OCR for devnagri scripts

OCR for devnagri scripts

Sumedh Pendurkar

Sumedh Pendurkar

Pune, Maharashtra

detecting text from printed documents in devnagri scripts.

Intel RealSense™

  • 0 Collaborators

  • 1 Followers

    Follow

Description

Opencv library was used for image processing. After removal of speckle, pepper noise and binarization contours in the images where obtained. These contours were the words as shirorekha is present above all characters. Each word was deskewed using hough line transformations and the word was divided into three parts. And shirorekha was removed The features used for classification were the raw pixels. As of now, basic characters(only क ख ग..) segmented by this process were tested using SVM with linear kernel (130 fonts). This was testing on similar fonts and the accuracy was found about to be around 95%

Gallery

Links

github link

Medium 0 0wmyd owtnjesu7fwikzsqeehlhf4bt e8kzjuff8wlbwu22m2kuszudrlefwc82i7kggevditwbvqsfeghavdsfntwfvqj7ighy7 q73u4avau6iq4g5qqh4l

Sumedh P. created project OCR for devnagri scripts

Medium 9d0d3da7 7767 48c3 86c7 6faf9c61ae94

OCR for devnagri scripts

Opencv library was used for image processing. After removal of speckle, pepper noise and binarization contours in the images where obtained. These contours were the words as shirorekha is present above all characters. Each word was deskewed using hough line transformations and the word was divided into three parts. And shirorekha was removed The features used for classification were the raw pixels. As of now, basic characters(only क ख ग..) segmented by this process were tested using SVM with linear kernel (130 fonts). This was testing on similar fonts and the accuracy was found about to be around 95%

No users to show at the moment.