Chinese Image Text Recognition on grayscale pixels
Authors:
- 1 May 2014
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 15206149,p. 1380-1384
Abstract
This paper presents a novel scheme for Chinese text recognition in images and videos. It's different from traditional paradigms that binarize text images, fed the binarized text to an OCR engine and get the recognized results. The proposed scheme, named grayscale based Chinese Image Text Recognition (gCITR), implements the recognition directly on grayscale pixels via the following steps: image text over-segmentation, building recognition graph, Chinese character recognition and beam search determination. The advantages of gCITR lie in: (1) it does not heavily rely on the performance of binarization, which is not robust in practical and thus severely affects the performance of OCR, (2) grayscale image retains more information of the text thus facilitates the recognition. Experimental results on text from 13 TV news videos demonstrate the effectiveness of the proposed gCITR, from which significant performance gains are observed.Keywords
This publication has 19 references indexed in Scilit:
- Linear Sequence Discriminant Analysis: A Model-Based Dimensionality Reduction Method for Vector Sequences2013
- ICDAR 2013 Robust Reading Competition2013
- ICDAR 2013 Chinese Handwriting Recognition Competition2013
- ICDAR 2011 Robust Reading Competition - Challenge 1: Reading Text in Born-Digital Images (Web and Email)2011
- ICDAR 2011 Robust Reading Competition Challenge 2: Reading Text in Scene Images2011
- 2DVTE: A two-directional videotext extractor for rapid and elaborate designPattern Recognition, 2009
- An empirical evaluation of deep architectures on problems with many factors of variation2007
- A comprehensive method for multilingual video text detection, localization, and extractionIEEE Transactions on Circuits and Systems for Video Technology, 2005
- Segmentation and recognition of characters in scene images using selective binarization in color space and GAT correlation2005
- Extraction of signatures from check background based on a filiformity criterionIEEE Transactions on Image Processing, 1998