Please use this identifier to cite or link to this item:
Title: Real-Time Speech Recognition System for Prosthetic Arm Control
Authors: Samant, Piyush
Supervisor: Agarwal, Ravinder
Keywords: Speech Recognition;Prosthetic Arm
Issue Date: 26-Aug-2014
Abstract: Prosthesis is an artificial approach, which is used to replace a disabled body part. Prostheses are typically used to replace and provide supplement to disabled/defective body parts. Disabled body parts can be of any reason like lost accidently, birth physical disability etc. In addition to the standard artificial limb for every-day use, many disable have special limbs and devices to aid in the participation of sports and recreational activities. The main requirement is that its function should be as natural as real arm. There are various designs of artificial arm that are available in the market, categorized as electrical, mechanical and Myo-electric arm. Mechanical prostheses use some motion of the body to provide the power necessary to control the prosthetic component. Electrical arms activate the hand by a motor which is driven by micro switches and relays. Voice is the most popular and easy tool for communication and simple tool for man machine interface, as it is user friendly and wireless. In this study the hardware of voice controlled prosthetic arm was designed and implemented. For man machine interface through speech a voice recognition module was used, which was initially trained for five voice commands to control the movements of hand. This prosthetic hand having two degrees of freedom was designed with the help of geared DC motor, which has the ability of simply picking up and placing the objects. Further, the prosthetic arm was tested using a glass that contains different quantity of water in it and their respective response time was calculated in placing the glass from one fix position to another fixed position. As soon as the condition of failure exists, the motor was changed with higher specifications and the same procedure was repeated using higher weights.
Description: ME-Dissertation
Appears in Collections:Masters Theses@EIED

Files in This Item:
File Description SizeFormat 
3070.pdf1.52 MBAdobe PDFThumbnail

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.