| 医学文本图像字符识别校正技术研究与应用 |
| |
| 引用本文： | 李琴,杨斌,郇宝贵,江俊龙.医学文本图像字符识别校正技术研究与应用[J].医学信息学杂志,2018,39(6):56-60. |
| |
| 作者姓名： | 李琴 杨斌 郇宝贵 江俊龙 |
| |
| 作者单位： | 青岛百洋智能科技股份有限公司 青岛 266042,青岛百洋智能科技股份有限公司 青岛 266042,青岛百洋智能科技股份有限公司 青岛 266042,青岛百洋智能科技股份有限公司 青岛 266042 |
| |
| 摘 要： | 针对医学文本图像字符识别的后处理技术进行研究，通过建立医学常用专业词库，基于中文汉字图像的Hog特征和相关系数计算字符之间的文字相似度，从而对识别后的文本进行拼写检查校正，提高字符识别准确率。
|
| 关 键 词： | 医学词库 Hog特征 字符识别 校正 |
| 收稿时间： | 2018/3/5 0:00:00 |
Study and Application of the Recognition and Correction Technology of Medical Text Image Characters |
| |
| LI Qin,YANG Bin,HUAN Bao-gui and JIANG Jun-long.Study and Application of the Recognition and Correction Technology of Medical Text Image Characters[J].Journal of Medical Informatics,2018,39(6):56-60. |
| |
| Authors: | LI Qin YANG Bin HUAN Bao-gui and JIANG Jun-long |
| |
| Institution: | Qingdao Baheal Intelligent Technology Co., LTD, Qingdao 266042, China,Qingdao Baheal Intelligent Technology Co., LTD, Qingdao 266042, China,Qingdao Baheal Intelligent Technology Co., LTD, Qingdao 266042, China and Qingdao Baheal Intelligent Technology Co., LTD, Qingdao 266042, China |
| |
| Abstract: | The paper studies on the post-processing technology of recognition and correction technology of medical text image characters. By building the common medical lexicon, it checks and corrects spelling of the recognized text on the basis of the Hog feature of Chinese character images and word similarity between correlation coefficient calculation characters, so as to enhance the accuracy rate of character recognition. |
| |
| Keywords: | Medical lexicon Hog character Optical Character Recognition (OCR) Correction |
|
| 点击此处可从《医学信息学杂志》浏览原始摘要信息 |
| 点击此处可从《医学信息学杂志》下载免费的PDF全文 |
|
