摘要
以专业领域公文和专业领域要图为基础,通过自然语言处理、OCR文字识别、图像特征提取和匹配、Python语言等技术,利用命名实体识别和事件关系抽取等方式对专业领域公文要素进行提取,利用OCR引擎对专业领域要图地名要素进行提取,利用机器学习算法中的SIFT特征提取算法、暴力匹配等算法对专业领域要图地物要素进行提取,最后依托自动化的底图添加操作,初步构建一个标图辅助标绘模型,探索标图作业人员提升标图效率的新方法。
This article takes professional field documents and expertise base map as the foundation,and uses natural language processing,OCR text recognition,image feature extraction and matching,Python language,and other technologies to extract elements from professional field documents through named entity recognition and event relationship extraction.It also uses OCR engines to extract geographical names from expertise base map,and extracts geographical features from expertise base map using SIFT feature extraction algorithms and brute force matching algorithms in machine learning algorithms.Finally,relying on automated map addition operations,this article initially builds a marking assistance model for marking maps,exploring new methods for improving marking efficiency for marking workers.
作者
兰嵩
郑雄
雷肖玲
郭安业
丁一
Lan Song;Zheng Xiong;Lei Xiaoling;Guo Anye;Ding Yi(Fujian Armed Police Corps,Fuzhou 350000,China)
出处
《网络安全与数据治理》
2023年第S01期207-211,共5页
CYBER SECURITY AND DATA GOVERNANCE
关键词
文本抽取
特征匹配
专业领域公文
专业领域要图
辅助标绘
text extraction
feature matching
professional field documents
expertise base map
auxiliary marking
