Need help in Braille character recognition

  • Thread starter mimitha
  • Start date
In summary, DaveE is trying to find a way to convert braille characters into normal text by scanning a real braille document. He advises using an array of braille characters corresponding to letters and numbers, and scanning the document to get the input. He also mentions that there is a full description of "level 2" Braille including the contractions which is a 200 page document. Level 3 Braille, which is not formally specified, would be a good topic for a PhD in artificial intelligence.
  • #1
mimitha
4
0
Need help in Braille character recognition!

Need Help on Braille character recognition!
Hai is there any1 to help me with the braille character recognition.. i need to convert the braille characters into normal text by using matlab!

Its my final year project i really need a help because i don't even know how to start it... PLS help me with the source code!


~mimitha
 
Technology news on Phys.org
  • #2


Wait, braille character recognition? Wouldn't that require some sort of physical sensor to read data? You could in theory do image recognition, although the important part about braille is the physical protrusion, making images unreliable (sort of defeats the purpose).

I don't know much of anything about Matlab, but how are you going about getting your input?

DaveE
 
  • #3


Use an array with 3x2 matrix representations of the dots, corresponding to letters and numbers?
 
  • #4


davee - by scanning the real braille document

jack the stri - i advised to form an array with the use of region probs, thresholdin, bw label and centroid,location.

and i do have to find the minimum maximum function to get the coordination area
 
  • #5


mimitha said:
davee - by scanning the real braille document

Scanning? As in, putting a piece of paper with Braille on it on a flatbed scanner, taking the resulting image, and then turning it into normal text? That'll take some doing. I'd assume the steps to take would be:

1) Determine X,Y positions on the image of all dots
2) Determine the angle of text (horizontal, vertical, angled)
3) Determine all the separate lines of text
4) For each line of text, cycle through an array of braille characters to see which one(s) match. Eliminate dots from your map as you recognize characters.
5) Make sure you don't have any dots remaining-- if you do, something's wrong.
6) Display the text translations you matched!

I would think the hardest parts would be recognizing a braille dot, and getting the text angle precisely enough to be accurate.

DaveE
 
  • #6


post what it looks like from your scanner.
 
  • #7


I don't really have a real braille documents yet.. but then i asked to define the 3x2 array dots into the matlabs 1st by using the normal braille characters & letters which is shown here http://en.wikipedia.org/wiki/Braille" then once define all the characters inside... and then scanned the real braille document n convert it into normal text so all can read it...
 
Last edited by a moderator:
  • #8


mimitha said:
i asked to define the 3x2 array dots into the matlabs [...] then once define all the characters inside...

That really seems like an incredibly trivial part of this assignment. Make each dot a "bit" and make a corresponding character translation table. The character "R" for example has bits 1, 2, 3, and 5 turned on, making for a numeric value of 23 (that's following the standard number labelling of Braille dot positions). A translation array might look something like:

0) space
1) A or 1
2) ,
3) B or 2
4) '
5) K
6) ;
7) L
8) N/A
9) C or 3
10) I or 9
... etc ...

Given that letters are the default, and numbers will be in use if following #60 ("number follows" symbol), you could even make TWO arrays-- the first being alphabetic characters, and the second being numeric characters. Probably put in the characters 0-9, and also the "." character (assuming that's how Braille denotes decimal points). Then, when you encounter a #60, you set a flag-- and as long as the flag is set, use the 2nd array to translate rather than the 1st. And if something pops up that isn't defined in your 2nd array, unset the flag and go back to using the 1st array to translate.

Anyway, that part's dead easy.

mimitha said:
and then scanned the real braille document n convert it into normal text so all can read it...

There's your REAL task. Especially if you use genuine Braille documents, which supposedly contain a lot of context-sensitive shorthand, like using the letter "B" to mean "BUT", rather than spelling it out. Here's hoping you don't have to do anything THAT fancy.

DaveE
 
Last edited:
  • #9


davee123 said:
Anyway, that part's dead easy.

It's not so easy for "real world" Braille, which uses many contractions, or abbreviations, similar to texting. The meaning of a character in the input often depends on its context.

The full description of "level 2" Braille including the contractions is 200 page document. Dealing with "level 3" Braille, which is widely used but not formally specified anywhere so far as I know, would probably be a good topic for a PhD in artificial intelligence.

But you are right that all this is independent of the other major problem, identifying the dot patterns from a scanned image.
 

1. What is Braille character recognition?

Braille character recognition is the process of converting written or printed Braille text into digital text that can be read by a computer or other electronic device.

2. Why is Braille character recognition important?

Braille character recognition is important because it allows visually impaired individuals to access printed materials and information. It also enables them to use technology and electronic devices, which can greatly improve their independence and quality of life.

3. How does Braille character recognition work?

Braille character recognition works by using a scanner or camera to capture an image of the Braille text. The image is then processed by software that uses optical character recognition (OCR) technology to analyze the patterns of dots and convert them into digital text.

4. What are the challenges of Braille character recognition?

One of the main challenges of Braille character recognition is the variation in the placement and spacing of Braille dots, which can lead to errors in the recognition process. Another challenge is the recognition of handwritten Braille, which can be more difficult to decipher than printed Braille.

5. What are some strategies for improving Braille character recognition?

Some strategies for improving Braille character recognition include using high-quality scanners or cameras, using specialized software designed specifically for Braille recognition, and training the software with a large database of Braille characters to improve its accuracy. Additionally, providing proper lighting and contrast can also help improve the recognition process.

Similar threads

  • Programming and Computer Science
Replies
1
Views
7K
Replies
8
Views
1K
  • Programming and Computer Science
Replies
1
Views
2K
Replies
12
Views
2K
  • Programming and Computer Science
Replies
19
Views
6K
  • Science Fiction and Fantasy Media
2
Replies
44
Views
5K
  • Programming and Computer Science
Replies
11
Views
1K
  • Programming and Computer Science
Replies
3
Views
1K
  • Programming and Computer Science
Replies
32
Views
2K
  • Sci-Fi Writing and World Building
Replies
22
Views
2K
Back
Top