Mathematical Formulas Extraction

As a universal technical language, mathematics has been widely applied in many fields, and it is more accurate than any other languages in describing information. Therefore, numerous mathematical formulas exist in all kinds of documents. There is no doubt that automatic mathematical formulas processing is very important and necessary, of which extract formulas from document images is the first step. In this paper, formulas extraction methods which are not based on recognition results are presented: isolated formulas are extracted based on Parzen window and embedded expressions are extracted based on 2-D structures detection. Experiments show that our methods are very effective in formulas extraction.

• Type: Conference
• Year: 2003
• Where: ICDAR
• Authors: Jianming Jin, Xionghu Han, Qingren Wang
• Keywords: Document Analysis, Formulas, Formulas Extraction, ICDAR 2003, Mathematical Formulas