Title :
Improving OCR accuracy through combination: a survey
Author :
Handley, John C.
Author_Institution :
Xerox Corp., Webster, NY, USA
Abstract :
Optical character recognition is perhaps the most studied application of pattern recognition. Recent work has increased accuracy in two ways. Combination of individual classifier outputs overcomes deficiencies of features and trainability of single classifiers. OCR systems take page images as input and output strings of recognized characters. Due to character segmentation errors, characters can be split or merged preventing output combination character-by-character. Merging of output strings is done using string alignment algorithms
Keywords :
image classification; image segmentation; optical character recognition; OCR; character segmentation errors; individual classifier output combination; optical character recognition; page images; string alignment algorithms; Bagging; Boosting; Character recognition; Classification tree analysis; Image recognition; Image segmentation; Neural networks; Optical character recognition software; Pattern recognition; Voting;
Conference_Titel :
Systems, Man, and Cybernetics, 1998. 1998 IEEE International Conference on
Conference_Location :
San Diego, CA
Print_ISBN :
0-7803-4778-1
DOI :
10.1109/ICSMC.1998.727527
