Technical Papers Parallel Session-V: Character recognition in natural scene images
Abstract/Description
Applications like content-based image indexing, automatic geocoding of businesses and real-time robotic navigation have generated research interest in the problem of text reading from natural images. This paper tackles the problem of character recognition which constitutes the last phase of the text detection and extraction pipeline. We employ the stroke width transform used for the text detection phase, and address the problem of character recognition by training KNNs, Random Forest and Neural Networks on the Char74 and ICDAR2003 datasets of images in-the-wild, containing English characters from natural images. We also compare the performance of our character classifier from the currently best-performing state-of-the-art OCR and demonstrate that both perform competitively.
Keywords
Text recognition, Character recognition, Images in-the-wild, Supervised learning, Natural images
Location
C-10, AMAN CED
Session Theme
Technical Papers Parallel Session-V (Information Retrieval)
Session Type
Parallel Technical Session
Session Chair
Dr. Imran Hayee
Start Date
13-12-2015 2:30 PM
End Date
13-12-2015 2:50 PM
Recommended Citation
Akbani, O., Gokrani, A., Khan, F., Behlim, S. I., & Syed, T. Q. (2015). Technical Papers Parallel Session-V: Character recognition in natural scene images. International Conference on Information and Communication Technologies. Retrieved from https://ir.iba.edu.pk/icict/2015/2015/24
COinS
Technical Papers Parallel Session-V: Character recognition in natural scene images
C-10, AMAN CED
Applications like content-based image indexing, automatic geocoding of businesses and real-time robotic navigation have generated research interest in the problem of text reading from natural images. This paper tackles the problem of character recognition which constitutes the last phase of the text detection and extraction pipeline. We employ the stroke width transform used for the text detection phase, and address the problem of character recognition by training KNNs, Random Forest and Neural Networks on the Char74 and ICDAR2003 datasets of images in-the-wild, containing English characters from natural images. We also compare the performance of our character classifier from the currently best-performing state-of-the-art OCR and demonstrate that both perform competitively.
