Surveyreport 1
Surveyreport 1
DEAF PEOPLE
Abstract - Hand Gesture recognition system is more familiar or easy to use to the human beings.
one of the best technique has developed in last few The communication between the dumb, visually
year, the reason behind to get success is ability to impaired person are made only by their expressions
cooperate with machine. Body gestures are and their hand gestures, We may design a system
consider most common way to communication which is able to communicate or can be median for
along human and personal computer in virtual society.
world. To developed effective algorithm to that can
recognize, interpret, process, and simulate human 2. BLOCK DIAGRAM
affections. Advance computing has ability to
humanize digital interactions. This pepper
introduces application using Computer Vision(CV)
to recognize hand gesture. A camera takes live
footage and analysis gestures on the basis of
analysis system takes decision and recognize
gestures.
1. INTRODUCTION
In country like India we found 40 million people are
visually impaired person and 2.42 people are deaf
and able to speak. The vision and voice are the main
problem in this people. They only have one median Fig1: Block Diagram
to connect to society that is sigh language. To the
society it is mandatory to understand there sign Above fig shows that the block of the proposed
language. Some time it go’s hard to society to system.it this system we have used computer to
understand there sign language. So the we have recognize hand gestures. The input of computer
design a system which may help society to system is given to through the camera. Which can
understand there sign language. The system has be systems camera or external USB camera. Here
design with computer system and camera. The in the system hand gestures and of human are the
system or algorithm is able to recognize sign input given to camera. Given hand gestures are
language. In this algorithm Image processing has getting recognized by applying image processing
used as to recognize sign. The complete algorithm on to given input to apply image processing we
has developed using python programming have used python machine algorithms have used.
language. Python is one of the best programing
language to do develop complex algorithm. 3.FLOW CHART
1.2. SOLUTIONS
5. RELATED WORK
4.1 OPEN CV
Several research works have been done in the area of
words prediction using hand gesture recognition. For
instance, Guo and Wang (2017) proposed a novel
method of using hand gesture recognition to predict
words for individuals. Their study utilized a
convolutional neural network (CNN) algorithm to
classify hand gestures, and a Hidden Markov Model
(HMM) to predict words. The proposed method
Open cv computer vision library which is achieved high accuracy and efficiency in predicting
dominated for image processing it also called open words.
source computer vision library. Mainly used for
real time computer vision and image processing. Another research work done in this field was carried out
The main library has design in c ++. Open CV open by Pei et al. (2018). They proposed a real-time hand
source library started in 2011. gesture recognition system that can predict words for
people. Their system employed a deep learning-based
4.2 MEDIA PIPE
framework, which was trained using a large dataset of
Media pipe python library which has design for hand gestures. Their proposed system achieved an
machine learning package which has built in trained accuracy of 97.23% and a response time of 0.12
data set in tensor flow. it can form the basis for sign seconds.
language understanding and hand gesture control,
and can also enable the overlay of digital content and Similarly, Jiang and Li (2018) proposed a hand gesture
information on top of the physical world in recognition system that uses machine learning
augmented reality. Media pipe most perfect solution algorithms to predict words for individuals. Their
for the hand and finger tracking solution. It employs system used a combination of hand feature extraction
Deep learning to infer 21 3D landmark on hand. and gesture classification techniques to recognize hand
MediaPipe Hands utilizes an ML pipeline consisting gestures, and a support vector machine (SVM)
of multiple models working together: A palm algorithm to predict words. Their proposed system
achieved an accuracy of 95.2%.
Furthermore, a research work done by Zhang et al. have several devices connected to it. So to correctly
(2019) proposed a system that can recognize hand detect the camera we should get the id of the device.
gestures and translate them into spoken words for the For that type command ‘imaqhwinfo (‘winvideo’)’
deaf. Their system utilized a combination of machine in the command window of MATLAB. The frames
learning algorithms, including the HMM and SVM per trigger and frame grab intervals are specified in
algorithms, to recognize hand gestures and translate the program. When the specified frame is acquired
them into spoken words. Their proposed system the image is captured with webcam. The software
achieved an accuracy of 92.7%. code and supporting tools used are based on the
leading software in the field, MATLAB and the
In addition, Wang and Chen (2020) proposed a system Image Processing Toolbox from The Math Works.
that can recognize hand gestures and translate them into
MATLAB is a high-performance language for
written words for the deaf. Their system utilized a
technical computing which is very popularly used
convolutional neural network (CNN) algorithm to
nowadays. It integrates computation, visualization
recognize hand gestures and a deep learning-based
and programming in an easy-to-use environment
model to translate them into written words. Their
where problems and solutions can be easily
proposed system achieved an accuracy of 96.78%. expressed in mathematical notations. The basic
commands used are given in “table 1”
COMMANDS FUNTIONS
Imread To read an image
REFRENCES:
[1].https://google.github.io/mediapipe/solutions/ha
nds