I am currently working on a project on Handwriting Matching algorithm where I am given some test data (images with text handwritten in English Language ) . I have to determine by matching this data from the training data that for each test data , which person from the training set has most probable similar writing . I searched internet for papers related to this but I found only papers based on text extraction in which we have to determine what is written rather than determine who has written .
I would be very helpful if I can get some insights on it and also some references or journals or papers so that I can work on an evolutionary algorithm rather than doing everything from scratch. It would be very useful If I can also get the links for paid papers/journals .
Asked
Active
Viewed 113 times
0
Kavish Dwivedi
- 735
- 6
- 24
-
1A quick search turned up that you're looking for writer identification. Scholar yields papers with quite a high reference count: [Automatic signature verification and writer identification — the state of the art](http://www.sciencedirect.com/science/article/pii/0031320389900599). I never did this myself, so you'll have to see through it yourself and come back with more specific questions – pdresselhaus Nov 14 '13 at 15:52
1 Answers
0
In the last years, Kaggle hosted contests with the goal of identifying a writer. The contests were based on Arabic alphabet but I guess you could apply similar techniques to English letters. You will have to dig the forums and results from these contests to actually find links to helpful ideas/papers.
rold2007
- 1,297
- 1
- 12
- 25