
effort of the user. The proposed system is easier to increase the interaction between people and computers by using speech
recognition; especially for those who suffer from health problems, for example, the proposed system helps physically challenged
persons. This application will help in reduction in hardware requirement and can be implemented in other electronic devices also.
In this case, we are using TIC-TAC-TOE GAME.
II. RELATED WORK
A few works have been proposed recently to use free hand gestures in games using computer vision. A multimodal multiplayer
gaming system combines a small number of postures, their location on a table-based interaction system and speech commands to
interact with games and discusses results of using this platform to interact with popular games. In this study, a colour pointer has

HCI based Application for Playing Computer Games
(J4R/ Volume 04 / Issue 01 / 010)
been used for object recognition and tracking. Instead of conventional finger tips a colour pointer has been used to make object
detection easy and fast. Other tools facilitate the use of gesture recognition for applications in general, not only games.
Speech technology is a very popular term right now. Speech recognition is highly demanded and has many useful applications.
Conventional system users use pervasive devices such a mouse and keyboard to interact with the system. Further more people with
physical challenges find conventional system hand to use. A good system has minimal restrictions in interacting with its user. The
speed of typing and hand writing is usually one word per second, so speaking may be the fastest communication form with a
computer. The applications with voice recognition can also be a very helpful to for handicapped people who have difficulties
with typing.
III. SPECIFIC REQUIREMENTS
For Hand Gestures, we are using the OpenCV. In this study, a colour pointer has been used for object recognition and tracking.
Blue and green colours are used as colour pointers. Colour detection is performed using inbuild functions in OpenCV.
OpenCV (Open Source Computer Vision) is a Library of programming functions mainly aimed at real-time computer vision.
Originally developed by Intel’s research centre. OpenCV is written in C++ and its primary interface is in C++, but it still retains a
less comprehensive though extensive older C interface. OpenCV’s applications areas include: Facial recognition system, Gesture
recognition, Human-computer interaction (HCI), Mobile robotics. OpenCV is an open source computer vision and machine
learning software library. OpenCV was built to provide common infrastructure for computer vision applications and to accelerate
the use of machine perception in the commercial products. Being a BSD-Licensed products, OpenCV makes it easy for business
to utilize and modify the code. It has C, C++, Python, Java and MATLAB interfaces and supports Windows, Linux, Android and
Mac OS. OpenCV means mostly towards real-time vision applications and takes advantage of MMX and SSE instructions when
available. A full-featured CUDA AND OpenCV interfaces are being actively developed right now. There are over 500 algorithms
and above 10 times as many functions that compose or support those algorithms. OpenCV is written natively in C++ and has a
template interface that works seamlessly with STL containers.
For Voice Input, we are using the Sphinx package. CMU Sphinx, also called Sphinx in short, is the general term to describe a
group of speech recognition systems. SPIHNX the first large-vocabulary speaker-independent continuous-speech recognizer using
multiple code books of various LPC-derived features. Two types of HMMs are used in SPHINX: context-independent phone model
and function-word-dependent phone model. On a task using bigram grammar, SPHINX achieved word accuracy. This demonstrates
the feasibility of speaker-independent continuous-speech recognition, and the feasibility of appropriateness of hidden Markov
models for such a task. These include a series of speech recognizers. The speech decoders come with acoustic models and sample
applications. The available resources include in addition software for acoustic model training.
These two specific requirements are used in the TIC-TAC-TOE GAME.
IV. DESCRIPTION
Hand Gestures
In the TicTacToe Game, the OpenCV works:
The user use one color for the movement and another color for click options in their fingers. The user move the hands into the
webcam and the images are captured by openCV tool.
The captured images go to the segment analysis by BG Removal, and the segmented image is recognizes by the recognizer. The
recognized movement is mapped to the mouse tracker, and if the another color will identified the action takes placed.
Both the use of gestures and having games as an application brings specific requirements to an interface and analyzing these
requirements was one of the most important steps in designing Gestures. Gestures are most often used to relay singular commands
or actions to the system, instead of tasks that may require continuous control, such as navigation. Therefore, it is recommended that
gestures be part of a multimodal interface. This also brings other advantages, such as decoupling different tasks in different
interaction modalities, which may reduce the user's cognitive load. So, while gestures have been used for other interaction tasks in
the past, including navigation. All this leads to the requirement that the vocabulary of gestures in each context of the interface,
while small, must be as simply and quickly modifiable as possible. Systems that require retraining for each set of possible gestures,
for instance, could prove problematic in this case, unless such training could be easily automated. The interface should also accept
small variations for each gesture. Demanding that postures and movements be precise, while possibly making the recognition task
easier, makes the interaction considerably harder to use and learn, demanding not only that the user remember the gestures and
their meanings but also train how to do them precisely, greatly reducing usability. It could be argued that, for particular games,
reducing the usability could actually be part of the challenge presented to the player (the challenge could be remembering a large
number of gestures, or learning how to execute them precisely, for instance). While the discussion of whether that is a good game
design practice or not is beyond the scope of this paper, Gestures opts for the more general goal of increasing usability as much as
possible. This agrees with the principle that, for home and entertainment applications, ease of learning, reducing user errors,
satisfaction and low cost are among the most important design goals. The system should also allow playing at home with minimal
setup time required. Players prefer games where they can be introduced to the action as soon as possible, even while still learning
the game and the interface. Therefore, the system should not require specific background or lighting conditions, complex

