1 / 16

Handwritten Character Recognition using Elastic Matching and PCA

Handwritten Character Recognition using Elastic Matching and PCA. Vanita Mane, Lena Ragha International Conference on Advances in Conputing , Communication and Control. Reporter: 資訊所 P78991121 Yung-Chih Cheng ( 鄭詠之 ). Outline. Introduction Data Collection System Architecture

red
Download Presentation

Handwritten Character Recognition using Elastic Matching and PCA

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Handwritten Character Recognition using Elastic Matching and PCA Vanita Mane, Lena Ragha International Conference on Advances in Conputing, Communication and Control Reporter: 資訊所 P78991121 Yung-Chih Cheng (鄭詠之)

  2. Outline • Introduction • Data Collection • System Architecture • Feature Extraction • Recognition Methods • Results and Discussion • Conclusion

  3. Introduction (1/2) • Character recognition is becoming more and more important in the modern world. • Handwritten recognition is not a new technology • Optical Character Recognition(OCR) • Automatic reading of optically sensor • Document text materials to translate human-readable character to machine readable codes

  4. Introduction (2/2) • Less work has been reported for the recognition of Indian Languages. • Complexity of the shape • Large set of different patterns • Propose a new elastic image matching technique based on an eigen-deformation • Offline isolated English uppercase handwritten characters • Offline isolated handwritten character of Devnagari

  5. Data Collection-English • ETL6 standard database is used • Japanese Characters • English Capital letters [A-Z] • Digits [0-9] • .pgn file format • Image size: 64 x 63 • Collect 500 data samples from different individuals of various professions for the experimetn.

  6. Data Collection-Devnagari • Devnagari is the most popular script in Indian and the most popular Indian language Hindi is written an Devnagari script • National language of India • The third most popular language in the world • 14 vowels and 33 consonants • Left to right • In this paper, considering 12 basic character

  7. System Architecture (1/2)

  8. System Architecture (2/2) • Preprocessing • Filtering • Morphological operation • Normalization • Segmentation • Horizontal Segmentation • Vertical Segmentation • Size Normalization

  9. Feature Extraction (1/2) • Elastic Matching

  10. Feature Extraction (2/2) • Properties of Elastic Matching(EM) • Anisotropic • Asymmetric • The distance between image T and R • R is deformed by F for optimal matching T and R • Symmetric • Sum of two asymmetric distances is simplest and gives result than individual asymmetric elastic matching tecnique

  11. Recognition Methods (1/2) • With Eigen-deformations(PCA) • Training Phase • Recognition Phase

  12. Recognition Methods (2/2) • Without Eigen-deformation

  13. Results and Discussion (1/2)

  14. Results and Discussion (2/2)

  15. Conclusion (1/2) • A elastic matching with PCA based system towards the recognition of off-line isolated uppercase English character and Devnagari handwritten character • Depend on the characteristics of the elastic matching emplyed to collect actual deformations

  16. Conclusion (2/2) • The distribution of the actual deformation is not isotropic and therefore can be approcimated by several principal axes.

More Related