Text Recognition
Character Spotting
Participants: Nishatul Majid (Fort Lewis College), Elisa Barney Smith (Luleå Tekniska Universitet)
Rather than segmenting a word into characters and recognizing the characters, we have adopted the approach of looking for every character in a word and based on detections identifying the resulting string. This works well for overlapping characters, and does not require a one-dimensional arrangement of the characters. This was developed for Bangla, but has also been successfully tested on Korean and English. Work is in progress to re-implement the algorithm using Python and switch the object spotting engine to YOLO. Code, YOLO weights and data will be publicly shared.
Publications:
Nishatul Majid, Elisa H Barney Smith, “Character spotting and autonomous tagging: offline handwriting recognition for Bangla, Korean and other alphabetic scripts,” International Journal on Document Analysis and Recognition (IJDAR), pp 1-19, 2022.
R Mondal, S Malakar, Elisa H Barney Smith, R Sarkar, “Handwritten English word recognition using a deep learning based object detection architecture,” Multimedia Tools and Applications, v 81, n 1, p 975-1000, January 2022
N Majid, "Offline Bangla Handwriting Recognition with Sequential Detection of Characters/Diacritics," Doctoral Dissertation, Boise State University, 2020.
S Kim, EH Barney Smith, N Majid, "Segmentation-Free Korean Handwriting Recognition Using Neural Network Training, Undergraduate Research Conference Poster, 2020
N Majid, EH Barney Smith, "Segmentation-free Bangla offline handwriting recognition using sequential detection of characters and diacritics with a Faster R-CNN," International Conference on Document Analysis and Recognition (ICDAR), 2019
N Majid, EH Barney Smith, "Performance Comparison of Scanner and Camera-Acquired Data for Bangla Offline Handwriting Recognition," Workshop on Camera-Based Document Analysis and Recognition (CBDAR), 2019
N Majid, EH Barney Smith, "Introducing the Boise State Bangla Handwriting dataset and an efficient offline recognizer of isolated Bangla characters," 16th International Conference on Frontiers in Handwriting Recognition (ICFHR), Niagara Falls, New York, USA, 2018 (publicly released dataset)
Online Handwriting Recognition
Particpants: Sukhdeep Singh, Elisa H. Barney Smith
Online text recognition considers the time sequence of the stroke points as well as their position. Tablets and other touch devices often oversample to assure they have adequate stroke information. Not all these points are necessary. We developed an approach that selects a subset of the points to enable a simpler training, while maintaining high recognition accuracy.
Publications:
Sukhdeep Singh, Vinod Kumar Chauhan, Elisa H. Barney Smith, “A self controlled RDP approach for feature extraction in online handwriting recognition using deep learning,” Applied Intelligence, Springer Nature, February 2020
Other Text Recognition Projects
There are many things that can be done to improve text recognition document analysis. I have been fortunate to collaborate with several people on assorted projects over the years. Here are some of the results of those collaborations.
Publications:
Pawan Kumar Singh, Iman Chatterjee, Ram Sarkar, Elisa H Barney Smith, Mita Nasipuri, “A new feature extraction approach for script invariant handwritten numeral recognition,” Expert Systems, v 38, n 6, e12699, 2021.
Suman Kumar Bera, Akash Chakrabarti, Sagnik Lahiri, Elisa H Barney Smith, Ram Sarkar, “Normalization of unconstrained handwritten words in terms of Slope and Slant Correction," Pattern Recognition Letters, September 2019
T. Hoang, E. H. Barney Smith, A. Tabbone, “Sparsity-based edge noise removal from bilevel graphical document images,” International Journal of Document Analysis and Recognition, Springer Verlag, Vol. 17, No. 2, June 2014, pp. 161-179.
Thai V. Hoang, Elisa H. Barney Smith, Salvatore Tabbone, “Edge Noise Removal in Bilevel Graphical Document Images Using Sparse Representation,” International Conference on Image Processing, Brussels, Belgium, September 2011, Paper #WP.PE.7.
L. Likforman-Sulem, J. Darbon, E.H. Barney Smith, “Enhancement of historical printed document images by combining Total Variation regularization and Non-Local Means filtering,” Image and Vision Computing, Elsevier, Vol. 29, No. 5, 2011, pp. 351-363.
Amy Winder, Tim Andersen, Elisa H. Barney Smith, “Extending Page Segmentation Algorithms for Mixed-Layout Document Processing,” Proc. International Conference on Document Analysis and Recognition, Beijing, China, September 2011, pp 1245-1249.
C. J. Stanger, Thanh Tran and Elisa H. Barney Smith. “Descreening of Color Halftone Images in the Frequency Domain,” Proc. SPIE Electronic Imaging - Color Imaging XVI: Displaying, Processing, Hardcopy, and Applications, San Francisco, CA, January 2011, Vol. 7866.
E. H. Barney Smith and J. Darbon and L. Likforman-Sulem. “A Mask-based enhancement method for historical documents,” Proc. SPIE Electronic Imaging - Document Recognition and Retrieval XVIII, San Francisco, CA, January 2011, Vol. 7874.
L. Likforman, J. Darbon, and E. H. Barney Smith, “Pre-processing of Degraded Printed Documents by Non-Local means and Total Variation,” International Conference on Document Analysis and Recognition 2009, Barcelona, Spain, 26-29 July 2009.
P. Garrison, D. Davis, T. Andersen, E. H. Barney Smith, “Study of style effects on OCR errors in the MEDLINE database,” Proc. SPIE Electronic Imaging, Document Recognition and Retrieval XII, Vol. 5676, Santa Clara, CA, January 2005, pp. 28-37