Ask Your Question
0

Best OpenCV technique for detect kids handwrite letter

asked 2014-08-09 06:55:09 -0500

kumaramtr gravatar image

Hello,

I'm doing a application to improve kids handwriting using mobile application. In first stage I just want to detect only letters. so I provide the mobile app page with white background and they have to write a letter in black color.

My approach is, when kids write a letter (for ex; 'A') and convert that screen to image, and compare with a template 'A'. so the based on the matching percentage, I want to give them some marks.

That is my initial idea. I did some research but still couldn't found way to identify the matching percentage using template matching technique.

Is template matching good for this kind of application? or another suggestions? What about the feature matching technique? is it better than template matching?

Can someone please give me some idea?

edit retag flag offensive close merge delete

1 answer

Sort by ยป oldest newest most voted
1

answered 2014-08-09 13:06:59 -0500

Guanta gravatar image

One popular choice is to use the shape-context-descriptor in conjunction with the Hausdorff-distance (s. http://www.eecs.berkeley.edu/Research/Projects/CS/vision/shape/sc_digits.html)

Exactly this combination has been integrated in OpenCV's master branch (see github). Have a look at the example samples/cpp/shape_example.cpp (unfortunately not very well commented and formated) and the documentation of the used methods: http://docs.opencv.org/trunk/modules/shape/doc/shape_distances.html

edit flag offensive delete link more
Login/Signup to Answer

Question Tools

2 followers

Stats

Asked: 2014-08-09 06:55:09 -0500

Seen: 352 times

Last updated: Aug 09 '14