0

I am newbie in text detection ans recognition, but I am confused about how to detect and recognize correctly a text in a image, I founded tessarec but it don't gave me a good result, I used it in a Gui java's Swing's screenshot, what i want to do is to implement my own program, but i need some advice to do that

I want use OpenCv and Python for that

Thanks,

Kamel BOUYACOUB
  • 617
  • 3
  • 7
  • 20
  • It is hard to beat tesseract (python bindings via pytesseract). I think you should try different ways to preprocess your images and play around with the tesseract config files before you try to roll your own. – Paul Brodersen Apr 26 '17 at 09:32
  • I don't mind if my program will be better or not than tessarct, but i just want to learn the algorithms to detect text, I want t understad the methodology and the differents ways to detect and recognize the text in a images – Kamel BOUYACOUB Apr 26 '17 at 10:10
  • for detecting possible text areas, there was a method, something with "stroke width" or sth... – Micka Apr 26 '17 at 18:27

1 Answers1

-1

use google cloud api for computer vision it just detect your text in image and convert it it to text and send back to you json file so you can use that in your project its easiest one and if you want to use in opencv use

OCRTesseract class

and check out this link for text recognition with opencv Python

http://www.danvk.org/2015/01/07/finding-blocks-of-text-in-an-image-using-python-opencv-and-numpy.html

Abubakar
  • 374
  • 3
  • 16