(J4R/ Volume 04 / Issue 01 / 010)
calibration or repeated training. Allowing the use of the gesture-based interface with conventional games is also advantageous to
the user, providing new options to enjoy a larger number of games. From the developer point of view, the system should be as easy
as possible to integrate within a game, without requiring specific knowledge of areas such as computer vision or machine learning.
The Abstract Framework
Figure 1 shows a UML Activity Diagram Representing Gesture object flow model. It is responsible for the gesture model, while
GestureAnalysis and GestureRecognition define the interfaces for the classes that will implement gesture analysis and recognition.
To these activities are added image capture and segmentation. GestureCapture provides an interface for capturing 2D images from
one or multiple cameras or prerecorded video streams (mostly for testing). The images must have the same size, but not necessarily
the same color depth. A device could provide, for instance, one or more color images and a grayscale image to represent a dense
depth map. GestureSegmentation should usually find in the original image(s) one or both hands and possibly the head (to determine
relative hand position).
Fig. 1: Hand Gesture Process
Figure 1 shows that the usual flow of information in Gestures in each time step is as follows: one or more images serve as input
to the image capture model, which makes these images available as an OpenCV's Image object. The segmentation uses this image
and provides a segmented image as an object of the same class (and same image size, but not necessarily color depth). Based on
the segmented image, the analysis provides a collection of features as a GestureFeatureCol object which is in turn used by the
recognition to output a gesture.
GestreFeatureCol is a collection of GestureFeature objects. GestureFeature contains an identifier string to describe the feature
and either a scalar or an array of values (more often used) or an image (useful, for instance, for features in the frequency domain).
GestureFeature already defines several identifiers, for those features most often found in the gesture recognition literature, to
facilitate the interface between analysis and recognition, but user-created identifiers may also be used. Input is an optional module
that accompanies but is actually separate from Gestures. It is responsible for facilitating, in a very simple way, both multimodal
input and integration with games or engines not necessarily aware of Gestures.
It simply translates its input, which is a description (a numerical or string ID or a XML description, for instance) that may be
supplied either by Gestures or any other system (and here lies the possibility of multimodal interaction), into another type of input,
such as a system input (like a key down event) or input data to a particular game engine. In one of the tests, for instance, gestures
are used for commands and a dancing mat is used for navigation. Because this architecture consists mostly of interfaces, it is
possible to create a single class that, through multiple inheritance, implements the entire system functionality. This is usually
considered a bad practice in object orientation (should be avoided) and is actually one of the reasons why aggregation is preferred
to inheritance. There are design patterns that could have been used to force the use of aggregation and avoid multiple inheritance,
but Gestures opts for allowing it for a reason. Gesture recognition may be a very costly task in terms of processing and must be
done in real time for the purpose of interaction. Many algorithms may be better optimized for speed when performing more than
one task (such as segmentation and analysis) together.
Furthermore, analysis and recognition are very tightly coupled in some algorithms and forcing their separation could be difficult.
So, while it is usually recommended to avoid using multiple inheritance and to implement each task in a different class, making it
much easier to exchange one module for the other or to develop modules in parallel and in teams, the option to do otherwise exists,
and for good reason.

