Title :
An Improved method for segmentation of touching symbols in printed mathematical expressions
Author :
Zhang, Dong-Yu ; Tian, Xue-dong ; Li, Xin-fu
Author_Institution :
Coll. of Math. & Comput. Sci., Hehei Univ., Baoding, China
Abstract :
The segmentation of touching symbols is one of the key factors which affect the performance of printed mathematical expression recognition system. An Improved method for segmentation of touching symbols in printed mathematical expressions is presented. This method is suitable for different types of touching symbols. Firstly, the outer contour of the symbol image is extracted based on contour tracing algorithm. Next, the concave corner points are detected by corner detection algorithm. These concave corner points are considered as the candidate segmentation points. Finally segmentation paths are constructed to achieve the segmentation of the touching symbol.
Keywords :
edge detection; feature extraction; image segmentation; concave corner point detection; contour tracing algorithm; corner detection algorithm; printed mathematical expression recognition system; symbol image outer contour extraction; touching symbol segmentation; Character recognition; Computational intelligence; Computer science; Detection algorithms; Educational institutions; Image segmentation; Laboratories; Machine learning; Mathematics; Optical character recognition software; Mathematical expression recognition; OCR; Symbol segmentation; Touching symbol segmentation;
Conference_Titel :
Advanced Computer Control (ICACC), 2010 2nd International Conference on
Conference_Location :
Shenyang
Print_ISBN :
978-1-4244-5845-5
DOI :
10.1109/ICACC.2010.5486679
