This document describes a sign recognition system to translate sign language gestures from deaf and dumb people into audible speech. The system uses Harris corner detection and SIFT to extract features from captured images of gestures. K-nearest neighbors algorithm is used to match features to a database of images and identify the sign. A serial board communication kit then converts the text result to speech using a text-to-speech module and speaker. Examples show the system correctly identifying the signs for letters A, B, and C from captured images.
Conversion of sign language to speech using kinect
Report
Share
1 of 20
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
More Related Content
Ppt final
1. SIGN RECOGNITION System FOR DEAF AND
DUMB PEOPLE
GUIDED BY PRESENTED BY
Ms.I.KANAGA SANGEETHA M.E
ASSISTANT PROFESSOR M.RINCY JULIET
DEPARTMENT OF ECE (98010106052)
A.SUBACHITHIRAI MEENAL
(98010106068)
M.SURIYA
(98010106072)
Department of ECE
2. OBJECTIVE
To develop a system for automatic translation
of static gestures used by dumb and deaf persons
into audible speech understandable by a common
person.
3. INTRODUCTION
• Sign recognition system act as a interpreter
between common people and socially aided
people.
• It captures the gesture shown by the deaf and
dumb people and convert immediately into
known language.
4. LITERATURE SURVEY 1
Title Sign language recognition system
Author Er.Aditi Kalsh, Dr.N.S.Garewal
Years June 2013
Journal
International Journal of Engineering
Research
Explanation
Canny edge and peak detection is used for feature
extraction.
5. LITERATURE SURVEY 2
Title Real Time Indian Sign Language Recognition System to
aid Deaf and Dumb people
Author Rajam, P. Subha and Dr G Bala krishnan
Years June 2011
Journal 13th International Journal Communication
Technology(ICCT)
Explanation Cam shift and HSV approach for recognizing gesture
6. LITERATURE SURVEY 3
Title SignPro-An Application Suite for Deaf and Dumb
Author Ashish Sethi, Hemanth, Kuldeep kumar
Years
May 2012
Journal
International Journal of Computer Science Engineering
Technology
Explanation
Skin segmentation and region growth approach is used
for image recognition
9. Harris Corner Detection
• In the harries corner detection we
calculate the cim , r, c.
• Cim - binary image marking corners.
• r - Row coordinates of corner points.
• c - Column coordinates of corner points
10. Histogram
• Histograms can be used to provide
useful image statistics Information
• It plots the number of pixels for each
value
• Histogram we are using the num-bins
function
12. K-NN Algorithm
• This approach is used to match the key points
calculated from the capture image using the
harries and sift.
• Now the key points matched with the database
image.
• IDX = knn search searches the reference data
set to find the k-nearest neighbours of each
query point represented by each row.
13. Text to Speech
• Serial board communication kit is used to
convert Text to Speech.
• This Kit contains PIC Microcontroller,
Selector, Voice recorder, Amplifier, Speaker,
Step down transformer.