| Status | 已發表Published |
| Two-dimensional PHMM for handwriting digits recognition | |
| Lin Dong; Chen X.X.; Tang Y.Y. | |
| 1996-12-01 | |
| Source Publication | International Conference on Signal Processing Proceedings, ICSP
 |
| Volume | 2 |
| Pages | 1316-1319 |
| Abstract | In this paper we used two-dimensional Pseudo Hidden Markov Models (2D PHMM) for handwriting digits recognition. We used 2D Pseudo Viterbi algorithm training the model. This algorithm compared to traditional Forward-Backward algorithm has some advantage as follow. 1) It can get rig of down over-flow that maybe happened in Forward-Backward algorithm. 2) As the algorithm used Log calculation the accurate of the calculation is much higher than Forward-Backward algorithm. 3) The speed of calculation is much faster than Forward-Backward algorithm. The data base of handwriting digits used in this paper was collected at Beijing Postal Center, the digits was scanned from letters zip code, altogether 4000 hand writing digits, 2000 are used for training set 2000 are used for testing set. We used pixel value as recognition feature. The result of the test shows, the recognition rate of training set is 95%. The recognition rate of testing set is 92.6%. We are horizontal direction and vertical direction as super plane to build two types model at same time, and at the recognition phase if the result of the two models are different as rejection condition, the recognition rate of training set is 98%, the recognition rate of testing set is 96%. If we used some feature which is rotate and move invariant the recognition rate may get higher. |
| URL | View the original |
| Language | 英語English |
| Fulltext Access | |
| Document Type | Conference paper |
| Collection | University of Macau |
| Affiliation | Beijing University of Posts and Telecommunications |
| Recommended Citation GB/T 7714 | Lin Dong,Chen X.X.,Tang Y.Y.. Two-dimensional PHMM for handwriting digits recognition[C], 1996, 1316-1319. |
| APA | Lin Dong., Chen X.X.., & Tang Y.Y. (1996). Two-dimensional PHMM for handwriting digits recognition. International Conference on Signal Processing Proceedings, ICSP, 2, 1316-1319. |
| Files in This Item: | There are no files associated with this item. | |||||
Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.
Edit Comment