Is there a framework for tests and research purpose, that contain the various levels of document recognition (preprocessing, segmentation, feature extraction, recognition and post-processing) allowing us to develop, test, compare and improve algorithms and contributions.
I found this http://gamera.informatik.hsnr.de/ , but it does not incorporate all the stages.
Thanks in advance
You could use https://github.com/janantala/handwriting-recognition and combine it with post processing step - smoothing using n-gram language models (e.g. SRILM Tool)
Please read my chapter:
Soft Clustering: Algorithms and Validity of Solutions
Fuzzy Computing, pp. 445-472, 1988.
It has a better approach and nine validity measures.
hi
refer the following artcile
http://www.cvc.uab.es/icdar2009/papers/3725a121.pdf
http://www.ijcaonline.org/journal/number18/pxc387572.pdf
- Badges
- Science topic
- Similar topics
- Engineering
- Electrical Engineering
- Signal Processing
- Image Processing