Top Banner
Phone Reader Project By: Hossein and Hadi Shayesteh Supervisor: Mr. James Connan
18

By: Hossein and Hadi Shayesteh Supervisor: Mr. James Connan.

Dec 24, 2015

Download

Documents

Gilbert Horn
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: By: Hossein and Hadi Shayesteh Supervisor: Mr. James Connan.

Phone Reader Project

By: Hossein and Hadi ShayestehSupervisor: Mr. James Connan

Page 2: By: Hossein and Hadi Shayesteh Supervisor: Mr. James Connan.

Introduction

Phone Reader

• Intended for Blinds and Illiterates

• People dealing with a Foreign Language

Potential Users

• Converting Image to text using OCR

• Reading out the text using TTS

Page 3: By: Hossein and Hadi Shayesteh Supervisor: Mr. James Connan.

Phone Reader Design / Phone Device

Generated Audio

Page 4: By: Hossein and Hadi Shayesteh Supervisor: Mr. James Connan.

Implementation / Phone Device

Page 5: By: Hossein and Hadi Shayesteh Supervisor: Mr. James Connan.

Implementation / Server

Page 6: By: Hossein and Hadi Shayesteh Supervisor: Mr. James Connan.

Implementation / Phone Device

Page 7: By: Hossein and Hadi Shayesteh Supervisor: Mr. James Connan.

Phone Reader Design / Server

Gray Scale

OCR engine

Thresholding

Server

Page 8: By: Hossein and Hadi Shayesteh Supervisor: Mr. James Connan.

Project Testing

Functionality Testing

• Ease of use

• Fulfillment of functionality

Usability Testing

• Unit testing

- Server - Phone Device

• Integration Testing

Page 9: By: Hossein and Hadi Shayesteh Supervisor: Mr. James Connan.

Functionality Testing / Server

15 different images captured from screen

Case 1 images captured by camera

Case 4 :

Noise and color applied to Case 1 images Case 2 :

Case 3 :

Case 1 :

Case 5 :

Case 2 images captured by camera

Image preprocessing applied to Case 1 images

Case 6 :

Case 7 :

Case 8 :

Image preprocessing applied to Case 2 images

Image preprocessing applied to Case 3 images

Image preprocessing applied to Case 4 images

Page 10: By: Hossein and Hadi Shayesteh Supervisor: Mr. James Connan.

Functionality Testing Results

1 2 3 4 5 6 7 8

0

10

20

30

40

50

60

70

80

90

100

OCR functionality test results15 different images captured from screen

Case 1 images captured by camera

4 :

Noise and color applied to Case 1 images 2 :

3 :

1 :

5 :

Case 2 images captured by camera

Image preprocessing applied to Case 1 images

6 :

7 :

8 :

Image preprocessing applied to Case 2 images

Image preprocessing applied to Case 3 images

Image preprocessing applied to Case 4 images

Page 11: By: Hossein and Hadi Shayesteh Supervisor: Mr. James Connan.

Functionality Testing Results (Contd)

• OCR engine produces near ideal results for high quality images

• Probable poor OCR results are due to noise, colored background or distractions

• OCR performance was greatly improved by applying image pre-processing techniques

What we understood from the test result

What we did to improve the performance

Page 12: By: Hossein and Hadi Shayesteh Supervisor: Mr. James Connan.

Functionality Improvement / Phone Device

Employing Phone’s Camera in Phone Reader

Improving the User Interface (Buttons,Options)

Page 13: By: Hossein and Hadi Shayesteh Supervisor: Mr. James Connan.

Usability Testing

Questionnaires

• 10 participants

• Task sheet provided

• Feedback analyzed

Page 14: By: Hossein and Hadi Shayesteh Supervisor: Mr. James Connan.

Usability Testing Result

12

34

5

0

10

20

30

40

50

60

70

80

90

100

OCR functionality test results

Successful completion of all the tasks

How well buttons were organized

4 :

How easy was working with Phone Reader 2 :

3 :

1 :

5 :

Function of each button was easy to understand

Overall impression of the Phone Reader application

Page 15: By: Hossein and Hadi Shayesteh Supervisor: Mr. James Connan.

Usability Testing Result (Continued)

What we understood from the test result

• Phone Reader is fairly easy to use

• The most challenging part was working with camera

• We tried to simplify the process of image capturing by eliminating unnecessary option buttons

What we did to improve the performance

Page 16: By: Hossein and Hadi Shayesteh Supervisor: Mr. James Connan.

User Manual

• Run the application

• Capture an Image

• Press OCR button

• Press TTS button

Page 17: By: Hossein and Hadi Shayesteh Supervisor: Mr. James Connan.

Demo

Page 18: By: Hossein and Hadi Shayesteh Supervisor: Mr. James Connan.

Question and Answer