You are on page 1of 4

Group: 02

Group members:
Name ID
Wasifa Mashiyath 1806003
Khadiza Sultana 1806013
Raihan Mahmud Chowdhury 1806014
M.M. Yeamin 1806016

Title of the project:


Voice Recognition Based Ticket System for BUET Bus service.

Abstract:
The main purpose of this project is to develop a voice-based ticket system for BUET
bus service instead of regular ticket system. Students face several problems while
going through the whole process of getting tickets for the existing bus service.
Through this project, we will try to solve those problems by converting the existing
system to a voice recognition-based ticket system where students will just need to
tell their roll before entering the bus. Our process will recognize the students by their
voice and store the data about how many times they get on the bus in a month. The
data will be added to their BIIS account monthly and through the BIIS account, they
will be able to easily do their payment. Hence, they won’t need to buy more tickets
than their need, won’t have to waste a lot of time in the billing process through bank
and won’t have to carry tickets all the time and also there will be no chance to lose
one.
The project will be implemented by MATLAB. The system has two parts – Voice
recognition process which contains speech analysis and pattern recognition, and data
storing and calculation part for showing the billing process.

Component:
1. MATLAB R2020a
2. Microphone as recording device
Data Acquisition & Verification Setup:
Just beside the area where students get on the bus, we can manage a place and keep
a door, a microphone connected to a mobile phone, and a screen. Whenever a student
wants to get on any BUET bus, he/she must pass through this set of arrangements.
They will stand in a line and gradually pass through the door after proper
verification. An app will be built which will do the work of speech recognition. The
microphone will record their voice, which will be sent to the mobile app. In this app,
our proposed speech recognition process will use a database already created to verify
whether the student is the one whose name and roll are uttered by him/her. And the
app will show the result on a moderate-sized screen. As a result, there will have no
possibility of illegally passing through the door. Again, the screen will show' not
allowed' if any student does not pay the bill for more than a particular period. We
will try to build an efficient model to verify the students quickly and properly,
otherwise there may arise a long line, and things may become more complex then.
To ensure more security, we can use Arduino in our project. Instead of showing on
the screen, the mobile app will send a signal to the door so that it will allow the
student to go, otherwise not. Or to ensure the system is more secured, we can keep
CCTV in this area

Speech Recognition Process:


Bill Payment through App:
Step 1 : Log in the app Step 2:

Step 3: Checking travelling history Step 4: Paying the bill: go to payment and click on ‘Pay’

Step 5: Choosing the payment gateway


Challenges:
• Extracting features from noisy speech.
• Interference of multiple speakers’ voice.
• Text independent speech recognition.

You might also like