Embedding a Mathematical OCR Module into OCRopus

在这项研究中，我们的目标是有助于全文OCR系统的发展，也能够读取数学表达式，并嵌入我们的数学OCR模块到开放源码的OCR软件OCRopus开发的DFKI（德国人工智能研究中心）。为了增加我们的模块和系统之间的亲和力，我们回顾了我们的模块的嵌入位置，并利用OCRopus的存在函数。对于数学表达式单独显示为一行的显示表达式，我们的模块能够自动定位并裁剪行，然后将裁剪后的行正确转换为Math-ML或LaTex格式。

In this research, we aim to contribute the development of the full text OCR system that is also able to read mathematical expression, and embed our mathematical OCR module to an open source OCR soft OCRopus developed by DFKI(German Research Center for Artificial Intelligence). To increase the affinity between our module and the system, we reviewed the embedding position of our module and utilized the existence functions of the OCRopus. As for the display expression in which the mathematical expression appears as a line independently, our module is able to locate and crop the line automatically, then the cropped line is correctly converted to the Math-ML or the LaTex format.