Abstract | ||
---|---|---|
India is a multi-lingual, multi-script country. Considerably less work has been done towards handwritten character recognition of Indian languages than for other languages. In this paper we propose a quadratic classifier based scheme for the recognition of off-line handwritten characters of three popular south Indian scripts: Kannada, Telugu, and Tamil. The features used here are mainly obtained from the directional information. For feature computation, the bounding box of a character is segmented into blocks, and the directional features are computed in each block. These blocks are then down-sampled by a Gaussian filter, and the features obtained from the down-sampled blocks are fed to a modified quadratic classifier for recognition. Here, we used two sets of features. We used 64-dimensional features for high speed recognition and 400-dimensional features for high accuracy recognition. A five-fold cross validation technique was used for result computation, and we obtained 90.34%, 90.90%, and 96.73% accuracy rates from Kannada, Telugu, and Tamil characters, respectively, from 400 dimensional features. |
Year | DOI | Venue |
---|---|---|
2006 | 10.1007/978-3-540-78199-8_15 | SACH |
Keywords | Field | DocType |
directional information,directional feature,high accuracy recognition,accuracy rate,popular south indian script,feature computation,handwritten character recognition,tamil character,high speed recognition,down-sampled block,indian language | Gaussian filter,Tamil,Pattern recognition,Speech recognition,Multilayer perceptron,Artificial intelligence,Engineering,Intelligent word recognition,Telugu,Quadratic classifier,Chain code,Minimum bounding box | Conference |
Volume | ISSN | ISBN |
4768 | 0302-9743 | 3-540-78198-6 |
Citations | PageRank | References |
12 | 0.68 | 18 |
Authors | ||
4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Umapada Pal | 1 | 1477 | 139.32 |
Nabin Sharma | 2 | 132 | 11.55 |
Tetsushi Wakabayashi | 3 | 361 | 43.25 |
Fumitaka Kimura | 4 | 584 | 67.24 |