Scale Invariant Braille Translator Student: Yaniv Tocker Final - - PowerPoint PPT Presentation

Scale Invariant Braille Translator

Student: Yaniv Tocker Final Project in 'Introduction to Computational & Biological Vision' Course

Motivation

2

 Optical Character Recognition (OCR):  Automatic translating of letters/digits in images to a

form that a computer can manipulate (Strings, ASCII codes)

Background Braille Translator Methods Conclusions

Scale Invariant Braille Translator

3

Why is this important?

 Replacing data entry clerks  Reading car plates  Making electronic copies of books researchable

Scale Invariant Braille Translator

Background Braille Translator Methods Conclusions

4

Braille OCR

 A less investigated field of OCR  Can assist the vision-impaired

Scale Invariant Braille Translator

Background Braille Translator Methods Conclusions

5

Braille Translator

 Goal: being able to translate braille language from an

image to English letters

Scale Invariant Braille Translator

Background Braille Translator Methods Conclusions

 The system is required to be robust to scale changes,

since circle can appear in different sizes

6

Methods

Braille Circle Detection Histogram

• f Radius

Find areas Create Dictionary Create Patch Letter Decision Braille Circle Detection Histogram

• f Radius

Find areas Create Dictionary Create Patch Letter Decision

Scale Invariant Braille Translator

Original Image Detected Circles Radius Histogram Circles filtered by common radius size Circles filtered by common radius size & amount of neighbors Circles Mask Dynamic mask using common radius Masks convolution results Points that are local maxima & convolution result above a threshold Braille meaningful circles Braille meaningful circles mask Braille meaningful circles mask with centers Braille meaningful – creating patches Braille patches Dictionary Creation Dictionary Original image First Patch Interpretation Result

Background Braille Translator Methods Conclusions

7

Scale Invariant Braille Translator

GUI

 A user friendly GUI was created to easily operate the

software

Background Braille Translator Methods Conclusions

8

Conclusions & Future Work

 A scale invariant Braille translator has been proposed  The main key is to find the common radius in the image

& build the dictionary according to it

Scale Invariant Braille Translator

 Future addition could be to detect if braille writing is in

an image in order to assist the vision impaired

Background Braille Translator Methods Conclusions

References

• 1. OCR: http://en.wikipedia.org/wiki/Optical_character_recognition
• 2. Braille OCR

http://en.wikipedia.org/wiki/Optical_braille_recognition

• 3. Braille OCR example

http://www.ni.com/white-paper/6470/en/

• 4. Braille code generator

http://braille.compelo.com/generate/

• 5. X. fernanadez et al, “A braille O.C.R for the blind”

http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.141.7727&rep=rep1&type= pdf

• 6. J.Mennans et al,”Optical recognition of braillle writing using standard equipment”.

http://bauhaus.ece.curtin.edu.au/~iain/PhD%20BU/A_Phd%20docs/To%20read/Ac cessibility%20info/Research/Braille_Articles/OCR%20of%20Braille.pdf

• 7. O. ben-shahar – lecture notes from ICBV 2014 –
• object classification
• hough transform

9

Scale Invariant Braille Translator

Background Braille Translator Methods Conclusions

10

Scale Invariant Braille Translator

Background Braille Translator Methods Conclusions