(J4R/ Volume 04 / Issue 01 / 010)
Speech Recognition
For Voice Input, we are using the Sphinx Library. Sphinx consists of in-build functions which are used to detect the speech input
given by a particular user. In this case, we are giving the position number as the input for TicTacToe Game through voice command.
The Speech Recognizer, recognize the user input and store it, so that it can be used as an input in TicTacToe Game.
Fig. 2: Speech Recognition Process
For Voice Input, human voice which is sampled at rate of 16,000per second. It should be given in live mode. It should be noted
that the system would have difficulty in recognizing accented English, hence it is recommended to give input as native English
speakers would do. In Speech Recognizer, Platform speed directly affected our choice of a speech recognition system for our work.
It is used to recognize the voice commands as spoken by user. The voice input is supposed to be given from its microphone. The
speech recognition process decodes these input files, identifies the command and generates output accordingly. The Dictionary file
is a separate file that describes the phonics of each word. It is a collection of pre-defined words that are relevant to various activities.
Each word is separated into syllables. The output in from the recognizer(i.e.) Recognition result is cross referenced with the
dictionary file and correct word is returned as the result. The Decoder module then parses this word and converts into equivalent
text. This text is then used by the command executer to execute the required functions. The decoder module is inbuild the SPHINX
System. The class Microphone from the sphinx for microphone control using java provides solve the essential functions for working
with microphone connected to the system. The StartRecording() function in the microphone class can be used to capture the audio
from the microphone connected to the system. It returns an object of the result class which can be converted to text command using
the recognize() function in Recognizer class.
Speech Recognition is the process of identifying the speech in the recorded audio by using the phonetics dictionary. The
phonetics for each word is stored in the dictionary file. The recognize() function in the recognizer class uses the grammar file to
recognize speech if any in the audio recorded by the microphone.
A different approach that has been studied very well is the analysis of the human voice for means of human computer interaction.
The topic of speech processing has been studied since the 1960’s and is very well researched. Even though the process of speech
processing, and more precise speech recognition. Speaker Recognition: The goal of a speaker recognition system is to determine,
to whom a recorded voice sample belongs. To achieve this, possible users of this system need first to enroll their voice. This is
used to calculate a so-called model of the user. This model is then later again used to perform the matching, when the system needs
to decide the owner of a recorded voice section. Such a system can either be built text-dependent, meaning that the user needs to
speak the same phrase during enrollment and usage of the system. This can be seen similar to the setting a password for a computer,
which is then required to gain access. A recognition system can also operate text independent, so that the user may speak a different
phrase during enrollment and usage. This is more challenging to perform but provides a more flexible way the system can be used.
Speaker recognition is mainly used for two tasks: Speaker verification assumes that a user claims to be of a certain identity. The

(J4R/ Volume 04 / Issue 01 / 010)
system is then used to verify or refute this claim. A possible use case is to control access for a building. Here, the identity claim
could be provided by the usage of a secondary system like smart cards or fingerprints. Speaker identification: It does not assume
a prior identity claim and tries to determine to whom a certain voice belongs. These systems can either be built for a closed group,
where all possible users are known to the system beforehand. Or it can be used for an open group, meaning that not all possible
users of the system are already enrolled. Most of the time this is achieved by building a model for an “unknown speaker”. If the
recorded voice sample fits this “unknown model” the speaker is not yet enrolled. The system can then either just ignore the voice
or build a new model for this speaker, so that he later can be identified.
Speech Recognition
The idea behind speech recognition is to provide a means to transcribe spoken phrases into written text. Such a system has many
versatile capabilities. From controlling home appliances as well as light and heating in a home automation system, where only
certain commands and keywords need to be recognized, to full speech transcription for note keeping or dictation. There exist many
approaches to achieve this goal. The most simple technique is to build a model for every word that needs to be recognized. As
shown in the section pattern matching this is not feasible for bigger vocabularies like a whole language. Apart from constraining
the accepted phrases, it must be considered if the system is only used by a single individual, so-called speaker dependent, or if it
should perform equally well for a broad spectrum of people, called speaker independent.
Speech Processing
Although speaker recognition and speech recognition systems achieve different goals, both are built from the same general
structure. This structure can be divided into following stages:
 Signal generation models the process of speaking in the human body.
 Signal capturing & preconditioning deals with the digitalization of voice. This stage can also apply certain filters and
techniques to reduce noise or echo.
 Feature extraction takes the captured signal and extracts only the information that is of interest to the system.
 Pattern matching then tries to determine to what word or phrase the extracted featured belong and concludes on the system
output.
V. CONCLUSION
The System architecture proposed will completely revolutionize the way gaming applications are performed. The system makes
use of web camera and which are an integral part of any standard system, eliminating the necessity of additional peripheral devices.
Our endeavor foe object detection and image processing in OpenCV for the implementation of the gaming console provide to be
practically successful. Here a person’s motions tracted and interpreted as commands. Most gaming application required additional
hardware which is often very costly. The motive was to create this technology in the cheapest possible way under a standardized
operating system using a set of wearable gestureable interfaces. This technique can be further extended to other services also. In
Speech Recognition, the proposed system provides a better performance and experience I user interaction. It reduces potential time
involved in conventional interactive methods. Voice recognition is irrefutably the feature of human computer interaction and our
study proposes a new approach to utilize this approach for enhanced user experience and extend in ability of computer interaction
to motor impaired users.
REFERENCES
[1] D. Bowman, E. Kruijff, J. LaViola, I. Poupyrev, 3D User Interface: Theory and Practice, Addison-Wesley, 2005.
[2] T. Stamer, B. Leibe, B. Singletary, J. Pair, "Mind-warping: towards creating a compelling collaborative augmented reality game", Proc. of the 5th international
conference on Intelligent user interfaces (IUI '00), pp. 256-259, 2000.
[3] M. Snider, "Microsoft unveils hands-free gaming", USA Today, June 2009, [online] Available: http://www.usatoday.com/tech/gaming/2009-06-01-hands-
free-microsoft_N.htm.
[4] V. Dobnik, Surgeons may err less by playing video games, 2004, [online] Available: http://www.msnbc.msn.com/id/4685909.
[5] S. LAAKSO, M. LAAKSO, "Design of a Body-Driven Multiplayer Game System", ACM Comput. Entertain. vol. 4, no. 4, 2006.
[6] A. Camurri, B. Mazzarino, G. Volpe, "Analysis of Expressive Gesture: The EyesWeb Expressive Gesture Processing Library" in Gesture-Based
Communication in Human-Computer Interaction, Springer, pp. 469-470, 2004.
[7] B. Shneiderman, Designing the User Interface: Strategies for Effective Human-Computer Interaction, Addison-Wesley, 1998.
[8] Cinneide, A., Linear Prediction: The Technique, Its Solution and Application to Speech. Dublin Institute of Technology, 2008.
[9] Gutierrez-Osuna, R. CSCE 689-604: Special Topics in Speech Processing, Texas A&M University, 2011, courses.cs.tamu.edu/rgutier/csce689 s11.
[10] Schalkwyk, J., et al., Google Search by Voice: A case study. In Advances in Speech Recognition. Springer, 2010.
[11] Rabiner, L., A tutorial on hidden Markov models and selected applications in speech recognition. In Proceedings of the IEEE, vol. 77, no. 2 pp. 257-286,
1989.
[12] Campbell, J.P. Jr., Speaker recognition: a tutorial. In Proceedings of the IEEE, vol. 85, no. 9 pp. 1437-1462, 1997.

HCI BASED APPLICATION FOR PLAYING COMPUTER GAMES | J4RV4I1014

More Related Content

What's hot (20)

Similar to HCI BASED APPLICATION FOR PLAYING COMPUTER GAMES | J4RV4I1014 (20)

More from Journal For Research (20)

Recently uploaded (20)

HCI BASED APPLICATION FOR PLAYING COMPUTER GAMES | J4RV4I1014